Adenomyosis (ADS) is one of the most challenging problems in benign uterine disease. This is because the mechanisms associated with this condition remain unclear and there are no unified and effective treatment options at present. The Endometriosis Committee of the Chinese Medical Doctor Association of Obstetricians and Gynecologists defines ADS as an eutopic endometrium that invades and migrates into the uterine myometrium, a condition that affects 7–23% of women of reproductive age (García-Solares et al., 2018; Leng et al., 2020). The recognized mechanisms of ADS include the invasion of the endometrial basalis into the myometrium, microtrauma of the endometrial-myometrial interface, genetics, immune disorders, and inflammatory mediators (An et al., 2017; Vannuccini et al., 2017; Guo, 2020; Zhai et al., 2020; Guo et al., 2022). However, the clinical presentations of patients with ADS are heterogeneous, thus indicating that unknown pathological mechanisms may exist (Bourdon et al., 2021).

Several studies have proposed that ADS is associated with an imbalance in the vagina microbiota (Chen et al., 2020; Chao et al., 2021; Kunaseth et al., 2022). Lactobacillus is known to dominate the bacterial composition of a healthy vagina and deters the growth of opportunistic pathogens; however, dysbiosis can potentially lead to the occurrence of disease. Chronic pelvic pain (CPP) has been associated with both endometriosis (EM) and ADS. The vaginal microbiota of patients with EM/ADS-associated CPP is known to feature a higher abundance of Clostridium butyricum, Clostridium disporicum, Alloscardovia omnicolens, and Veillonella montpellierensis (Chao et al., 2021). In a previous study, Kunaseth et al. (2022) reported that Alloscardovia, Oscillospirales, Ruminoccoccaceae, Oscillospiraceae, Enhydrobacter, Megamonas, Selenomonadaceae, and Faecalibacterium all showed an increased abundance in the vaginal microbiota of patients with ADS. Atopobium is a known biomarker for patients with EM combined with ADS of the vaginal microbiota (Chen et al., 2020).

The uterus has long considered as a sterile environment. However, several studies have shown that the endometrium is colonized by bacterial strains that differ from the vaginal microbiota (Moreno et al., 2016; Chen et al., 2017). Dysbiosis of the endometrial microbiota is likely to lead to endometrial cancer (Walther-António et al., 2016; Sobstyl et al., 2022), polyps (Fang et al., 2016), endometriosis (Akiyama et al., 2019; Wessels et al., 2021), endometritis (Moreno et al., 2018; Hernandes et al., 2020), and recurrent pregnancy loss (Łaniewski et al., 2020). A recent study confirmed that Porphyromonas somerae represents a suitable biomarker for endometrial cancer (Walsh et al., 2019). Fang et al. (2016) found that Lactobacillus, Bifidobacterium, Gardnerella, and Streptococcus were increased in women with endometrial polyps. A previous study found that the relative abundances of the Streptococcaceae, Moraxellaceae, Staphylococcaceae, and Enterobacteriaceae families were significantly increased among females with endometriosis (Akiyama et al., 2019). However, research on the endometrial microbiota in patients with ADS is rare. In this study, we aimed to investigate the composition and biomarkers of the microbiota in endometrial samples from females with and without ADS using 5R 16S rRNA gene sequencing (Nejman et al., 2020; Poore et al., 2020). We also investigated whether biomarkers of the endometrial microbiota could be used to diagnose ADS and then predict the potential interactions between microbiota and ADS.

The study was performed according to the guidelines of the Declaration of Helsinki and was approved by the Ethics Committee of the Beijing Obstetrics and Gynecology Hospital, Capital Medical University No. 2016-KY-012-02.

In this study, we compared the endometrial microbiota from patients with and without adenomyosis, as confirmed by pathological analysis. We demonstrated the presence of an endometrial microbiota in the uteri of patients with adenomyosis that had a significantly lower richness than the control group. Until now, we know little about the precise etiology involved in the reduced richness of the endometrial microbiota in patients with adenomyosis. There were no significant differences between the two groups of patients with regard to their age, BMI, menstrual cycle phase, parity, and methods of contraception. In addition, we excluded patients with autoimmune disorders and genital infections. In contrast to our findings (mean age > 45 years old), one previous study found that the richness of the endometrial flora in patients with endometriosis (mean age < 36 years old) was increased (Wessels et al., 2021); this finding might be due to differences in participant age. Wang et al. (2021) reported that the richness of the uterine microbiome in the 40–50-year age group was lower than that in younger women. To a certain extent, such changes in the diversity of the endometrial microbiota may be the result of changes in the endometrial immune microenvironment, including endometrial immune response and inflammation (abnormal activation and differentiation of dendritic cells, macrophages, and Natural killer cells) which ultimately reduce endometrial receptivity, disrupt embryo implantation, and promote the progression of disease (Wira et al., 2005; Chen et al., 2021).

Furthermore, we found that the majority of samples could be clustered according to disease status by performing beta diversity analysis; this indicated that the structure and composition of the endometrial microbiota of patients with or without adenomyosis were different. LEfSe analysis also identified some biomarkers that were strictly associated with adenomyosis that might be useful for the diagnosis of adenomyosis. A high abundance of Firmicutes and Actinomycetes was detected in endometrial samples from patients with adenomyosis, whereas Proteobacteria and Thermotogae were present in high abundance in patients from the control group. These results are consistent with previous studies of endometrial cancer, intrauterine adhesions, endometrial polyps, and patients with infertility (Fang et al., 2016; Benner et al., 2018; Walsh et al., 2019; Qiu et al., 2021). In a previous study, Khan et al. studied patients with endometriotic cysts and observed a higher proportion of Corynebacterium (Actinomycetes, Gram-positive), Enterococcus (Firmicutes, Gram-positive), Staphylococcus (Firmicutes, Gram-positive), Caulobacter (Proteobacteria, Gram-negative), and Citrobacter (Proteobacteria, Gram-negative) in patients with adenomyosis (Akiyama et al., 2019). However, Chen et al. (2017) reported that Methylophilaceae (Proteobacteria, Gram-negative) was a useful biomarker for the endometrial flora in patients with adenomyosis. In the present study, blank and negative controls were used to eliminate potential bacterial contamination from sampling collection, DNA processing, library preparation, and from the local environment. Differences between our results and those published previously may be due to analytical differences, along with different collection methods and study populations. Moreover, our results align with the bacterial contamination hypothesis of endometriosis that Gram-negative bacteria show enrichment in the endometrium of patients with endometriosis. Lipopolysaccharide (LPS) is a component of the outer membrane of Gram-negative bacteria and regulates the uncontrolled activation of inflammatory and immune responses in the pelvis and the growth of endometriosis via the LPS/TLR4 cascade (Khan et al., 2018). The major taxa of relative richness (TOP 10) represented in the population of patients with adenomyosis were non-Lactobacillus dominant (Lactobacillus species <90%, NLD), including Lactobacillus, Corynebacterium, Streptococcus, Prevotella, Hyphomicrobium, Geobacillus (3.78%), and Vibrio. These results are consistent with those reported previously by Chen et al. (2017) but differed from several other studies which reported that the most dominant endometrial microbiota are Lactobacillus (Lactobacillus species >90%, LD) (Moreno et al., 2016; Tao et al., 2017; Kyono et al., 2018).

Although the precise mechanisms underlying the endometrial microbiota with regard to human health and disease remain unclear, an imbalance of the microbiota is closely related to the pathophysiological mechanisms of endometriosis, endometrial cancer, and uterine fibroids. Currently, most of our knowledge relating to the interaction between bacteria and their host is derived from the highly diverse intestinal microbiota (Belkaid and Hand, 2014). The diversity of the endometrial microbiota is lower than that of the intestinal microbiota and higher than that of the vaginal microbiota (Chen et al., 2017). At the species level, our results indicated that Citrobacter freundii (Gram-negative), Burkholderia cepacia (Gram-negative), and Prevotella copri (Gram-negative) could be biomarkers for adenomyosis. Citrobacter freundii is a symbiotic bacterium found in the intestines of humans and animals (Liu et al., 2015). However, another study reported that Citrobacter freundii activated inflammation via the nucleotide oligomeric Domain-Like receptor family, a pyridine domain containing 3 (NLRP3) inflammatory body, in a T6SS-dependent manner (Liu et al., 2021). Citrobacter freundii can also invade, reproduce, and transfect human microvascular endothelial cells in vitro (Badger et al., 1999); this was associated with neonatal meningitis and brain abscesses (Joaquin et al., 1991). Burkholderia cepacia was previously shown to secrete BceF Tyrosine Kinase which participates in the formation biofilm and contributes to respiratory tract infection. BceF tyrosine kinase does not show any homology with human kinase and is therefore useful for the development of target antibacterial drugs (Mayer et al., 2021). Another study revealed that ‘quorum sensing’ (QS) enabled Burkholderia cepacia to produce virulent factors and biofilms, thus destroying the host’s innate immunity, especially epithelial cells, neutrophils, and megaphagocytes in patients with cystic fibrosis (Ganesh et al., 2020). Similar results were observed in patients with keratitis (Ho et al., 2021). Wessels et al. also reported that Burkholderia enriched the endometrium of symptomatic patients with pelvic pain but without endometriosis, as confirmed by surgery (Wessels et al., 2021).

In addition, Prevotella copri, a potential pro-inflammatory bacterium, has been associated with inflammation and autoimmunity. In a mouse model of glucan sulfate-induced colitis, elevated levels of Prevotella copri richness were associated with more severe intestinal inflammation than mice without Prevotella copri (Cao et al., 2020). Prevotella copri was also positively correlated with the severity of rheumatoid arthritis (Scher et al., 2013). Recent studies have revealed that Prevotella copri play an important role in anti-tumor therapy; patients with a high abundance of Prevotella copri could benefit from treatment (Jin et al., 2019). In addition, Yu et al. (2019) proved that Prevotella copri caused damage to the mucosal barrier by expressing reduced levels of tight junction proteins (ZO-1, Occludin, and Claudin-1) and increased expression levels of inflammatory factors (IFN-γ, IL-1 β, IL-6, and MCP-1). The eutopic endometrium is thought to play a vital role in the pathogenesis of adenomyosis. Studies have predicted that the IL-6 and ERK/MAPK pathways are abnormally activated in the eutopic endometrium of patients with adenomyosis (Xiang et al., 2019). Furthermore, the numbers of natural killer cells and macrophages have been found to be elevated in the eutopic endometrium of patients with adenomyosis along with increased levels of inflammatory markers (IL-1b and IL-6) (Zhihong et al., 2016; Carrarelli et al., 2017).

Although the role of the microbiota has been revealed in some diseases, we explored the potential mechanisms that may act between the host and microbes by applying the PICRUSt algorithm. Differentially regulated pathways between the two groups that may be of relevance included genetic information processing and metabolism; these were both enriched in the ADS group when compared to the control group. Furthermore, nucleotide metabolism, replication and repair, translation, enzyme families, and carbohydrate metabolism pathways were all upregulated in the ADS group when compared to the control group. Furthermore, protein export, glycolysis/gluconeogenesis, alanine aspartate, and glutamate metabolism were all enriched in the ADS group when compared to the control group. Only limited research has targeted the endometrial microbiota of women with adenomyosis. Citrobacter freundii, Burkholderia cepacia, and Prevotella copri are pathogenic microbiota that may be associated with adenomyosis; these are Gram-negative bacteria that secrete LPS via protein export, which causes mucosal inflammation and damages the mucosal barrier integrity (Mobley et al., 2001). Microorganisms and their metabolites can also trigger DNA damage and mutation of the epithelial cells and directly regulate the biological function of host cells via increased proliferation or reduced apoptosis (Lim et al., 2017; Chattopadhyay et al., 2019; Helmink et al., 2019).

Our study has several strengths that should be considered. We used a blinded approach to ensure the accuracy in all statistical analyses. We also took multiple methods to minimize the contamination. Furthermore, all patients in this study received pathological examination to guarantee the precision and accuracy of diagnosis. However, this study has several limitations that should be considered. First, the patients in this study were ethnically and genetically homogenous. Second, the endometrial microbiota was susceptible to contamination as a specimen with low microbial abundance. Researchers tried to avoid contamination at all times during sampling; however, environmental contamination cannot be ruled out. Third, all participants received hysterectomy in this study; therefore, we were unable to recruit healthy samples for the control group; this would have provided a better understanding of the differences in endometrial microbiota between the patients with adenomyosis and healthy women. Finally, another limitation of this study is its small sample size; however, other studies on similar topics also involved small sample sizes. In the future, studies should aim to recruit a larger population size for investigation.

In conclusion, females possess an endometrial microbiota that contains various microbes. Adenomyosis is highly associated with endometrial bacterial dysbiosis. A potential variation in the microbial environment of the endometrium might contribute to adenomyosis. Pathogenic bacteria, such as Citrobacter freundii, Burkholderia cepacia, and Prevotella copri, could be considered as potential emerging causes of adenomyosis. Further studies should investigate the cross-talk between potential pathogenic microbiota in patients with adenomyosis, evaluate whether the early diagnosis of adenomyosis could be assessed according to the endometrial microbiota profile, and investigate whether the endometrial microbiota could be modified to increase treatment outcomes.

The data presented in the study are deposited in the Sequence Read Archive repository (https://www.ncbi.nlm.nih.gov/sra/), accession number PRJNA891223.

The studies involving human participants were reviewed and approved by Medical Ethics Committee of Beijing Obstetrics and Gynecology Hospital, Capital Medical University No. 2016-KY-012-02. The patients/participants provided their written informed consent to participate in this study.

QL, HD, ShW, ZG, and YC conceived the study and reviewed and edited the manuscript. QL, SiW, CC, MS, HS, and CZ participated in sample collection. QL and ZG conducted the data analysis and prepared the figures and tables. QL helped to write the original draft. All authors contributed to the article and approved the submitted version.

This research was funded by the National Natural Science Foundation of China (grant number 81571412) and supported by the Beijing Obstetrics and Gynecology Hospital, Capital Medical University (grant number FCYY201920).

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.