Vaginal microbiome differences between patients with adenomyosis with different menstrual cycles and healthy controls

Background Adenomyosis is a commonly observed benign gynecological disease that affects the quality of life and social psychology of women of childbearing age. However, because of the unknown etiology and incidence of adenomyosis, its pathophysiological mechanism remains unclear; further, because no noninvasive, accurate, and individualized diagnostic methods are available, treatment and efficacy evaluations are limited. Notably, the inter- action between the changes in the microecological environment of the female reproductive tract and human immu- nity, endocrine, and other links leads to the occurrence and development of diseases. In addition, the vaginal micro- biome differs in different menstrual cycles; therefore, assessing the differences between the microbiomes of patients with adenomyosis and healthy individuals in different menstrual cycles will improve the understanding of the disease and provide references for the search for noninvasive diagnosis and individualized precision treatment of adenomyo- sis. This study aimed to explored the data of individuals in different menstrual cycles.

Differences in the vaginal microbiome between patients with adenomyosis and healthy individuals were observed. At phylum level, the relative abundance of Firmicutes in the adenomyosis group was higher than that in the control group, which contributed the most to the species difference between the two groups. At the genus level, Lactobacillus was the most dominant in both groups, Alpha-diversity analysis showed significant differences in the adenomyosis and control group during luteal phase (Shannon index, p = 0.0087; Simpson index, p = 0.0056). Beta-diversity index was significantly different between the two groups (p = 0.018). However, based on Weighted Unifrac analysis, significant differences were only observed throughout the luteal phase (p = 0.0146). Within the adeno- myosis group, differences between women with different menstrual cycles were also observed. Finally, 50 possible biomarkers including were screened and predicted based on the random forest analyse.

The vaginal microbiome of patients with adenomyosis and healthy individuals differed during men- strual periods, especially during the luteal phase. These findings facilitate the search for specific biological markers *Correspondence: Ming Yuan [email protected] Guoyun Wang [email protected] Full list of author information is available at the end of the article Page 2 of 14Pan et al. BMC Microbiology (2024) 24:281 within a limited range and provide a more accurate, objective, and individualized diagnostic and therapeutic evalua- tion method for patients with adenomyosis, compared to what is currently available.

Adenomyosis, Vaginal microbiome, Menstrual cycles

Adenomyosis is a benign uterine myometrial lesion commonly found in women of reproductive age and is characterized by compensatory hypertrophy in the peripheral myometrium, with endometrioid glands and stroma found in the myometrium [1]. Pathological diag - nosis after surgery is the gold standard for clinical diag - nosis; however, the exact incidence and pathogenesis of adenomyosis remain unknown [2]. Studies have shown that a history of uterine surgery is a high risk factor for adenomyosis. For example, the incidence of adenomyo - sis in patients with the aforementioned surgical history is 1.5 times higher than in patients with a different history [3, 4]. In the treatment of adenomyopathy, in addition to surgical treatment, conservative programs are used to regulate endocrine and immune system functions. Diagnostic methods include magnetic resonance imag - ing (MRI), transvaginal ultrasonography, and CA125 test, however, no specific, individualized diagnostic method is available. Adenomyosis and other benign gynaecologi - cal diseases, such as uterine fibroids, endometriosis, and endometrial polyps, have a high comorbidity rate, and attributing specific symptoms to adenomyosis in clinical diagnosis and treatment is difficult. The vagina is an important organ of the female lower genital tract and is an important habitat for microor - ganisms in the human body. Lactobacillus is the pre - dominant bacterial species and is affected by various exogenous and endogenous factors; furthermore, the species composition of the vaginal microbiome has a strong dynamic change [5]. The vaginal microbiome is an important defence mechanism that regulates and main - tains reproductive function and relative homeostasis in healthy environments. The stability of the microbiome can prevent the proliferation of symbiotic microorgan - isms and the colonization of pathogens [6]. Microorgan - isms affect the balance of the microenvironment through nutritional competition, intraspecific and interspecific signal transduction, metabolic pathways, and product interactions. The mechanism of microenvironmental imbalance remains unclear; however, this imbalance can disrupt normal homeostasis, resulting in certain patho - logical signs. The female upper reproductive tract was once considered a sterile environment; however, this theory has been challenged. The presence of micro - biota in the endometrial microbiota [7] was confirmed by the isolation of microbiota from female endometrial aspirated fluid samples. Studies have shown that bacte - rial DNA can be detected in 95% of post-hysterectomy samples [8]. Microbial switching occurs in the female reproductive tract, and the microbiota of the upper and lower reproductive tracts work synergistically to regulate the uterine environment. With increasing age, synchro - nous changes in the microbiome of the uterus and vagina increasingly converge, showing a mutually parallel rela - tionship. Animal studies have verified the damaging and protective effects of vaginal bacteria on the endometrium using microbiota transplantation techniques [9]. This also indicates that lower reproductive tract bacteria affect or directly interfere with the regulation of some benign and malignant diseases, to some extent, through certain mechanisms. Initial research on vaginal microbes mainly relied on microscopy and microbial culture techniques; however, the vast majority of microorganisms in the physiologi - cal or natural environment are difficult to obtain through culture. Using bioinformatics, high-throughput sequenc - ing and analysis technology were performed to minimise the dependence on bacterial culture technology used in the literature and enhance our understanding of the structure and function of the microbial community, as well as of the association between the bacterial commu - nity of this "non-visual organ" and benign and malignant diseases of the female reproductive system. The 16S-rRNA is a subunit of ribosomal RNA. With improvements in sequencing technology, 16S-rDNA amplicon sequencing has become an important method to evaluate the microenvironment, structure, and com - position [10–13]. As research progresses, sequenc - ing platforms are updated and iterated. Relying on the upgraded Illumina NovaSeq sequencing platform, we compensated for the inefficiency of single-ended read - ing and realized double-ended sequencing; that is, small fragment libraries were built according to the character - istics of the amplified regions. According to our review of the literature, no study has investigated the differences in the vaginal microbiome between adenomyosis patients with different menstrual cycles and healthy individuals. Therefore, this study aimed to elucidate the differences in the vaginal micro - biota between women with and without adenomyosis, with different menstrual cycles. Our results provide a ref- erence for the subsequent screening of characteristic bio- logical markers, disease diagnosis, non-invasive precision Page 3 of 14 Pan et al. BMC Microbiology (2024) 24:281 treatment, and efficacy prediction based on microbial detection.

The case group in this study comprised patients with aden- omyosis in the gynecological outpatient department of Affiliated Hospital of Shandong University from Novem - ber 2021 to October 2022 were selected as the case group. They were evaluated by professional gynecologists, and adenomyosis was confirmed by ultrasound or magnetic resonance imaging (MRI). The control group comprised healthy individuals. The inclusion criteria were as follows: (1) 18–49 years old; (2) no unhealthy lifestyle; (3) Regular menstrual cycle; (4) non-pregnant, non-puerperal, non- lactation, not during the menstrual phase of the estrogen cycle; (6) pre-menopause. The exclusion criteria were as follows: (1) no medical history could be provided; (2) cer- vical intraepithelial lesions, cervical malignancies, vulva lesions and other HPV-related diseases; (3) virus or bacte- rial infection; (4) history and treatment of endocrine sys - tem diseases; (5) autoimmune diseases; (6) acute/chronic inflammation of the urogenital tract; (7) sexually transmit- ted diseases and infectious diseases; (8) malignant tumors; (9) history of sexual life, vaginal bleeding, vaginal douch - ing, vaginal medication, sitting bath, pelvic bath, trans - vaginal examination 48 h before sampling; (10) history of use of antibiotics, antifungals, and hormonal treatments within 30  days before sampling; (11) intrauterine device implantation; (12) recent history of pelvic and abdominal surgery and intrauterine operation. Sample collection The individuals who fulfilled the inclusion criteria had a clinical sample collected on the day of the clinical visit before they received a transvaginal gynecologic exami - nation or gynecologic ultrasound. The posterior vaginal fornix was fully sampled using disposable sterile swabs. During the procedure, contact between the swab head and the speculum, vaginal wall, and other non-sampling sites was avoided. The swab head was cut off with sterile scissors and placed in a sterile centrifuge tube containing Amies culture medium (JINAN BABIO BIOTECHNOL - OGY CO,.LTD.), and stored at -80 ℃ in the laboratory. Extraction of genome DNA The genomic DNA of the sample is extracted by cetyltri - methylammonium bromide (CTAB) method. DNA con - centration and purity was monitored on 1% agarose gels. According to the concentration, DNA was diluted to 1 ng/ µL using sterile water. Using the diluted genomic DNA as a template, the V3-V4 region of 16S-rDNA gene was ampli- fied. The primer sequence was as follows: ① F:CCT AYG GGRBGCASCAG; ②R:GGA CTA CNNGGG TAT CTAAT (Phusion® High-Fidelity PCR Master Mix with GC Buffer, New England Biolabs,lnc.). Polymerase Chain Reaction (PCR) was performed using specific primers with Bar - code and high-efficiency high-fidelity enzyme according to the selection of sequencing region to ensure amplifica- tion efficiency and accuracy. All PCR reactions were car - ried out with 15µL of Phusion® High-Fidelity PCR Master Mix (New England Biolabs); 2 µM of forward and reverse primers, and about 10 ng template DNA. Thermal cycling consisted of initial denaturation at 98℃ for 1 min, followed by 30 cycles of denaturation at 98℃ for 10 s, annealing at 50℃ for 30 s, and elongation at 72℃ for 30 s. Finally 72℃ for 5 min. Library construction and sequencing Sequencing libraries were generated using TruSeq ® DNA PCR-Free Sample Preparation Kit (Illumina, USA) following manufacturer’s recommendations and index codes were added. The library quality was assessed on the [email protected] Fluorometer (Thermo Scientific) and Agilent Bioanalyzer 2100 system. At last, the library was sequenced on an Illumina NovaSeq platform and 250 bp paired-end reads were generated. Paired‑end reads assembly and quality control Paired-end reads was assigned to samples based on their unique barcode and truncated by cutting off the barcode and primer sequence. Paired-end reads were merged using FLASH (V1.2.7, http:// ccb. jhu. edu/ softw are/ FLASH/) [14], which was designed to merge paired-end reads when at least some of the reads overlap the read generated from the opposite end of the same DNA fragment, and the splic- ing sequences were called raw tags. Quality filtering on the raw tags were performed under specific filtering condi - tions to obtain the high-quality clean tags [15] according to the QIIME(V1.9.1, http:// qiime. org/ scrip ts/ split_ libra ries_ fastq. html) [16] quality controlled process. The tags were compared with the reference database (Silva data - base, https:// www. arb- silva. de/) [17] to detect chimera sequences, and then the chimera sequences were removed [18]. Then the Effective Tags finally obtained.

The study enrolled 43 patients with adenomyosis and 40 healthy people. There were no significant differences in demographic background between the two groups of participants (Table 1). The vaginal samples were collected from all partici - pants; however, 7 samples in total were excluded from the control group due to poor DNA quality after library quality check. Therefore, 83 samples were used in the subsequent analysis. (Fig. 1). Page 4 of 14Pan et al. BMC Microbiology (2024) 24:281 Next, the vaginal microbiota was analyzed using 16  s rDNA sequencing techniques. The Raw PE data sequenced by Illumina Novaseq were splicing and quality control to obtain Clean Tags, and then chimeric filtering was performed to obtain Effective Tags for subsequent analysis (S1 Table). Species relative abundances At phylum level, the relative abundance of Firmicutes in adenomyosis group was higher than that in control group (80.70% and 69.72% in adenomyosis and con - trol groups). At the genus level, the Lactobacillus rela - tive abundance in both adenomyosis group and control group was the highest (72.10% and 66.08%). But the rel - ative abundance of Gardnerella and Atopobium in the adenomyosis group was lower than that in the control group (9.67% and 1.04% in adenomyosis and 14.95% and 4.69% in control groups); At the Species level, the Lactobacillus_iners abundance in the adenomyosis group was higher than that in the control group(43.74% and 32.14%), and showed a diversity of Lactobacillus, including Lactobacillus_delbrueckii and Lactobacillus_ jensenii (Fig. 2 ). Different menstrual cycles The top 35 species with the average abundance of all sam- ples of the same level and different groups are selected for clustering, and the heatmap is drawn by heatmap pack - age of R software, which is convenient to find the number or content of species in the sample (Fig. 3). Sample complexity analysis In order to study the influence of menstrual cycle on vag- inal microecology, we named all the samples in the luteal Table 1 Demographic data of the subjects Adenomyosis (N = 40) Control (N = 40) P‑value Age, years (mean ± SD) 39.81 ± 5.62 38.38 ± 5.51 0.243 BMI, kg/m2, (mean ± SD) 23.73 ± 2.81 22.32 ± 3.88 0.060 Gestation 2.19 ± 1.20 1.77 ± 1.07 0.055 Delivery 1.07 ± 0.67 1.00 ± 0.56 0.687 Menstrual cycle, days 26.88 ± 3.67 27.97 ± 2.15 0.066 Menstrual period, days 5.81 ± 1.33 5.47 ± 1.32 0.153 Fig. 1 Study process Page 5 of 14 Pan et al. BMC Microbiology (2024) 24:281 phase of the adenomyosis group as group C and the fol - licular phase as group D. All the samples in the luteal phase of the control group were named group E and group F. in the follicular phase. Alpha-diversity analysis showed significant dif - ferences in the adenomyosis and control group dur - ing luteal phase (Shannon index, p = 0.0087; Simpson index, p = 0.0056), but we didn’t find the statistically difference in ACE and chao 1 index (Fig.  4). It was verified that the amount of sequencing data was pro - gressive and reasonable, and more data would only produce a few new species, thus suggesting a uniform distribution of species (Fig. 5 ). Comparative analysis of multiple copies The species distributions in the adenomyosis group and the control group were not completely separated, but were similar (Fig. 6). We analyzed the Beta-diversity index by using the t-test and found that the species Beta-diversity index was significantly different between the adenomyosis group and the control group (p = 0.018). However, based on Weighted Unifrac analysis, significant differences between the disease group and the control group were only observed throughout the luteal phase (p = 0.0146) (Fig. 7 A, B, C, D). R value was between (-1, 1), and R value was greater than 0, indicating that the difference between groups was greater than the difference within groups, which was sig - nificant (P < 0.05). The reasonableness of the grouping in this study was proved. (Table 2). At the phylum level, there were no significant spe - cies differences between the adenomyosis group and the control group. At the class level the significant differ - ences was in Coriobacteriia and Gammaproteobacteria (p < 0.01). At the class level the significant differences Fig. 2 Taxonomy bar charts of vaginal microbiame at (A) phylum, (B) class, (C) order, (D) family, (D) genus and (E) species level Page 6 of 14Pan et al. BMC Microbiology (2024) 24:281 was in Lactobacillales, Coriobacteriales (p < 0.01),and in Pseudomonadales (p < 0.05). At the class level the signifi - cant differences was in Beijerinckiaceae and Listeriaceae (p < 0.05). At the genus level, that were in Listeria, Ral- stonia, Acinetobacter, and Haemophilus (p < 0.01), and Alloscardovia,Ureaolasma (p < 0.05). Finally, at the spe - cies level,there was significant difference in Alloscardo - via_omnicolens and Lactobacillus_delbrueckii (p < 0.01) (Fig. 8). At the phylum level, Firmicutes showed the highest species abundance in both the adenomyosis group and the control group, and at the same time, contributed the most to the species difference between the two groups (Fig. 9). Random forest is a classical machine learning model based on classification tree algorithm to screen fea - tures (biomarkers) that play an important role in classi - fication or grouping. A default tenfold cross-validation was performed for each model, and Receiver Operating Characteristic Curve (ROC) curves were drawn to select potential Biomaker 50 as shown in Fig. 10.

Species diversity was analyzed using alpha diversity indices (Shannon index, chao1 index, ACE, and Simp - son indices), and the number of microbial species and proportion of each species in a single sample were cal - culated. Results showed that species diversity of the two groups did not show significant differences, similar to the

of Chen et al. [19]. Although the species compo - sition of the two groups was similar, species abundance significantly differed. At the phylum level, the relative abundance of Firmicutes was higher in the adenomyo - sis group than in the control group. At the genus level, except for the absolute species dominance of Lactoba - cillus in both groups, the relative abundance of Gard - nerella in the adenomyosis group was significantly lower than that in the control group, which differed from the

of Kunaseth [20]. Other groups of vaginal bacilli were also detected, second only to Lactobacillus in over - all abundance. Lactobacillus vegetation in the female reproductive tract is critical for the maintenance of genital health. Fig. 3 Heatmap of species abundance clustering during different menstrual cycles. The top 35 species with the average abundance of all samples of the same level and different groups are selected for clustering at (A) phylum, (B) class, (C) order, (D) family, (D) genus and (E) species level.The heatmap is drawn by heatmap package of R software, which is convenient to find the number or content of species in the sample Page 7 of 14 Pan et al. BMC Microbiology (2024) 24:281 However, the exact pathogenesis of Gardnerella vagi - nalis remains unclear [21]. Lactobacillus and Gard - nerella interact in the female reproductive tract; when the abundance of Lactobacillus decreases to a certain extent, the growth of Gardnerella can decrease or stop [22], and the imbalance of the two bacteria can change the acid–base environment of the vagina and produce mucosal adsorption and biofilm, promoting chronic, persistent infection and inflammation [23, 24]. A data analysis using the dominance network analysis frame - work found that Lactobacillus is not the dominant spe - cies in some healthy African women, and very few bacteria have a cooperative and mutually beneficial relationship with Gardnerella and Lactobacillus iners [25], contrary to previous views [26]. L. iners cooperate with Gardnerella but are inhibited by other species [27]. A high abundance of Gardnerella genomospecies indi - cate the presence of gene variants coding for virulence factors, such as cholesterol-dependent pore-forming cytotoxin vaginolysin and neuraminidase sialidase [28]. In this study, the abundance of L. iners in the adeno - myosis group was found to be significantly higher than that in the control group, which was verified using the MetaStat method. Microbiomes from women diagnosed with Amsel-bacterial vaginosis (BV) were enriched for host immune response evasion and colonization func - tions by L. iners, and its role in the vaginal microbiome has been widely debated. A study has identified a specific set of L. iners genes associated with positive Amsel-BV diagnoses, and their data suggested that certain L. Iners strains may adhere to epithelial cells, contributing to the appearance of clue cells and becoming more difficult to displace in the vaginal environment [27]. In conclu - sion, the variation in L. iners and Gardnerella abundance may be a potential cause of adenomyosis, and maintain - ing the balance of Lactobacillus and Gardnerella in the Fig. 4 Alpha-diversity analysis. A shannon index, (B)Simpson index, (C) ACE index, (D) chao1 index. Alpha-diversity analysis indices for different samples at 97% consistency thresholds Page 8 of 14Pan et al. BMC Microbiology (2024) 24:281 body may be a self-mechanism to maintain the stability of vaginal microecology. However, little is known about how the genital micro - biota affects host immune function and regulates disease susceptibility. Lactobacillus imbalance and high ecologi - cal diversity may be closely related to the concentration of pro-inflammatory cytokines in genital organs [29]. Patients with adenomyosis show leukocyte infiltration in the endometrial functional layer, and the number of mac- rophages and natural killer (NK) cells increased [30, 31]. Transcriptional analysis showed that antigen-presenting cells sense gram-negative bacterial products in  situ via Toll-like receptor 4 (TLR-4) signalling, promoting geni - tal organ inflammation by activating the nuclear factor kappa-B (NF-κB) signalling pathway and recruiting lym - phocytes through chemokine production [29]. Immune dysregulation is present in the ectopic endometrium of patients with adenomyopathy and manifests as elevated T Cell Immunoglobulin Domain and Mucin Domain-3/ Galectin-9 (Tim-3/Gal-9) expression and differential RNA methylation [32, 33]. Therefore, we speculated that vaginal microecological changes affect the important role of Tim-3/Gal-9 in immunosuppression through some mechanism, causing the persistence of infection, affect - ing the growth environment of the endometrial tissue, and causing adenomyosis. In addition, the expression of Type I interferon (IFN-I) inducers is increased in the ectopic endometrium in adenomyosis. The increased lev- els of IFN-Is and expression of IFN-stimulating genes and pro-inflammatory cytokines in tissues may be related to host immunity under the influence of certain microor - ganisms [34]. Recent literature has suggested that micro - biota-induced interferon activation does not require direct host-bacterial interaction but the remote trans - port of bacterial DNA into host cells via bacteria-derived membrane vesicles [35]. In contrast with our finding that the beta diversity index was significantly higher in the adenomyosis group than in the control group, the increased bacterial diversity in the vagina prob - ably explains the activation of the host’s innate immune response in the ectopic endometrium in adenomyosis [5, 20]. Endometriosis and adenomyosis are closely related disorders. Their pathophysiology and clinical symptoms such as chronic pain are extremely similar [36]. There is a correlation in the microbial composition of both intes - tinal and cervicovaginal microbial niches, and over 50% overlap in species abundance and cell density [37]. Cen - tral sensitisation is known to be significantly involved in endometriosis-associated chronic pelvic pain [38]. Dysbi- osis may potentially lead to incorrect immune responses, triggering the development of inflammatory pain [39], such as that seen in endometriosis and adenomyosis. All the patients with adenomyosis included in the study have obvious dysmenorrhea. However, further studies are may elucidate the association between microbial changes and chronic pain. The microbiota of the female reproductive system is influenced by changes in age and system physiology, and Fig. 5 Rarefaction curve and Rank Abundance curve. In the (A) Rarefaction curve, horizontal coordinate is the number of sequencing strips randomly selected from a sample, and the vertical coordinate is the number of Operational Taxonomic Units (OTUs) that can be constructed based on the number of sequencing strips, which is used to reflect the sequencing coverage, and different samples are represented by different colored curves; in the (B) Rank Abundance curve, the horizontal coordinate is the serial number sorted by the abundance of OTUs, and the vertical coordinate is the relative abundance of the corresponding OTUs, and different samples are represented by different colored fold lines Page 9 of 14 Pan et al. BMC Microbiology (2024) 24:281 the menstrual cycle is a major disruptor of the vaginal microbiome. Different microbiota characteristics are observed in women at different physiological stages [40]. In healthy women of reproductive age, the vaginal micro- biome composition changes dramatically before and after menstruation [41]. Menstrual blood flowing through the vagina leaves sufficient iron necessary for pathogens, and the iron necessary for pathogen metabolism [42], which is reduced by the iron-binding affinity of lactoferrin, is replenished. Additionally, studies measuring oestradiol levels and vaginal microbiome composition in women who use oral contraceptives to inhibit ovulation have shown that the high diversity observed during menstrua - tion is mainly driven by oestradiol withdrawal before menstruation rather than by the dynamic drive of pro - gesterone. Lactobacillus abundance increases during the follicular and luteal phases, gradually normalising the vaginal microecology [41, 43]. Under the influence of Fig. 6 AWeighted Unifrac based distance from Principal Co-ordinates Analysis (PCoA) analysis. Horizontal coordinates indicate one principal component, vertical coordinates indicate another principal component, and percentages indicate the contribution of the principal component to the sample variance; each point in the graph indicates a sample, and samples from the same group are indicated using the same color (B) Unweighted Unifrac based distance from PCoA analysis. C Euclidean based distances from Principal Component Analysis (PCA) analysis. The horizontal coordinate indicates the first principal component, and the percentage indicates the contribution value of the first principal component to the sample difference; the vertical coordinate indicates the second principal component, and the percentage indicates the contribution value of the second principal component to the sample difference; each point in the graph indicates a sample, and samples in the same group are indicated using the same color; in PCA graphs with clustering circles, the clustering circle is added with the grouping information (clustering circles need more than 3 samples in the group) Page 10 of 14Pan et al. BMC Microbiology (2024) 24:281 this periodicity, combined with our test results, differ - ent types of dominant bacterial profiles were observed in patients with adenomyosis in both luteal and follicular stages, which provided a reference for the detection of biomarkers in patients with specific menstrual cycles or to evaluate their efficacy. In summary, in this study, an increase in microbial richness was associated with adenomyosis, and the microbiome characteristics of patients with and with - out adenomyosis differed according to the menstrual cycle. This study has three notable limitations: 1) the final sample size was limited because of coronavirus disease 2019 (COVID-19), 2) large sample of clini - cal data for verification was not available, and 3) the different methods used in each study may have led to different conclusions. Furthermore, adenomyosis diagnosis remains unconfirmed without histological Fig. 7 A Weighted Unifrac based distance from Beta-diversity analysis. B Unweighted Unifrac based distance from Beta-diversity analysis. The box plots of Beta-diversity between-group difference analysis can visualize the median, dispersion, maximum, minimum, and outliers of within-group sample similarity. At the same time, the T-test test was used to analyze whether the Beta diversity differences of species between groups were significant or not. C Weighted unifrac ased distance from Beta-diversity analysis during different menstrual cycles. D Unweighted unifrac ased distance from Beta-diversity analysis during different menstrual cycles Table 2 Anosim analysis based on the Bray–Curtis distance. Anosim analysis is a non-parametric test used to test whether the difference between groups is significantly greater than the difference within groups, so as to determine whether the grouping is meaningful. We conducted the significance test of the difference between groups based on the rank of the Bray– Curtis distance value Group R value P value B-A 0.03067 0.044 Page 11 of 14 Pan et al. BMC Microbiology (2024) 24:281 Fig. 8 MetaStat analysis at (A) phylum, (B) class, (C) order, (D) family, (D) genus and (E) species level. For the species with significant differences between study groups, MetaStat method was used to screen the species with significant differences based on the species abundance tables of different levels Fig. 9 Simper analysis. It is a breakdown of the Bray–Curtis difference index that quantifies how much each species contributes to the difference between two groups. The results show the top 10 species with the highest contribution to the difference between the two groups and their abundance Page 12 of 14Pan et al. BMC Microbiology (2024) 24:281 Fig. 10 A MeanDecreaseAccuracy based analysis and MeanDecreaseGin based analysis. B proportion of false positive (Specificity), ordinate: proportion of true Sensitivity; (C) ROC curve of the test pair, abscess: proportion of false positive (Specificity), ordinate: proportion of true Sensitivity (specificity) Mean Decrease Accuracy measures the extent to which the prediction accuracy of random forest is reduced when the value of a variable is changed to a random number. The greater the value, the greater the importance of the variable. MeanDecreaseGini compared the importance of the variables by calculating the effect of each variable on the heterogeneity of the observed values at each node of the classification tree using the Gini index Page 13 of 14 Pan et al. BMC Microbiology (2024) 24:281 assessment. This may have led to misclassification in both cases (false positives) and controls (false nega - tives). In future research, we plan to develop stand - ardized analysis software and large databases to continue our investigation of the mechanisms behind this association. Abbreviations CA125 Carbohydrate Antigen 125 PCR Polymerase Chain Reaction CTAB Cetyltrimethylammonium Bromide. OTUs Operational Taxonomic Units PCoA Principal Co-ordinates Analysis MRI Magnetic resonance imaging BV Bacterial Vaginosis NK cells Natural killer TLR-4 Toll-like receptor 4 NF-κB Nuclear Factor kappa-B Tim-3 T Cell Immunoglobulin Domain and Mucin Domain-3 GAL-9 Galectin-9 IFN-I Type I interferon COVID-19 Coronavirus disease 2019 Supplementary Information The online version contains supplementary material available at https:// doi. org/ 10. 1186/ s12866- 024- 03339-9. Supplementary Material 1.

We would like to thank all of the volunteers who participated in this study. Authors’ contributions All authors read and approved the final manuscript. GYW and MY designed the experiments. ZYP conceived and carried out sample collection, analysis, and interpretation. JD, PZ, QHR, XYW, HS, SMY and XJ carried out sample col- lection. ZYP drafted the manuscript and prepared the figures and tables. All authors finalized the final manuscript. Funding This study was supported by the National Key R&D Program of China [2023YFC2705400], the Major Basic Research of Natural Science Foundation of Shandong [ZR2021ZD34], and the National Natural Science Foundation of China [grant numbers 82071621 and 81901458]. Availability of data and materials The datasets generated and analyzed during the current study are available in the [China National Center of Bioinformation (CNCB)]database. The number of this project is CRA012802. [https:// ngdc. cncb. ac. cn/ gsub/ submit/ gsa/ list]. Declarations Ethics approval and consent to participate This study was reviewed and approved by the Medical Ethics Committee of Qilu Hospital of Shandong University (ethics approval No.: KYLL-20211–092-1). The research has been performed in accordance with the Declaration of Hel- sinki. The patients were informed about the sample collection and had signed informed consent forms. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Author details 1 Department of Obstetrics and Gynecology, Shandong Provincial Hospital, Jinan 250000, Shandong, China. 2 Medical Integration and Practice Center, Cheeloo College of Medicine, Shandong University, Jinan 250000, Shandong, China. 3 JiNan Key Laboratory of Diagnosis and Treatment of Major Gynae- cological Disease, Jinan 250000, Shandong Province, China. 4 Gynecology Laboratory, Shandong Provincial Hospital, Jinan 250000, Shandong Province, China. 5 Gynecology Laboratory, Medical Science and Technology Innovation Center, Shandong First Medical University & Shandong Academy of Medical Sciences, Jinan 250000, Shandong Province, China. 6 Qilu Hospital of Shandong University, Jinan 250000, Shandong Province, China. Received: 1 September 2023 Accepted: 16 May 2024

1. Yen CF, Huang SJ, Lee CL, et al. Molecular characteristics of the endome- trium in uterine adenomyosis and its biochemical microenvironment[J]. Reprod Sci. 2017;24(10):1346–61. 2. Cunningham RK, Horrow MM, Smith RJ, et al. Adenomyosis: a sono- graphic diagnosis [J]. Radiographics. 2018;38(5):1576–89. 3. Parazzini F, Vercellini P , Panazza S, et al. Risk factors for adenomyosis[J]. Hum Reprod. 1997;12(6):1275–9. 4. Riggs JC, Lim EK, Liang D, et al. Cesarean section as a risk factor for the development of adenomyosis uteri[J]. J Reprod Med. 2014;59(1–2):20–4. 5. Abou Chacra L, Fenollar F, Diop K. Bacterial Vaginosis: what do we cur- rently know? [J]. Front Cell Infect Microbiol. 2022;11: 672429. 6. Deidda, F, Amoruso, A, Allesina, S, et al. In Vitro Activity of Lactobacil- lus fermentum LF5 Against Different Candida Species and Gardnerella vaginalis: A New Perspective to Approach Mixed Vaginal Infections?[J]. J CLIN GASTROENTEROL. 2016;50 (Suppl 2):S168-70. 7. Moreno I, Codoñer FM, Vilella F, et al. Evidence that the endometrial microbiota has an effect on implantation success or failure[J]. Am J Obstet Gynecol. 2016;215(6):684–703. 8. Mitchell CM, Haick A, Nkwopara E, et al. Colonization of the upper genital tract by vaginal bacterial species in nonpregnant women[J]. Am J Obstet Gynecol. 2015;212(611):1–611.e9. 9. Wang J, Li Z, Ma X, et al. Translocation of vaginal microbiota is involved in impairment and protection of uterine health[J]. Nat Commun. 2021;12(1):4191. 10. Youssef N, Sheik CS, Krumholz LR, et al. Comparison of species richness estimates obtained using nearly complete fragments and simulated pyrosequencing-generated fragments in 16S rRNA gene-based envi- ronmental surveys[J]. Appl Environ Microbiol. 2009;75(16):5227–36. 11. Hess M, Sczyrba A, Egan R, et al. Metagenomic discovery of bio - mass-degrading genes and genomes from cow rumen[J]. Science. 2011;331(6016):463–7. 12. Lundberg DS, Yourstone S, Mieczkowski P , et al. Practical innova- tions for high-throughput amplicon sequencing[J]. Nat Methods. 2013;10(10):999–1002. 13. Hickey RJ, Zhou X, Settles ML, et al. Vaginal microbiota of adolescent girls prior to the onset of menarche resemble those of reproductive- age women [J]. mBio. 2015;6(2):e00097–15. 14. Magoč T, Salzberg SL. FLASH: fast length adjustment of short reads to improve genome assemblies[J]. Bioinformatics. 2011;27(21):2957–63. 15. Bokulich NA, Subramanian S, Faith JJ, et al. Quality-filtering vastly improves diversity estimates from Illumina amplicon sequencing[J]. Nat Methods. 2013;10(1):57–9. 16. Caporaso JG, Kuczynski J, Stombaugh J, et al. QIIME allows analysis of high-throughput community sequencing data[J]. Nat Methods. 2010;7(5):335–6. 17. Edgar RC, Haas BJ, Clemente JC, et al. UCHIME improves sensitivity and speed of chimera detection[J]. Bioinformatics. 2011;27(16):2194–200. 18. Haas BJ, Gevers D, Earl AM, et al. Chimeric 16S rRNA sequence formation and detection in Sanger and 454-pyrosequenced PCR amplicons[J]. Genome Res. 2011;21(3):494–504. Page 14 of 14Pan et al. BMC Microbiology (2024) 24:281 19. Chen S, Gu Z, Zhang W, et al. Microbiome of the lower genital tract in Chinese women with endometriosis by 16s-rRNA sequencing tech- nique: a pilot study[J]. Ann Transl Med. 2020;8(21):1440. 20. Kunaseth J, Waiyaput W, Chanchaem P , et al. Vaginal microbiome of women with adenomyosis: a case-control study [J]. Plos One. 2022;17(2):e0263283. 21. Djukic M, Schmidt-Samoa C, Lange P , et al. Cerebrospinal fluid findings in adults with acute Lyme neuroborreliosis[J]. J Neurol. 2012;259(4):630–6. 22. Haedicke J, Iftner T. Human papillomaviruses and cancer[J]. Radiother Oncol. 2013;108(3):397–402. 23. Machado D, Castro J, Palmeira-de-Oliveira A, et al. Bacterial vaginosis biofilms: challenges to current therapies and emerging solutions[J]. Front Microbiol. 2016;6:1528. 24. Garcia EM, Kraskauskiene V, Koblinski JE, et al. Interaction of Gard- nerella vaginalis and Vaginolysin with the Apical versus Basolateral face of a three-dimensional model of vaginal epithelium. Infect Immun. 2019;87(4):e00646–e718. 25. Doyle R, Gondwe A, Fan YM, et al. A lactobacillus-deficient vaginal micro- biota dominates postpartum women in Rural Malawi[J]. Appl Environ Microbiol. 2018;84(6):e02150–e2217. 26. Ravel J, Gajer P , Abdo Z, et al. Vaginal microbiome of reproductive-age women[J]. Proc Natl Acad Sci U S A. 2011;108(Suppl 1):4680–7. 27. Li W, Ma ZS. Dominance network analysis of the healthy human vaginal microbiome not dominated by Lactobacillus species[J]. Comput Struct Biotechnol J. 2020;18:3447–56. 28. Holm JB, France MT, Gajer P , et al. Integrating compositional and func- tional content to describe vaginal microbiomes in health and disease[J]. Microbiome. 2023;11(1):259. 29. Anahtar MN, Byrne EH, Doherty KE, et al. Cervicovaginal bacteria are a major modulator of host inflammatory responses in the female genital tract[J]. Immunity. 2015;42(5):965–76. 30. Orazov MR, Radzinsky VE, Nosenko EN, et al. Immune-inflammatory predictors of the pelvic pain syndrome associated with adenomyosis[J]. Gynecol Endocrinol. 2017;33(sup1):44–6. 31. Tremellen KP , Russell P . The distribution of immune cells and mac- rophages in the endometrium of women with recurrent reproduc- tive failure. II: adenomyosis and macrophages[J]. J Reprod Immunol. 2012;93(1):58–63. 32. Huang P , Lv C, Zhang C, et al. Expression and significance of T-cell immu- noglobulin mucin molecule 3 and its ligand galectin-9 in patients with adenomyosis[J]. Gynecol Endocrinol. 2020;36(7):605–10. 33. Zhai J, Li S, Sen S, et al. m6A RNA methylation regulators contribute to eutopic endometrium and myometrium dysfunction in adenomyosis[J]. Front Genet. 2020;11:716. 34. Qu H, Li L, Wang TL, et al. Epithelial cells in endometriosis and adenomyo- sis upregulate STING expression[J]. Reprod Sci. 2020;27(6):1276–84. 35. Erttmann SF, Swacha P , Aung KM, et al. The gut microbiota prime systemic antiviral immunity via the cGAS-STING-IFN-I axis[J]. Immunity. 2022;55(5):847–61. 36. Bulun SE, Yildiz S, Adli M, et al. Endometriosis and adenomyosis: shared pathophysiology[J]. Fertil Steril. 2023;119(5):746–50. 37. Amabebe E, Anumba DOC. Female Gut and genital tract microbiota- induced crosstalk and differential effects of short-chain fatty acids on immune Sequelae[J]. Front Immunol. 2020;11:2184. 38. Bajaj P , Bajaj P , Madsen H, et al. Endometriosis is associated with central sensitization: a psychophysical controlled study[J]. J Pain. 2003;4(7):372–80. 39. Salmeri N, Sinagra E, Dolci C, et al. Microbiota in irritable bowel syndrome and endometriosis: birds of a feather flock together-a review[J]. Microor- ganisms. 2023;11(8):2089. 40. Galkin F, Mamoshina P , Aliper A, et al. Human gut microbiome aging clock based on taxonomic profiling and deep learning [J]. IScience. 2020;23(6):101199. 41. Krog MC, Hugerth LW, Fransson E, et al. The healthy female microbiome across body sites: effect of hormonal contraceptives and the menstrual cycle[J]. Hum Reprod. 2022;37(7):1525–43. 42. Roberts SA, Brabin L, Diallo S, et al. Mucosal lactoferrin response to genital tract infections is associated with iron and nutritional biomarkers in young Burkinabé women[J]. Eur J Clin Nutr. 2019;73(11):1464–72. 43. Song SD, Acharya KD, Zhu JE, et al. Daily vaginal microbiota fluctua- tions associated with natural hormonal cycle, contraceptives, diet, and exercise[J]. MSphere. 2020;5:1–14. Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in pub- lished maps and institutional affiliations.