Multi-omics profiling identifies inosine as a key metabolite associated with embryonic arrest in hydrosalpinx-associated infertility.

Source data are provided with this paper. T&T-seq generated in this study were deposited to GSA ( https://ngdc.cncb.ac.cn/gsa/ ), with the accession number: GSA: CRA030324.

The clinical samples collected in this study were approved by the Peking University Third Hospital Medical Science Research Ethics Committee (IRB00006761-M2023838). Patients were recruited and samples were collected between March 2023 and December 2024. As a pilot study, fallopian tubes were obtained from 52 premenopausal control patients (aged 33–52 years old) undergoing hysterectomy for benign conditions, including uterine fibroids (n = 33), adenomyosis (n = 13), endometriosis (n = 5), and uterine prolapse (n = 1). Additionally, 49 patients with hydrosalpinx (aged 25–48 years old) undergoing salpingectomy were enrolled. All patients signed written informed consent before surgery. Luminal fluids were collected immediately after the fallopian tubes were removed by flushing the lumen of the fallopian tubes with 10 mL of sterile normal saline solution, as previously described. 23 Notably, fluids from each side were collected and processed separately without pooling. After verifying the pathological diagnosis, fluid from a randomly selected side from each patient was aliquoted into 1.5 mL microcentrifuge tubes and stored at −80 °C for subsequent analysis. All patients had no prior history of metabolic diseases, such as diabetes, polycystic ovarian syndrome (PCOS), hypothyroidism, or metabolic dysfunction-associated fatty liver disease (MAFLD), or any other malignant tumours. Demographic and clinical information including age, body mass index (BMI), ethnicity and the pathological report of the fallopian tube were recorded ( Table S1 ). This study was conducted in accordance with the STROBE guidelines. After thawing on ice at 4 °C for 30 min and vortexing for 10 s, 100 μL of fallopian tube luminal fluid was mixed with 300 μL extract solution (ACN: Methanol = 1:4, V/V) containing an internal standard. After vortexing for 3 min, the mixture was centrifuged at 13,400 g for 10 min at 4 °C for protein precipitation. A 150 μL aliquots of the supernatant was collected and placed at −20 °C for 30 min, and then centrifuged at 13,400 g for 3 min (4 °C). A total of 120 μL of supernatant was transferred for LC-MS analysis. All samples were analysed using two LC/MS methods. For one aliquot, positive-ion conditions were used for analysis, eluted from a T3 column (Waters ACQUITY Premier HSS T3 Column 1.8 μm, 2.1 mm ∗ 100 mm) with 0.1% formic acid in water as solvent A and 0.1% formic acid in acetonitrile as solvent B using the following gradient: 5–20% B over 2 min, increased to 60% B over the next 3 min, increased to 99% B over 1 min and held for 1.5 min, then returned to 5% mobile phase B within 0.1 min, followed by holding for 2.4 min. Analytical conditions were as follows: column temperature, 40 °C; flow rate, 0.4 mL/min; injection volume, 4 μL. Another aliquot was analysed using negative ion conditions with the same elution gradient as in the positive mode. Data acquisition was operated in the data-dependent acquisition (DDA) mode using Analyst TF 1.7.1 Software (Sciex, Concord, ON, Canada). Source parameters were as follows: ion source gas 1 (GAS1), 50 psi; ion source gas 2 (GAS2), 50 psi; curtain gas (CUR), 25 psi; temperature (TEM), 550 °C; declustering potential (DP), 60 V or −60 V in positive or negative modes respectively; ion spray voltage floating (ISVF), 5000 V or −4000 V in positive or negative modes, respectively. TOF MS scan parameters were as follows: mass range, 50–1000 Da; accumulation time, 200 ms; dynamic background subtraction, enabled. Product ion scan parameters were as follows: mass range, 25–1000 Da; accumulation time, 40 ms; collision energy, 30 or −30 V in positive or negative modes, respectively; collision energy spread, 15; resolution, UNIT; charge state, 1 to 1; intensity, 100 cps; exclude isotopes within 4 Da; mass tolerance, 50 ppm; maximum number of candidate ions to monitor per cycle, 18. The levels of inosine, adenosine, hypoxanthine, 8-OHG, PGE 2 , and IL-6 were determined using Human ELISA kits following manufacturer's instructions (Jiangsu Meimian industrial Co., Ltd, MM-927291O1; MM-926439O1; MM-926421O1; MM-0331H2; MM-0162H1; MM-0049H1). The fallopian tube fluid samples were centrifuged at 4 °C for about 20 min (2000 rpm) and the supernatant was carefully collected. A total of 10 μL of diluted sample (1:5) was added to each pre-coated well along with 40 μL of sample diluent. After adding 50 μL of enzyme reagent, the plate was incubated at 37 °C for 30 min, then washed five times. Colour reagents A and B (50 μL each) were added, followed by incubation at 37 °C for 10 min. The reaction was stopped with 50 μL of stop solution, and absorbance was measured at 450 nm within 15 min. Concentrations were calculated using a standard curve. The C57BL/6J mice were purchased and housed in the Laboratory Animal Science Department at Peking University Health Science Centre (Beijing, China). Mice were kept in a temperature-regulated environment with a 12-h light/dark cycle, maintained at 18–23 °C and 40–60% humidity, with food and water ad libitum. All animal procedures were performed in compliance with the ARRIVE guidelines and were approved by the Peking University Third Clinical Medical School Ethical Committee of Animals (Approval No. A2023127). 6–8-week-old female mice were intraperitoneally injected with 7.5 IU pregnant mare's serum gonadotropin (PMSG, Ningbo Second Hormone Factory) followed by 7.5 IU human chorionic gonadotrophin (hCG, Ningbo Second Hormone Factory) 46 h later. Female mice were placed in the same cage with adult male mice overnight for mating after hCG injection. The next morning, female mice with vaginal plugs were used for zygote collection. Zygotes were extracted from the oviducts, and cumulus cells were removed using a 0.4% hyaluronidase (Sigma, H3506) solution. To minimise potential maternal bias, zygotes collected from different mice were pooled together and then randomly and equally allocated to each treatment group. At least three independent experiments were conducted for each condition. Embryos at different stages were observed or collected: Zygotes (22–24 h), 2-cell embryos (36–40 h), late 2-cell embryos (48–50 h), 4-cell embryos (52–54 h), late 4-cell embryos (58–60 h), 8-cell embryos (66–70 h), morula (78–82 h) and blastocysts (94–98 h) after hCG injection were cultured in mini-drops of KSOM medium (AIVFO, C003-3) covered with mineral oil (Sigma–Aldrich, M8410) at 37 °C, 5% CO 2 and 100% humidity. For compound treatment, 1.5‰ DMSO, inosine and forodesine at different concentrations were supplemented in KSOM medium. At least 30 embryos were collected in each group. Following culture in KSOM medium until the E4.0 stage, embryos were surgically transferred into the uterine horns of pseudopregnant female mice at E2.5. The development of deciduae was assessed five days post-transfer at E7.5. Successful decidualisation was identified by the presence of uterine swellings that were significantly wider than the non-implanted uterine segments. A 20 μM solution of pooled siRNAs was injected into the cytoplasm of zygotes utilising Eppendorf FemtoJet 4i micromanipulators. Following injection, the zygotes were maintained in culture at 37 °C with a 5% CO 2 atmosphere until they reached the blastocyst stage. Detailed sequences and specifications for the siRNAs used in this study are listed in Table S2 . The procedure of ultrasensitive translatomics and total RNA-seq was conducted as a previous study described. 24 Each sample contained five mouse embryos. Embryos were carefully picked by mouth pipetting after removing zona pellucida and transferred into a 200 μL PCR tube containing 10 μL of lysis buffer, which included an 8-nucleotide barcode, RNase inhibitor (EO0381, Thermo), 10% Triton X-100 and dNTPs. The samples were thoroughly vortexed and divided into two parts, of which 2 μL of lysate was used for the transcriptome and the remaining for translatome. The Ribolace beads (Immagina, RL001) were prepared as manufacturer's instructions. The lysate for the translatome was mixed with binding buffer (Tris–HCl, NaCl, MgCl 2 , DTT and cycloheximide) and functional Ribolace beads. Then the mixture was incubated at 4 °C, 10 rpm, for 70 min on a rotator. Then the samples were washed, and ribosome-bound mRNAs were released using 1% SDS and 0.1 mg of proteinase K. After purification with RNAClean XP beads (A63987, Beckman Coulter), the ribosome-bound full-length RNA was recovered with RNase inhibitor, 10% Triton X-100, an 8-nucleotide barcode and dNTPs. Both total RNA (2 μL lysate) and ribosome-bound full-length RNA (5 μL) were incubated at 72 °C for 3 min, then underwent reverse transcription and pre-amplification using the SMART-seq2 method, in accordance with the published protocol. 25 Then, the cDNA products were fragmented into 300bp fragments using a Covaris S3 ultrasonicator. After purification, 30 ng of cDNA fragments was used to construct libraries with the NEBNext Ultra II RNA Library Prep Kit (E7770, New England Biolabs). The sequencing was performed on the Illumina NovaSeq X Plus platform. Embryos were collected, lysed and subjected to reverse transcription in 4 μL of lysis buffer as described above. qPCR was carried out using SYBR Master Mix (Applied Biosystems, A25742) and QuantStudio 3 Real-Time PCR System. The qPCR conditions were 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s, and 60 °C for 1 min. Relative gene expression levels were quantified using the 2 −△△CT method and normalised to Gapdh expression, based on data from at least three independent experiments. The primers used in this study are given in Table S2 . For immunofluorescence staining, the sections underwent heat-mediated antigen retrieval with sodium citrate buffer (pH of 6.0) for 16 min, then blocked in 10% normal donkey serum (YEASON) at room temperature for 1 h. Primary antibodies against goat anti-Purine Nucleoside Phosphorylase (PNP) (1:100, Thermo scientific, PA5-37877, RRID: AB_2554485 ), mouse anti-acetylated Tubulin (1:200, Sigma, T7451, RRID: AB_609894 ), rabbit anti-γH2AX (1:200, Cell Signalling Technology, 9718, RRID: AB_2118009 ) and rabbit anti-Cleaved Caspase-3 (Asp175) (1:200, Cell Signalling Technology, 9661, RRID: AB_2341188 ) were diluted in PBS and incubated with sections in a humidified chamber for 16 h at 4 °C. After 3 washes in PBS, the sections were incubated with the corresponding secondary antibody conjugated to Alexa Fluor 594 or 488 (Invitrogen; 1:200) and 4’,6-diamidino-2-phenylindole (DAPI) (1:200, ab285390, Abcam) for 1 h at room temperature, followed by three washes with PBS. Images were captured with a ZEISS LSM880 inverted confocal microscope. For immunohistochemistry staining, the sections were treated with hydrogen peroxide after immunofluorescence staining to inhibit endogenous peroxidase activity. After incubation with HRP-conjugated secondary antibodies (ZSGB-BIO), the sections were visualised with DAB and counterstained with haematoxylin to label nuclei. Images were obtained through WISLEAP WS-10. Embryos were fixed in 4% paraformaldehyde in phosphate buffered saline (PBS) at room temperature for 20 min. After washing three times in wash buffer, the embryos were incubated in PBS with 0.5% Triton X-100 at room temperature for 30 min to permeabilise. After being blocked with 1% bovine serum albumin in PBS, the embryos were incubated with primary antibodies overnight at 4 °C and sequentially labelled with Alexa Fluor 594-conjugated or 488-conjugated secondary antibodies and DAPI for 30 min. Primary antibodies used were rabbit anti-OCT4 (abcam, ab181557; 1:50, RRID: AB_2687916 ), mouse anti-CDX2 (Biogenex, AM392; 1:200, RRID: AB_2650531 ), rabbit anti-checkpoint kinase 1 (CHEK1) (abcam, ab32531, 1:200, RRID: AB_726821 ), rabbit anti-CCNT2 (Immunoway, YT6222; 1:200, RRID: AB_3740855 ), and rabbit anti-BTG4 (abcam, EPRZJU-21, RRID: AB_2861140 ). Embryos were imaged using a confocal microscope. Images were captured with ZEISS LSM880 and LSM980 inverted confocal microscopes. After being fixed with 4% paraformaldehyde, dehydrated with ethanol, clarified with xylene, the FT sections were embedded in paraffin and then serially sliced into 5 μm sections and stained with H&E. Images were obtained through WISLEAP WS-10. Fresh FT tissue was washed with PBS to remove blood, then immediately fixed in electron microscopy fixative for 2 h at room temperature. The tissue blocks were subsequently transferred to 1% OsO 4 in 0.1 M PBS (pH 7.4) for 1–2 h at room temperature. Dehydration was performed sequentially with 15-min incubations in 30%, 50%, 70%, 80%, 90%, and 95% ethanol, followed by two changes in 100% ethanol and a 15-min treatment with isoamyl acetate. The samples were then dried using a Critical Point Dryer. Specimens were mounted on metallic stubs with carbon stickers and sputter-coated with gold for 30 s. Finally, the samples were examined and imaged using a scanning electron microscope. The terminal deoxynucleotidyl transferase dUTP nick end labelling (TUNEL) assay was performed on different stages of mouse early embryos using the One Step TUNEL Apoptosis Assay Kit (Beyotime, C1086) according to the manufacturer's instructions. The embryos were washed three times with PBS-HSA and fixed in 4% PFA in PBS for 30 min, then permeabilized in 0.25% Triton X-100 in PBS-HSA for 1 h. The embryos were stained in TdT-FITC (fluorescein) solution at 37 °C for 60 min, washed with PBS-HSA and stained with DAPI for 10 min. Embryos mounted on glass slides were examined by Zeiss LSM 880 confocal laser scanning microscope. Positive labelling for nuclear accumulation of FITC indicated dead cells. The numbers of total cells and labelled dead cells were analysed using Fiji. For metabolite annotation, this study drew on the workflow proposed by Alexandra et al. 26 To reduce the error rate of metabolite annotation, subsequent analyses only included metabolites with level 1 or 2 identification confidence. After log2-transforming the metabolite intensities, orthogonal partial least squares discriminant analysis (OPLS-DA), the supervised multivariate analytic methods, were utilised to classify patients depending on their metabolic profiles with the ropls R package. 27 Then, we did a difference analysis on the metabolomes of the hydrosalpinx group and the control group, calculated the fold change (FC) and student t-test (two-tailed), and used Benjamini & Hochberg to adjust the P-value. Differentially expressed metabolites (DEMs) were defined as FC > 1.5 or <0.67 and P-adjust <0.05. DEMs were further included in pathway analysis based on Kyoto Encyclopedia of Genes and Genomes (KEGG) with the MetPath R package. 28 A generalised linear model was established using the glm function to explore the relationship between age and the levels of metabolites of interest. Processed scRNA-seq data of hydrosalpinx disease state and healthy state of human fallopian tubes were downloaded from GSE178101 , 22 loaded in the R environment and created a Seurat subject. 29 In quality control, genes expressed in fewer than 5 cells were removed. Cells were reserved if nFeature_RNA >300 and <9000, but cells with the percentage of mitochondrial content over 15%, ribosomal content less than 2%, or red blood cell content over 3% were removed. The following steps were conducted according to standard procedures of Seurat tutorial. After data normalising, identifying highly variable features, data scaling, and performing linear dimensional reduction, we also introduced Harmony to integrate samples. 30 Clusters were found with a resolution of 0.05. We selected marker genes used in the original study to annotate cell type: ciliated epithelial cell (EPCAM, FOXJ1, and CAPS), non-ciliated secretory epithelial cell (KRT7, PAX8, and OVGP1), fibroblast/myofibroblast (DCN, COL1A1, CD34, PDGFRA, POSTN, and NR2F2), smooth muscle cell (DES, ACTA2, and MYH11), pericyte (PDGFRB, MCAM, and CSPG4), blood endothelial cell (KDR, PECAM1, and VWF), lymphatic endothelial cell (PROX1, and PDPN), B cell (SDC1, and JCHAIN), T/NK cell (RUNX3, CD3E, and PTPRC), mast cell (KIT, MS4A2, TPSB2, and TPSAB1), and macrophage (FOLR2, CD68, CD163, and ITGAX). 22 Metabolism activity was quantified by scMetabolism with default parameters. 31 Data quality control of raw reads was first performed via Trim Galore (version 0.6.10) [ https://zenodo.org/record/7598955 ], including trimming adaptors, and removing low-quality reads (<Q20) and very short reads (read length <37 bp after adaptor trimming). Subsequently, STAR (version 2.7.0e) 32 was used to align clean reads to the Mus musculus genome (mm39) with default parameters. Reads with mapping quality less than 99.999% were removed with ‘samtools view -q 50` (samtools, version 0.1.18). 33 Uniquely aligned reads were counted using featureCounts (version 1.6.4) 34 and gencode annotations VM29. After filtering out outlier samples from the transcriptome and translatome data, DESeq2 35 was used to perform gene expression differential analysis and principal component analysis (PCA) on the inosine-treated vs control groups at the late two-cell (L2C) and late four-cell (L4C), respectively. Differentially expressed genes (DEGs) were defined as FC > 1.5 or <0.67 and P-adjust <0.05. At the L2C, genes were identified with consistent changes in both the transcriptomic and translatomics profiles. Specifically, genes that met the criteria of P-adjust <0.05 were screened from the DEG analysis and visualised using a scatter plot, with the x-axis and y-axis representing the log 2 FC of the transcriptome and translatome, respectively. The clusterProfiler package 36 was used to perform Gene Set Enrichment Analysis (GSEA) (genes with both P-adjust 1.5 & P-adjust <0.05) and downregulated (genes with both FC < 0.67 & P-adjust <0.05), respectively. Translation efficiency (TE) was evaluated at two distinct levels: global TE, calculated as ribo_sum FPKM/rna_sum FPKM, representing the overall translation efficiency across the transcriptome; and individual gene TE, calculated using the formula TE = ribo_individual FPKM/rna_individual FPKM, for each specific gene (High TE gene: TE > 1.5, low TE gene: TE < 0.67). Further comparison of global TE differences between the guanosine treatment group and the control group. Scatter plots were generated for genes differentially expressed at the translatome level, with the log 2 FC in the translatome plotted on the x-axis and the log 2 TE of individual genes plotted on the y-axis. Based on data and definitions from Sha's and Yu ’s study, 37 , 38 2279 zygotic genome activation (ZGA) genes (including 79 minor ZGA genes, 1877 major ZGA genes, and 323 persistent ZGA genes) and 4242 maternal mRNAs (comprising 1633 M-decay genes, 2236 Z-decay genes, and 373 persistent decay genes) were identified. Similar to the preceding analysis, gene expression changes (DEG: FC > 1.5 or <0.67 and P-adjust <0.05) from the L2C to L4C stage under identical treatment conditions were further characterised. The overlap DEGs between the transcriptome and translatome were visualised using the UpSetR package. 39 Subsequently, GO enrichment analysis was performed on subsets of these overlapping genes. Data are expressed as mean ± SD from a minimum of three independent replicates. Unless otherwise specified, group comparisons were performed using two-tailed unpaired Student's t-tests. P-values less than 0.05, 0.01, and 0.001 were considered statistically significant and are denoted by ∗, ∗∗, and ∗∗∗, respectively. The funders had no involvement in the conception, data acquisition, analysis, interpretation, or writing of this manuscript.

To elucidate metabolomic patterns of fallopian tube fluid in patients with hydrosalpinx (HSPX) compared to healthy fallopian tube controls (HFT), we recruited 52 premenopausal patients undergoing hysterectomy with benign diseases and 49 patients with hydrosalpinx. The clinical characteristics are summarised in Table S1 . Untargeted metabolomic analysis was subsequently performed ( Fig. 1 A). OPLS-DA revealed an apparent discrimination between FT fluid from HSPX and that from HFT ( Fig. 1 B). A total of 126 DEMs were identified between HSPX and HFT ( Table S3 ). Among these, 95 metabolites, such as 8-hydroxyguanosine (8-OHG), dinoprostone (PGE 2 ), phosphoenolpyruvate (PEP) and inosine, had increased levels and 31 metabolites, such as deoxycholic acid (DC), inosinic acid (IMP) and tryptamine, had lower levels in HSPX compared with HFT ( Fig. 1 C). The DEMs were categorised into several classes, including aldehyde, ketones, esters, amino acids, fatty acids (FA) and nucleotide and its metabolites. Among upregulated metabolites, amino acids and nucleotides and their metabolites were the most abundant, each comprising 24.24% ( Fig. S1A and B ). Of the 31 downregulated metabolites, 19.35% were amino acids and 6.45% were nucleotides and derivatives ( Fig. S1C ). Purine metabolism was identified as the most significantly upregulated metabolic pathway in HSPX compared to HFT ( Fig. 1 D). Additionally, metabolites associated with the AMPK signalling pathway and caffeine metabolism were reduced in HSPX group ( Fig. S1D ). These results suggest that hydrosalpinx exhibits a disease-specific metabolic microenvironment that differs significantly from that of the healthy fallopian tube. Fig. 1 Metabolic characterisation of tubal fluid between hydrosalpinx and healthy fallopian tubes . (A) Schematic overview of the study design and workflow of the analysis. (B) OPLS-DA score plot showing metabolic separation between HSPX and Healthy FT. (C) Volcano plots highlighting DEMs in HSPX vs Healthy FT. P-adjust value < 0.05 (two-tailed Student's t-test), Fold Change ≥1.5 or <0.67. (D) KEGG pathway enrichment of upregulated metabolites in the HSPX. (E) ELISA validation of adenosine, inosine, and hypoxanthine concentrations in FT fluid (n = 15 per group). Data are presented as mean ± SD. Statistical analysis was performed by unpaired two-tailed Student's t tests. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. Metabolic characterisation of tubal fluid between hydrosalpinx and healthy fallopian tubes . (A) Schematic overview of the study design and workflow of the analysis. (B) OPLS-DA score plot showing metabolic separation between HSPX and Healthy FT. (C) Volcano plots highlighting DEMs in HSPX vs Healthy FT. P-adjust value < 0.05 (two-tailed Student's t-test), Fold Change ≥1.5 or <0.67. (D) KEGG pathway enrichment of upregulated metabolites in the HSPX. (E) ELISA validation of adenosine, inosine, and hypoxanthine concentrations in FT fluid (n = 15 per group). Data are presented as mean ± SD. Statistical analysis was performed by unpaired two-tailed Student's t tests. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. To validate the upregulation of purine metabolites in HSPX FT fluid, we used ELISA to quantify the levels of three representative compounds (adenosine, inosine, and hypoxanthine) within the purine metabolic pathway identified by untargeted metabolism, each group with 15 biological replicates. As expected, adenosine, inosine, and hypoxanthine levels were significantly higher in the HSPX group compared to the HFT group ( Fig. 1 E). In metabolomics data, the top two significantly upregulated compounds, 8-OHG and PGE 2 , are inflammation-related molecules that the former acts as a biomarker for oxidative damage in DNA 40 , 41 and the latter functions as a pro-inflammatory mediator in vivo . 42 The levels of 8-OHG and PGE 2 and pro-inflammatory factor IL-6, were higher in the HSPX group in ELISA results ( Fig. S1E ). These results indicate that the purine metabolism pathway is significantly upregulated in the HSPX group. Previous studies have revealed a significant increase in purine metabolism at the blastocyst stage of mouse embryo compared to earlier developmental stages. In particular, inosine levels are significantly elevated in blastocysts relative to the 2-cell stage, 6 implying a limited or adverse role for purine metabolism during early stages of embryogenesis. Given the ethical and limited availability of clinical materials, validation in human embryos is challenging. Therefore, we used mouse embryos to investigate the effects of inosine on embryonic development. To investigate the impact of inosine on early embryo development, zygotes were cultured with varying concentrations (0–50 μM) of permeable inosine through the blastocyst stage ( Fig. 2 A). While 5 μM inosine had no discernible effect, 12.5 μM inosine significantly reduced the blastocyst formation rate, at concentrations with low cytotoxicity that did not affect the viability of embryos ( Fig. S2A ). At higher concentrations (25 and 50 μM), inosine induced a dosage-dependent developmental arrest at the 2- to 4-cell stage ( Fig. 2 B and C). Although a small fraction of embryos exposed to 25 μM inosine were able to reach the blastocyst stage, the total number of blastomeres in these blastocysts was significantly reduced compared with those in the control group, indicating inosine may affect the proliferation and differentiation in early embryo development ( Fig. S2B and C ). Inner cell mass development was also impaired in inosine treatment group, as evidenced by a reduced number of OCT4-positive cells ( Fig. S2D ). However, no significant differences in OCT4 protein expression were observed between the two treatment groups ( Fig. S2E ). Fig. 2 Inosine impedes mouse early embryo development at the 2–4 cell stage . (A) Experimental schematic. In vivo fertilised zygotes (2 PN stage) were cultured in KSOM medium with or without inosine and monitored for developmental progression. (B) Representative images of embryos from 2-cell to blastocyst stages following 25 μM inosine or 1.5% DMSO (control) treatment. Scale bar, 100 μm. n = 30 embryos per group. (C) Developmental rates based on initial 2 PN embryo counts. Data represented mean ± SD from three biological replicates. (D) Schematic of different 25 μM inosine treatment windows (green lines). (E) Developmental stage distributions under treatment schemes shown in (D). DMSO (n = 69), A (n = 59), B (n = 58), C (n = 59), D (n = 40), and E (n = 73). This experiment was repeated independently at least three times. Inosine impedes mouse early embryo development at the 2–4 cell stage . (A) Experimental schematic. In vivo fertilised zygotes (2 PN stage) were cultured in KSOM medium with or without inosine and monitored for developmental progression. (B) Representative images of embryos from 2-cell to blastocyst stages following 25 μM inosine or 1.5% DMSO (control) treatment. Scale bar, 100 μm. n = 30 embryos per group. (C) Developmental rates based on initial 2 PN embryo counts. Data represented mean ± SD from three biological replicates. (D) Schematic of different 25 μM inosine treatment windows (green lines). (E) Developmental stage distributions under treatment schemes shown in (D). DMSO (n = 69), A (n = 59), B (n = 58), C (n = 59), D (n = 40), and E (n = 73). This experiment was repeated independently at least three times. To evaluate whether these developmental defects compromise functional outcomes in vivo , we performed embryo transfer experiments ( Fig. S2F ). Blastocysts pre-treated with 25 μM inosine exhibited a significantly lower decidualization rate at E7.5 compared to controls ( Fig. S2G–I ). These results demonstrate that early exposure to inosine not only arrests cleavage but also impairs the implantation competency of the surviving embryos. To identify the specific developmental stage sensitive to inosine and assess whether its inhibitory effects were reversible upon withdrawal, we treated embryos with 25 μM inosine starting at 0.5, 1.5, 2.5 and 3.5 days post coitum (d.p.c.), respectively ( Fig. 2 D). Interestingly, we found that inosine treatment starting at 0.5 d.p.c. had the most detrimental effect on mouse early embryo development, suggesting that inosine primarily affects development during the 2-cell to 4-cell stage ( Fig. 2 E and Fig. S2J ). Notably, embryos exposed to inosine only prior to 2.5 d.p.c. (Group E) failed to recover their developmental potential upon transfer to inosine-free medium, mirroring the arrest observed in the continuous treatment group (Group A). These results suggest that inosine-mediated damage during early cleavage is persistent despite the withdrawal of treatment, identifying the period prior to 2.5 d.p.c. as a critical developmental window of vulnerability. As a pivotal enzyme in the downstream metabolism of inosine, PNP is essential for inosine-induced proliferation. 43 We speculated that inhibiting PNP may salvage the adverse effects of inosine on mouse early embryo development. To confirm this, we assessed the embryonic development rate in the presence of forodesine (foro, as PNP's inhibitor). 44 Preliminary experiments confirmed that forodesine alone (up to 50 μM) did not impact embryonic development ( Fig. 3 A and B). Notably, the addition of 5 μM forodesine significantly rescued the inhibitory effect of inosine on the blastocyst formation rate ( Fig. 3 C–F). Fig. 3 Inosine supplementation impairs mouse early embryo development but is rescuable by forodesine . (A) Developmental rates of mouse embryos cultured in KSOM with DMSO or forodesine. Data are presented as mean ± SD of three biological replicates. (B) Representative confocal max-projections of TUNEL-stained embryos treated with indicated forodesine concentrations (0–50 μM). Scale bar, 50 μm. (C) Experimental schematic and representative images of embryos from 2-cell to blastocyst stages treated with DMSO, 25 μM inosine, or inosine plus forodesine. Scale bar, 100 μm. n > 30 embryos per group. (D, E) Developmental rates at sequential stages (D) and specifically at the blastocyst stage (E) under indicated treatment schemes. (F) qPCR analysis of Pnp mRNA levels in late 2-cell embryos (n = 5 embryos per sample). Gapdh served as the internal control. Data are presented as mean ± SD from three independent experiments. Statistical significance determined by two-tailed unpaired t-tests. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. Inosine supplementation impairs mouse early embryo development but is rescuable by forodesine . (A) Developmental rates of mouse embryos cultured in KSOM with DMSO or forodesine. Data are presented as mean ± SD of three biological replicates. (B) Representative confocal max-projections of TUNEL-stained embryos treated with indicated forodesine concentrations (0–50 μM). Scale bar, 50 μm. (C) Experimental schematic and representative images of embryos from 2-cell to blastocyst stages treated with DMSO, 25 μM inosine, or inosine plus forodesine. Scale bar, 100 μm. n > 30 embryos per group. (D, E) Developmental rates at sequential stages (D) and specifically at the blastocyst stage (E) under indicated treatment schemes. (F) qPCR analysis of Pnp mRNA levels in late 2-cell embryos (n = 5 embryos per sample). Gapdh served as the internal control. Data are presented as mean ± SD from three independent experiments. Statistical significance determined by two-tailed unpaired t-tests. ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001. To further strengthen these findings and exclude potential off-target effects of the chemical inhibitor, we performed siRNA-mediated knockdown of PNP in zygotes ( Fig. S3A ). Consistent with the pharmacological results, PNP knockdown significantly alleviated the embryonic developmental arrest induced by 25 μM inosine, mirroring the rescue effect observed with forodesine ( Fig. S3B–D ). Collectively, these pharmacological and genetic lines of evidence demonstrate that PNP is indispensable for inosine-mediated arrest during mouse early embryo development. To elucidate the mechanisms underlying the arrest of mouse embryos at the two- and four-cell stages induced by inosine, we applied ultrasensitive translatomic and total RNA-seq to L2C and L4C embryos treated with 25 μM inosine or 1.5‰ DMSO as control ( Fig. 4 A, Tables S4 and S5 ). PCA of the transcriptome and translatome data showed a clear separation between control and inosine-treated L2C embryos, whereas L4C embryos displayed considerable overlap between the groups ( Fig. 4 B), indicating that inosine primarily impacts L2C embryos. Volcano plots illustrate activated and repressed genes identified via transcriptomics and translatomics in L2C and L4C embryos ( Fig. 4 C and Fig. S4A ). Interestingly, translatomic profiles showed more pronounced alterations than transcriptomics in treated L2C embryos, with a greater number of differentially translated genes, suggesting that inosine primarily impacts translational regulation. We therefore assessed global TE at both stages ( Fig. 4 D). TE was reduced in the inosine-treated group relative to controls at the L2C and L4C stages, with a significant difference at L4C, implying that translational deficiency at the two-cell stage may undermine subsequent developmental potential, including progression to the four-cell stage and blastocyst formation. Furthermore, we validated downregulation of eukaryotic translation initiation factor 4e1b (Eif4e1b) in inosine-treated L2C embryos ( Fig. 4 E, Fig. S4B ), a factor essential for selective translation activation during MZT in mammals. 45 Previous studies report that proteins downregulated in Eif4e1b -cKO GV oocytes are involved in microtubule cytoskeleton organisation, and Eif4e1b -deficient embryos arrest at the two-cell stage, 45 consistent with our results. Fig. 4 Translatomics and transcriptomics of control and inosine-treated late 2-/4-cell embryos . (A) Schematic of the T&T-seq workflow for late 2-cell (L2C) and 4-cell (L4C) embryos (n = 3–4 replicates of 5 embryos per group) after control or inosine treatment. (B) PCA plot of T&T-seq data showing clustering L2C embryos and L4C embryos. (C) Volcano plot of DEGs in L2C embryos. P-adjust 1.5 or <0.67. (D) Global translation efficiency (TE) in control and inosine-treated embryos at L2C and L4C stages. (E) Transcriptional and translational levels of the Eif4e1b in L2C embryos. (F) Scatter plot of DEGs in L2C embryos. Class I (yellow) and Class II (blue) represent genes with consistent up- and down-regulation in both transcriptome and translatome, respectively. (G) Volcano plot showing TE differences (log 2 TE) in L2C embryos. (H) Heatmaps of downregulated genes in both the transcriptome and translatome of L2C. (I) Immunofluorescence of CHEK1 (yellow) and DAPI (blue) in L2C embryos (n = 1 embryo per dot). Scale bar, 20 μm ∗∗P < 0.01, ∗∗∗P < 0.001. P-values were calculated by two-tailed Student's t-test. n indicates the number of embryos were analysed. Translatomics and transcriptomics of control and inosine-treated late 2-/4-cell embryos . (A) Schematic of the T&T-seq workflow for late 2-cell (L2C) and 4-cell (L4C) embryos (n = 3–4 replicates of 5 embryos per group) after control or inosine treatment. (B) PCA plot of T&T-seq data showing clustering L2C embryos and L4C embryos. (C) Volcano plot of DEGs in L2C embryos. P-adjust 1.5 or <0.67. (D) Global translation efficiency (TE) in control and inosine-treated embryos at L2C and L4C stages. (E) Transcriptional and translational levels of the Eif4e1b in L2C embryos. (F) Scatter plot of DEGs in L2C embryos. Class I (yellow) and Class II (blue) represent genes with consistent up- and down-regulation in both transcriptome and translatome, respectively. (G) Volcano plot showing TE differences (log 2 TE) in L2C embryos. (H) Heatmaps of downregulated genes in both the transcriptome and translatome of L2C. (I) Immunofluorescence of CHEK1 (yellow) and DAPI (blue) in L2C embryos (n = 1 embryo per dot). Scale bar, 20 μm ∗∗P < 0.01, ∗∗∗P < 0.001. P-values were calculated by two-tailed Student's t-test. n indicates the number of embryos were analysed. Moreover, we integrated transcriptomic and translatomic data to analyse transcriptional and translational dynamics in L2C embryos under different treatments. In the inosine group, 368 genes were co-upregulated and 200 co-downregulated in both transcriptome and translatome ( Fig. 4 F). Co-upregulated DEGs were linked to ribosome biogenesis and cytoplasmic translation ( Fig. S4C ), while co-downregulated DEGs were associated with actin cytoskeleton, microtubules and mitotic cell cycle phase transition ( Fig. S4D ). GSEA showed that downregulated transcribed and translated genes were enriched in actin cytoskeleton regulation ( Fig. S4E ). TE was suppressed in inosine-treated embryos: among 1189 translationally downregulated genes, 22% had TE 1.5 ( Fig. 4 G). Low-TE genes were enriched for “tubulin binding”, “GTPase regulator activity”, and “actin binding”, implying impaired actin cytoskeleton function after inosine treatment ( Fig. 4 H and Fig. S4F ). CHEK1 mutations perturb microtubule by mainly affecting F-actin excessive depolymerisation and resulting cell cycle arrest. 46 CHEK1 showed downregulation in inosine-treated L2C as revealed by immunofluorescence ( Fig. 4 I). We also compared changes in mRNA and translating mRNA levels between control and inosine-treated embryos from L2C to L4C stages ( Fig. S5A and B ). Analysis focused on two gene sets: those upregulated in controls but unchanged with inosine (light blue bar), and those stable in controls but downregulated with inosine (dark blue bar). GO analysis showed that light bar genes were enriched in “mRNA splicing” and “translation regulator activity” ( Fig. S5C and D ), suggesting inosine may disrupt splicing and translation activation during this transition. Zygotic splicing activation is crucial for ZGA and totipotency-to-pluripotency transition. 47 Dark bar genes were enriched for “microtubule cytoskeleton organisation involved in mitosis”, “mitotic spindle organisation”, and “RNA polymerase II transcription regulator complex” ( Fig. S5E and F ), indicating inosine disturb mitotic spindle formation, consistent with prior findings. During MZT, maternal mRNAs are degraded as ZGA initiates, primarily through maternal-decay (M-decay) and zygotic-decay (Z-decay), with elimination largely complete by the two-cell stage in mouse embryos. 48 , 49 ZGA occurs in two sequential waves (minor ZGA and major ZGA) during MZT. 50 , 51 , 52 Inosine-induced developmental arrest in mouse 2- to 4-cell stage suggests that inosine may impair the MZT process at the two-cell stage. Based on published RNA-seq data 37 and previously defined gene sets from previous report, 38 we compared transcriptomic and translatomic DEGs from inosine-treated L2C embryos against two reference groups: 2279 ZGA genes and 4242 maternal mRNAs. GSEA demonstrated that minor ZGA, persistent ZGA genes and Z-decay genes were globally decreased in inosine-treated L2C embryos at transcriptome and translatome level ( Fig. 5 A and B). Representative ZGA genes include Usp17le , 53 Zscan4a , 54 Ddit4l , 55 and Obox3 56 ( Fig. 5 C), representative Z-decay genes include Ccnt2 , 57 Dazl , 58 and Nlrp5 59 ( Fig. 5 D). Furthermore, expression of MERVL, a murine endogenous retrovirus-L element and totipotency marker essential for preimplantation development, 60 was downregulated following inosine treatment, suggesting potential ZGA defects ( Fig. 5 E). Furthermore, the protein levels of CCNT2 (Cyclin T2), which is involved in the Z-decay pathway that regulate ZGA during early embryogenesis, were significantly reduced in inosine-treated L2C embryos ( Fig. 5 F). Depletion of CCNT2 during the zygotic stage is known to cause developmental arrest at the 2- or 4-cell stage, 57 a phenotype consistent with inosine-induced arrest. In summary, our results indicate that inosine exposure in mouse 2-cell embryos is associated with repression of ZGA genes, particularly minor ZGA genes and persistent ZGA genes, and degradation of Z-decay genes. Fig. 5 Inosine impairs MZT in late 2-cell stage mouse embryos . (A, B) GSEA showing downregulation of minor ZGA, persistent ZGA, and maternal Z-decay genes at the transcriptome (A) and translatome (B) levels. (C, D) Expression (TPM) of representative ZGA (C) and Z-decay (D) genes downregulated in both analyses. (E, F) Representative immunofluorescence images and relative intensity quantification of MERVL (E, purple) and CCNT2 (F, green) in L2C embryos. Scale bars, 20 μm. Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, two-tailed unpaired t-tests. n indicates the number of embryos analysed. Inosine impairs MZT in late 2-cell stage mouse embryos . (A, B) GSEA showing downregulation of minor ZGA, persistent ZGA, and maternal Z-decay genes at the transcriptome (A) and translatome (B) levels. (C, D) Expression (TPM) of representative ZGA (C) and Z-decay (D) genes downregulated in both analyses. (E, F) Representative immunofluorescence images and relative intensity quantification of MERVL (E, purple) and CCNT2 (F, green) in L2C embryos. Scale bars, 20 μm. Data are presented as mean ± SD. ∗P < 0.05, ∗∗P < 0.01, two-tailed unpaired t-tests. n indicates the number of embryos analysed. To identify the cellular source of inosine, we reanalysed scRNA-seq data from FT tissues ( Fig. 6 A). 22 Pathway enrichment analysis revealed that NCSE cells exhibit the highest purine metabolism activity among all cell subtypes ( Fig. 6 B). Further comparison showed that while purine metabolism scores were statistically higher in both total cell types and NCSE cells from the HSPX group, the absolute magnitude of this metabolic increase was relatively modest ( Fig. 6 C). We characterised the morphology of FT tissue from patients with HFT and HSPX using H&E stanning and scanning electron microscopy (SEM) ( Fig. S6A–D ). HSPX tissue exhibited an increased presence of chronic inflammatory cells, including lymphocytes (black arrows) and plasma cells (red arrows) ( Fig.S6B ). Furthermore, compared to HFT tissue, HSPX tissue showed extensive damage and sloughing of the mucosal epithelial layer, with adhesion of ciliated cells and destruction of secretory cells, accompanied by oedema and mucous deposition ( Fig. S6C and D ). We further examined whether HSPX tissue exhibited more apoptosis-associated cellular features than HFT tissue. NCSE cells in HSPX also had higher levels of γH2AX, active caspase-3/7 and TUNEL-positive cells than NCSE cells in HFT, respectively, as indicators of DNA damage and apoptosis ( Fig. 6 D and E). Therefore, we hypothesised that the elevated levels of purine metabolites in hydrosalpinx FT fluid may stem from the breakdown of NCSE cells and subsequent release of purine metabolites. Purine metabolic pathway as summarised in Fig. S6E . Fig. 6 Non-ciliated secretory epithelial (NCSE) cells injury elevates inosine levels in hydrosalpinx . (A) UMAP visualisation of FT tissue clusters from the Ulrich et al. dataset. (B) Dot plot of metabolic pathways expression across 11 major FT cell subtypes. (C) Purine metabolism scores in total cells and NCSE cells from healthy controls and patients with hydrosalpinx. (D, E) Representative immunofluorescence of HFT and HSPX fallopian tube epithelia, showing staining for (D) activated caspase-3 and γH2AX (green), and (E) TUNEL (green), with co-staining for acetylated tubulin (AcTub, red; a ciliated cell marker) and DAPI (blue). Scale bar, 100 μm. Non-ciliated secretory epithelial (NCSE) cells injury elevates inosine levels in hydrosalpinx . (A) UMAP visualisation of FT tissue clusters from the Ulrich et al. dataset. (B) Dot plot of metabolic pathways expression across 11 major FT cell subtypes. (C) Purine metabolism scores in total cells and NCSE cells from healthy controls and patients with hydrosalpinx. (D, E) Representative immunofluorescence of HFT and HSPX fallopian tube epithelia, showing staining for (D) activated caspase-3 and γH2AX (green), and (E) TUNEL (green), with co-staining for acetylated tubulin (AcTub, red; a ciliated cell marker) and DAPI (blue). Scale bar, 100 μm. Taken together, these findings indicate that inosine may function as a dysregulator in the hydrosalpinx fallopian tube that is harmful to early embryo development.

The fundamental question regarding fertility in hydrosalpinx (HSPX) involves identifying the factors that trigger reduced implantation rate and understanding how they influence embryo development. Here, we systematically unravel the metabolic perturbations in the HSPX microenvironment and their impact on early embryo development. HSPX exhibits a distinct metabolic signature, with purine metabolism emerging as the most prominently upregulated pathway. This study identified markedly elevated purine metabolites, particularly inosine, in the HSPX microenvironment. Functionally, inosine-mediated cell cycle disruption leads to MZT failure and reduced translation efficiency, resulting in embryo arrest at the 2- to 4-cell stage and defects in microtubule cytoskeleton organisation. This inhibitory effect is dependent on PNP, as its inhibitor, forodesine, partially rescues embryonic development. These findings establish a mechanistic link between purine metabolic dysregulation in HSPX and early embryonic failure, offering fresh insight into the metabolic aetiology of tubal infertility. In this study, upregulated DEMs were enriched in purine metabolism, a profile that aligns with the specialised metabolic signature of NCSE cells in the fallopian tube. Consequently, we speculate that purine-related compounds, such as inosine, primarily originate from these cells. Our observation of increased apoptotic signalling in HSPX epithelial cells aligns with findings by Ulrich et al., who noted a significant reduction in the proportion of NCSE cells in HSPX (23.9% in health and 15.2% in hydrosalpinx). 22 This suggests that the elevated luminal inosine levels result from the leakage of purine metabolites following NCSE cell apoptosis, rather than purely from upregulated synthesis. Furthermore, the chronic inflammatory microenvironment of hydrosalpinx, which is characterised by lymphocyte infiltration and hypoxia, likely triggers purinergic signalling and leads to the additional extracellular release of nucleotides. 61 , 62 , 63 Based on these findings, we propose that while the pathological microenvironment mildly stimulates local metabolism as a secondary factor, the extensive damage and apoptosis of epithelial cells serve as the primary driver of elevated inosine in hydrosalpinx. Inosine-arrested L2C embryos exhibited a marked reduction in CHEK1, a critical kinase that ensures proper mitotic progression and prevents chromosomal instability. 64 This finding aligns with recent evidence suggesting that excessive inosine can trigger DNA replication stress and polymerase η-dependent cell cycle arrest. 65 , 66 Such cell cycle blockage may serve as one of the main reasons for ZGA and M-decay abnormalities. Furthermore, microtubule cytoskeletal abnormalities are established causes of early embryo arrest in both humans and mice. 67 Consistent with this, we observed the downregulation actin cytoskeleton-related genes in inosine-treated L2C embryos exhibiting low TE. Of note, maternal deficiency in Eif4e1b has been shown to cause 2-cell embryo arrest in mice and disrupt microtubule organisation. 45 This highlights the critical role of translational regulation in maintaining cytoskeletal integrity and supports the notion that inosine impairs development from the 2- to 4-cell stage by disrupting translation efficiency, thereby compromising cytoskeletal function. Our study has certain limitations. The lack of accessible tubal fluid from healthy individuals precluded the inclusion of ideal healthy controls; we therefore used samples from patients with benign gynaecological conditions as the control group. Secondly, owing to limitations in existing reference databases and available annotation methodologies, it remains challenging to accurately identify all metabolites. As described in the Methods section, we excluded metabolites supported by low levels of evidence to ensure the highest possible accuracy. Furthermore, while our functional assays confirmed that inosine alone is sufficient to induce embryonic arrest, it likely acts in concert with other upregulated purines such as adenosine and hypoxanthine. The potential synergistic effects or independent toxicity of these interrelated metabolites within the purine family require further investigation. Additionally, the paucity of available human embryos presents a significant constraint, hampering further direct exploration of the effects of metabolites on their regulatory mechanisms. Although we propose inosine as a potential clinical biomarker, this study lacks validation in an independent human cohort and a direct correlation with clinical reproductive outcomes, such as implantation or live birth rates. While our findings provide a robust mechanistic proof of concept for embryonic arrest caused by inosine, large scale prospective clinical studies are required in the future to confirm its diagnostic sensitivity and predictive value for personalised infertility management. In conclusion, our study unveils the metabolic aetiology of hydrosalpinx-related infertility and demonstrates that inosine, a metabolite markedly elevated in hydrosalpinx fluid, is a key causative factor in preimplantation embryonic arrest. Mechanistically, inosine-induced cell cycle arrest impairs MZT and suppresses the translation efficiency of cytoskeleton-related genes. Intervention in the apoptosis and necrosis processes of fallopian tube epithelial cells may reduce the release and accumulation of inosine at the source. While PNP inhibition with forodesine partially alleviates inosine-induced embryotoxicity in vitro , its therapeutic potential remains highly exploratory. It is important to note that forodesine functions as a targeted molecular intervention rather than a comprehensive treatment for hydrosalpinx. Furthermore, whether forodesine can reduce the clinical severity or inflammatory state of the fallopian tube itself requires further investigation. Given these uncertainties, significant translational barriers exist regarding systemic safety and long-term feasibility. Extensive preclinical studies are required to bridge the gap between these in vitro findings and clinical application, particularly in assessing the long-term safety of modulating the tubal metabolic environment.

Conceptualisation: LY, PY, HL and JQ. Data curation: YP and YF. Formal Analysis: YP and YF. Funding acquisition: HL, LY, YF and PY. Investigation: YP, YF, LY, PY, HL and JQ. Methodology: YP, YF, NW and XW. Project administration: LY., PY., HL., and JQ. Visualisation: YP and YF. Resources: DL, ML, YL, HL and JQ. Software: YP and YF. Supervision: LY, PY, HL and JQ. Writing—original draft: YP and YF. Writing—review & editing: YP, YF, LY, PY, HL and JQ. YP and YF had full access to all the data in the study and verified the underlying data. All authors read and approved the final manuscript. HL and JQ made the decision to submit the manuscript.

In mammals, the fallopian tube (FT) is a pivotal female reproductive organ that connects the ovary to the uterus, providing a suitable and safe microenvironment for gamete transport, fertilisation and preimplantation embryo development. After fertilisation, both mouse and human embryos develop to the blastocyst stage within the FT before entering the uterus for implantation. This process takes approximately 4 days in mice 1 and 5–6 days in humans. 2 During this period, the fertilised embryo in the FT undergoes critical biological processes such as the maternal-to-zygotic transition (MZT), and epigenetic reprogramming. 3 , 4 , 5 , 6 Moreover, the abundant nutrients and factors within the FT fluid support early embryonic metabolic switching and protect embryos against stress and immune responses. Simulating the physiological conditions of FT fluid has been shown to optimise in vitro fertilisation-embryo transfer (IVF-ET) culture systems for enhanced embryonic development. Tubal and peritoneal pathologies account for 30–35% of female infertility. 7 Hydrosalpinx, a common manifestation of tubal factor infertility, accounts for 10–30% among infertile women. 8 Hydrosalpinx adversely affects clinical IVF outcomes, reducing pregnancy and implantation rates and increasing early pregnancy loss. 9 , 10 , 11 Current understanding suggests two primary mechanisms by which hydrosalpinx leads to infertility. First, cytotoxic constituents within the tubal fluid, or their reflux into the uterine cavity, exert deleterious effects on endometrial receptivity, ultimately contributing to adverse reproductive outcomes. Increasing evidence implies that toxic molecules may be present in tubal fluid. Exposure to hydrosalpinx fluid significantly impairs blastocyst formation in both humans and mice, inhibits blastocyst hatching, and reduces the formation of high-quality blastocysts. 12 , 13 , 14 , 15 , 16 In addition, mechanical obstruction and impaired ciliary function in the FT may also contribute to infertility by interfering with gamete transport and embryonic development. 17 , 18 , 19 , 20 However, the composition of hydrosalpinx has not been fully explored, and the specific mechanisms by which hydrosalpinx fluid affects preimplantation embryogenesis remain poorly understood. Previous investigations employing single-cell transcriptomic and proteomic approaches have identified alterations in the microenvironment associated with hydrosalpinx. 21 , 22 In addition to cells and large molecules in tubal fluid, small molecules such as metabolites remain underexplored. Metabolites are directly involved in embryo development and gene regulation. For example, L-2-HG inhibits α-KG-dependent dioxygenases, thereby impairing histone demethylation (H3K4me3/H3K9me3) and leading to aberrant epigenetic reprogramming and developmental defects. 6 Metabolomics reflects alterations in endogenous substances and is directly associated with phenotypic characteristics. In this study, we investigated the metabolic profiles and differential metabolites between hydrosalpinx fluid and normal FT flushing fluid. Key metabolites potentially associated with impaired preimplantation embryo development in mice were identified, and further explored their underlying mechanisms.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.