Hyperhomocysteinemia reduces the high-quality embryo rate in PCOS patients undergoing IVF/ICSI: clinical evidence and a preliminary exploration of mechanisms in KGN cells.

The workflow for patient selection is illustrated in Fig.  1 . Initially, 7,337 cycles were screened, including cases with potential confounding variables. After applying predefined inclusion and exclusion criteria, 1,083 patients were considered eligible for analysis. Among the remaining cohort, 484 cases were excluded due to unavailable Hcy measurements. The final analytical population was then categorized into three groups based on serum Hcy levels using a trichotomous stratification approach (Fig.  1 ). Univariate analysis revealed statistically significant differences in high-quality embryo rates among the three Hcy groups. Aside from Hcy levels and related metabolic parameters (e.g., triglycerides), which were used for stratification, no statistically significant differences were observed among the groups in key demographic and clinical baseline characteristics, including age, body mass index (BMI), and duration of infertility (Table  1 ). Table 1 Baseline characteristics of participants GROUP Group 1 Group 2 Group 3 P -value number 199 200 200 Hcy(umol/L) ≤ 8.63 8.63–11.10 ≥ 11.10 < 0.001 AMH(ng/mL) 7.04 ± 3.59 7.48 ± 4.08 7.51 ± 4.34 0.430 TSH(mIU/mL) 3.64 ± 2.56 3.64 ± 2.92 3.27 ± 1.93 0.232 E2(pg/mL) 37.04 ± 21.78 39.48 ± 28.53 39.59 ± 27.86 0.549 PRL(ng/mL) 22.59 ± 11.35 21.36 ± 11.71 22.35 ± 12.27 0.542 fasting plasma glucose(mmol/L) 4.84 ± 0.54 5.01 ± 1.65 4.85 ± 0.80 0.220 total cholesterol(mmol/L) 4.32 ± 0.80 4.29 ± 0.74 4.23 ± 0.78 0.552 triglyceride(mmol/L) 1.43 ± 1.07 1.46 ± 0.98 1.19 ± 0.64 0.004 high density lipoprotein(mmol/L) 1.33 ± 0.29 1.32 ± 0.28 1.38 ± 0.29 0.068 low density lipoprotein(mmol/L) 2.47 ± 0.63 2.41 ± 0.63 2.46 ± 0.67 0.683 duration of infertility (year) 3.06 ± 1.89 3.54 ± 2.26 3.40 ± 2.13 0.068 body mass index(kg/m 2 ) 24.26 ± 2.86 23.89 ± 3.20 24.65 ± 3.28 0.127 antral follicle count 18.17 ± 8.08 20.39 ± 8.87 20.58 ± 10.29 0.014 menstrual cycle(day) 33.69 ± 10.87 34.43 ± 13.24 36.88 ± 16.93 0.057 duration of ovarian stimulation(day) 10.40 ± 1.91 10.20 ± 1.71 10.21 ± 2.04 0.512 Cumulative gonadotropin dose (IU) 2337.50 ± 799.32 2304.44 ± 768.20 2277.81 ± 813.95 0.754 endometrial thickness of HCG day(mm) 10.94 ± 2.83 10.83 ± 2.67 10.91 ± 2.73 0.371 MII rate(%) 87.87 ± 15.78 88.63 ± 14.44 88.47 ± 14.93 0.868 2PN rate(%) 64.11 ± 19.75 65.72 ± 20.56 63.92 ± 21.24 0.628 high quality embryos rate(%) 59.35 ± 21.20 54.71 ± 19.77 46.97 ± 22.50 < 0.001 type of infertility(%) 0.369  primary infertility 61.3 54.5 56.5  secondary infertility 38.7 45.5 43.5 fertilization method(%) 0.978  IVF 77.9 77.5 77.0  ICSI 22.1 22.5 23.0 pregnancy(%) 0.575  yes 44.2 40.0 39.5  no 55.8 60.0 60.5 Baseline characteristics of participants Multivariate regression analysis revealed a significant negative association between serum Hcy levels and the high-quality embryo rate (β = -1.04, 95% confidence interval [CI]: -1.36 to -0.72; p  < 0.0001). This relationship remained robust after adjustment for potential confounders, i.e., creatinine clearance rate, female age, anti-Müllerian hormone (AMH), antral follicle count (AFC), thyroid-stimulating hormone (TSH), total cholesterol (TC), triglycerides (TG), and endometrial thickness (β = -1.06, 95% CI: -1.40 to -0.72; p  < 0.0001) (Table  2 ). No statistically significant differences in clinical pregnancy rates were observed among the Hcy-level groups. However, higher Hcy levels were inversely associated with the cleavage-stage high-quality embryo rate, with each 1 µmol/L increase in Hcy corresponding to an approximate 1.4% reduction in this parameter. Table 2 Multivariate analysis of factors associated with high-quality embryo rate Non-adjusted Adjusted HCY -1.04 (-1.36, -0.72) < 0.0001 -1.06 (-1.40, -0.72) < 0.0001 Adjusted: female age; AMH; TSH; FSH; LH; E2; PRL; fasting plasma glucose; total cholesterol; triglyceride; high density lipoprotein; lowdensity lipoprotein; duration of infertility; body mass index; antral follicle count; menstrual cycle; endometral thickness of HCG day Multivariate analysis of factors associated with high-quality embryo rate Adjusted: female age; AMH; TSH; FSH; LH; E2; PRL; fasting plasma glucose; total cholesterol; triglyceride; high density lipoprotein; lowdensity lipoprotein; duration of infertility; body mass index; antral follicle count; menstrual cycle; endometral thickness of HCG day Curve-fitting analysis was performed to further characterize the relationship between serum Hcy levels and the high-quality embryo rate (Fig.  2 ). The results showed a progressive decline in the high-quality embryo rate as serum Hcy levels increased, indicating a dose-dependent negative association. Fig. 2 Curve modeling of the association between serum Hcy and high-quality embryo rate. (The red curve represents the curve fitting of the high-quality embryo rate, and the blue curve represents the 95% confidence interval) Curve modeling of the association between serum Hcy and high-quality embryo rate. (The red curve represents the curve fitting of the high-quality embryo rate, and the blue curve represents the 95% confidence interval) We found treating with 80 µM Hcy for 24 h significantly decreased cell viability (Fig.  3 ). Fig. 3 The effect of Hcy on the activity of KGN cells. (“ns” indicates no statistically significant difference; “****” denotes a P -value < 0.0001) The effect of Hcy on the activity of KGN cells. (“ns” indicates no statistically significant difference; “****” denotes a P -value < 0.0001) To elucidate the mechanism underlying KGN cell death, flow cytometry analysis using Annexin V staining was performed to observe apoptosis (Fig.  4 ). Early apoptotic cells and late apoptotic or necrotic cells were represented in the Q3 and Q2 quadrants, respectively. The results showed that, compared with the control group, treatment with progressively increasing concentrations of Hcy significantly elevated the overall apoptotic cell rate, indicating that Hcy promotes apoptosis in KGN cells in a concentration-dependent manner. Fig. 4 The effect of Hcy exposure in KGN cells The effect of Hcy exposure in KGN cells Based on target identification, 102 potential Hcy-related therapeutic targets for PCOS were identified (Fig.  5 A). A PPI network was constructed using the STRING database (Fig.  5 B). Fig. 5 Venn diagram, PPI network of potential targets, KEGG, and GO enrichment analysis Venn diagram, PPI network of potential targets, KEGG, and GO enrichment analysis KEGG pathway enrichment analysis of these 102 targets identified the top 20 significantly enriched pathways (Fig.  5 C), primarily involving lipid metabolism and atherosclerosis, hepatitis B, cytomegalovirus infection, the AGE-RAGE signaling pathway in diabetic complications, and the TNF signaling pathway. GO enrichment analysis further revealed 293 biological process (BP) terms, 33 cellular component (CC) terms, and 56 molecular function (MF) terms ( p  < 0.05). The top 10 enriched GO categories, selected based on gene counts and statistical significance, are presented in Fig.  5 D.

This study enrolled women who sought infertility treatment at the First Hospital of Lanzhou University between April 2022 and July 2023. The study protocol was approved by the Ethics Committee of Lanzhou University First Hospital (Approval No. LDYYSZLLKH2022-03). All participants provided written informed consent before enrollment. Participants were included based on the Rotterdam diagnostic criteria for PCOS. PCOS was diagnosed when at least two of the following three criteria were present: (1) oligo-ovulation or anovulation, clinically manifested as menstrual irregularity; (2) clinical and/or biochemical hyperandrogenism, including hirsutism, severe acne, androgenic alopecia, or higher serum androgen levels; and (3) polycystic ovarian morphology (PCOM) detected by ultrasound. The exclusion criteria were as follows: (1) application of preimplantation genetic testing (PGT) due to heritable pathogenic variants in either female or male partners; (2) uterine abnormalities, i.e., congenital malformations, intrauterine adhesions, endometrial lesions, uterine leiomyomas, adenomyosis, or cervical insufficiency; (3) endometriosis, ovarian tumors (benign or malignant), or hydrosalpinx; (4) systemic conditions such as thyroid dysfunction, hyperprolactinemia, immune-allergic disorders, hypertension, cardiovascular disease, diabetes mellitus, or tuberculosis; (5) abnormal semen parameters, including moderate to severe oligospermia, azoospermia, or necrospermia; and (6) endocrine disorders (congenital adrenal hyperplasia, Cushing’s syndrome, or androgen-secreting tumors). A detailed patient selection flowchart is presented in Fig.  1 . Fig. 1 Flowchart of data exclusion and inclusion for the study Flowchart of data exclusion and inclusion for the study Based on serum Hcy levels, patients were stratified into three groups (low, medium, and high) using tertile distribution. All patients were managed at the Reproductive Medicine Center. Demographic data, including age and body mass index (BMI), were recorded, and medical histories were documented in the electronic medical record system. At enrollment, venous blood samples were collected for the assessment of complete blood count, anti-Müllerian hormone (AMH), homocysteine (Hcy), cancer antigen 125 (CA-125), and thyroid-stimulating hormone (TSH), with all analyses performed in the institutional laboratory. Transvaginal ultrasonography was conducted on menstrual cycle days 2–3 during the early follicular phase to evaluate uterine and ovarian morphology, including antral follicle count (AFC) and pelvic structures. Oocyte retrieval procedures were performed by three senior physicians. Fertilization (IVF/ICSI), embryo culture, and morphological analysis were performed by two experienced embryologists in accordance with standardized protocols. The initial gonadotropin (Gn) dose was determined based on age, AMH, BMI, and AFC. In the long GnRH agonist protocol, 3.75 mg of agonist was administered intramuscularly during the mid-luteal phase of the preceding cycle. Gonadotropin stimulation was initiated after confirmation of pituitary downregulation (endometrial thickness ≤ 5 mm, serum estradiol ≤ 50 pg/mL, and follicular diameter ≈ 5 mm). Ovulation triggering was performed when at least one follicle reached ≥ 18 mm in diameter, with hormone levels guiding clinical decisions. The trigger medication type, dosage, and timing were selected based on the specific stimulation protocol, follicle number, and serum hormone levels. Before oocyte retrieval, the vagina was irrigated 2–3 times. Transvaginal ultrasound-guided follicular aspiration was performed 36 h post-trigger administration. All embryos were cultured in G-1™ PLUS medium (Vitrolife, Sweden) within a tri-gas incubator maintained at 37.0 ± 0.5 °C with 6% CO₂ and 5% O₂ in a humidified atmosphere. Pronuclear (PN) stage evaluation was performed 16–18 h post-fertilization using phase-contrast microscopy by experienced embryologists. Normal fertilization was defined by the presence of two distinct pronuclei (2PN) in the cytoplasm. Cleavage-stage embryos were graded according to the Istanbul consensus criteria. Grades I and II were classified as high-quality embryos: Grade I: Blastomeres of equal size with regular morphology, clear homogeneous cytoplasm, and < 10% cytoplasmic fragmentation, Grade II: Blastomeres of slightly unequal size with 10–20% cytoplasmic fragmentation, Grade III: Blastomeres of unequal size with 21–50% cytoplasmic fragmentation, potentially with cytoplasmic vacuolation, Grade IV: Severely irregular blastomeres with > 50% fragmentation, frequently showing prominent cytoplasmic vacuoles or granulation. In this study, embryo transfer was performed using Grade I, Grade II, and/or Grade III embryos according to patient-specific clinical considerations. The following reproductive outcome parameters were calculated: MII oocyte rate = (number of metaphase II oocytes / total number of retrieved oocytes) × 100%; 2PN fertilization rate = (number of 2PN zygotes / total number of inseminated or injected oocytes) × 100%; High-quality embryo rate = (number of Grade I + Grade II embryos/number of 2PN zygotes) × 100%; Clinical pregnancy rate = (number of clinical pregnancy cycles/number of embryo transfer cycles) × 100% Clinical pregnancy was defined as ultrasonographic visualization of at least one intrauterine gestational sac with or without fetal cardiac activity, excluding ectopic and biochemical pregnancies diagnosed solely by serum β-hCG elevation without sonographic confirmation. Cell viability was observed using the CCK-8 assay (Coolaber, Beijing, China). Experimental groups included a blank control (medium only), a negative control (KGN cells without Hcy), and treatment groups exposed to Hcy at increasing concentrations (5, 10, 20, 40, 80, and 160 µM). Briefly, KGN cells (5 × 10³ cells/well) were seeded in 96-well plates and incubated for 24 h. The medium was replaced with Hcy-containing medium, followed by another 24-h incubation. Thereafter, 10 µL of CCK-8 reagent and 90 µL of medium were added to each well, and plates were incubated for 3 h at 37 °C. Absorbance was measured at 450 nm using a microplate reader (TECAN, Switzerland). Cell viability was calculated as: Cell viability (%) = [(OD_treatment - OD_blank) / (OD_control - OD_blank)] × 100. All experiments were independently performed in triplicate. Apoptosis was examined using an Annexin V-FITC/PI apoptosis detection kit (MultiSciences, Hangzhou, China). Annexin V (or Annexin A5) is a member of the intracellular annexin protein family. The assay principle is based on the high-affinity, calcium-dependent binding of Annexin V to phosphatidylserine (PS). In viable cells, PS is asymmetrically distributed to the inner leaflet of the plasma membrane; however, during early apoptosis, membrane phospholipid asymmetry is disrupted, resulting in PS externalization to the outer leaflet of the plasma membrane. Fluorescently conjugated Annexin V specifically binds to externalized PS, thus identifying apoptotic cells. For this assay, KGN cells were treated with Hcy at concentrations of 0, 20, 40, and 80 µM. Cells (1 × 10⁶ − 3 × 10⁶) were collected by centrifugation, washed twice with pre-cooled PBS, and resuspended in apoptosis-positive control solution (500 µL), then incubated on ice for 30 min. After washing with ice-cold PBS, cell pellets were resuspended in an appropriate volume of pre-chilled 1X Binding Buffer. An equivalent number of untreated, viable cells was added as a negative control. The cell suspension volume was adjusted to 1.5 mL with 1X Binding Buffer and divided equally into three aliquots: (1) unstained blank control, (2) Annexin V-FITC single-stain control, and (3) PI single-stain control. For single-stain controls, 5 µL of Annexin V-FITC or 10 µL of PI was added, respectively, followed by incubation at room temperature (25 °C) in the dark for 5 min. Flow cytometry was performed using a flow cytometer (Beckman Coulter, Cytoflex). Instrument voltage settings for forward scatter (FSC), side scatter (SSC), and fluorescence channels were optimized using the blank control. Fluorescence compensation was calibrated using the single-stain control samples. Annexin V-FITC fluorescence was detected through the FITC channel (excitation λ = 488 nm; emission λ = 530 nm), and PI fluorescence was detected through the PE/PI channel (excitation λ = 535 nm; emission λ = 615 nm. For microscopic visualization, stained cells were transferred onto glass slides and coverslipped. Before observation, cells could be fixed with 2% formaldehyde; however, Annexin V-FITC staining must be performed before fixation to preserve membrane integrity. Fluorescence imaging was performed using an inverted fluorescence microscope equipped with FITC and rhodamine filter sets. Apoptotic cells showed a characteristic green fluorescent ring at the plasma membrane (Annexin V binding), while cells with compromised membrane integrity showed red nuclear fluorescence (PI staining) in addition to peripheral green fluorescence. All experiments were independently repeated at least three times. To predict potential molecular targets of Hcy, a multi-database approach was used, drawing on four publicly accessible online target prediction platforms: SwissTargetPrediction ( http://www.swisstargetprediction.ch/ ) [ 23 ], TargetNet ( http://targetnet.scbdd.com/home/index/ ) [ 24 ], BATMAN-TCM ( http://bionet.ncpsb.org/batman-tcm/ ) [ 25 ], and PharmMapper ( http://www.lilab-ecust.cn/pharmmapper ) [ 26 ]. Target prediction searches were restricted to Homo sapiens . Retrieved gene identifiers and protein names were standardized and validated using the UniProt database ( http://www.uniprot.org/ ) [ 27 ]. Bioinformatics analyses were conducted using publicly available, quality-controlled databases derived from validated scientific datasets. Molecular targets associated with PCOS were retrieved from four complementary disease-gene databases: OMIM ( http://www.omim.org ) [ 28 ], GeneCards ( https://www.genecards.org/ ) [ 29 ], DrugBank ( https://go.drugbank.com/ .) [ 30 ], and DisGeNET databases ( https://www.disgenet.org/ ) [ 31 ], using “polycystic ovary syndrome” as the primary search query. For the GeneCards database, which assigns relevance scores that reflect the strength of gene-disease associations, only targets with a relevance score exceeding the median were retained to enhance specificity and reduce false-positive associations [ 32 ]. Target lists retrieved from all four databases were consolidated, and duplicate entries were removed to generate a comprehensive, non-redundant compendium of PCOS-associated molecular targets. The bioinformatics analyses conducted in this study are based on standard, publicly available genomic or proteomic databases. The data in these databases are derived from scientific projects that have undergone rigorous quality control and ethical review, serving as recognized, authoritative resources in the field, rather than on unverified “personal sequencing datasets.” Venn diagram analysis was performed to identify overlapping molecular targets between Hcy and PCOS using the online platform at ( https://bioinfogp.cnb.csic.es/tools/venny/ ). These intersecting targets were used to construct a PPI network from the STRING database (V11.5), accessible at https://string-db.org/ [ 31 ]. The species parameter was set to Homo sapiens , and the minimum required interaction score was defined as high confidence (> 0.7). The PPI and component-target gene–disease (CTD) networks were visualized using Cytoscape v3.9.1. The CentiScaPe plugin was used to calculate degree centrality (DC), closeness centrality (CC), and betweenness centrality (BC). Nodes showing DC, CC, and BC values above their respective mean thresholds were identified as core genes. Statistical analysis was conducted using SPSS v26.0 and Empower Stats. Normally distributed data were expressed as mean ± standard deviation. Independent sample t -tests or Mann-Whitney U tests were used as appropriate. Categorical variables were analyzed using chi-square tests. Hcy levels were categorized into tertiles, and intergroup comparisons were performed using analysis of variance (ANOVA) for continuous variables and chi-square tests for categorical variables. Logistic regression analysis was conducted to evaluate the association between Hcy levels and clinical outcomes, adjusting for potential confounders. Curve-fitting analysis was performed to assess the relationship between serum Hcy levels and high-quality embryo rates.

Variations in Hcy levels were not associated with differences in clinical pregnancy outcomes across the study groups. However, higher Hcy levels were negatively correlated with the cleavage-stage high-quality embryo rate, with each 1 µmol/L increase associated with an approximate 1.4% decline. The results further indicated that Hcy exposure reduced KGN granulosa cell viability and promoted apoptosis. Network pharmacology analysis provided preliminary evidence supporting a potential mechanistic association between Hcy and PCOS.

In this study, across the analyzed cohorts, there were no statistically significant differences in clinical pregnancy, miscarriage, or live birth rates. The role of Hcy in determining IVF/ICSI-ET outcomes remains controversial, with divergent findings reported in the literature. Evidence presented by Nafiye et al. aligns with the current observations, suggesting that although serum Hcy level may influence embryo number and quality, they do not significantly affect clinical pregnancy rates [ 33 ]. Similarly, Liu et al. reported no association between Hcy levels and oocyte maturation, embryo quality, or pregnancy outcomes [ 34 ]. In comparison, Ocal et al. showed a significant relationship between follicular fluid Hcy level and pregnancy outcomes, despite no correlation with oocyte quality. Lower Hcy levels were observed in the pregnancy group (9.6 ± 2.02 µmol/L) than in the non-pregnancy group (14.9 ± 2.93 µmol/L; p < 0.001), indicating its potential as a predictor of pregnancy failure [ 35 ]. Consistent with this, Berker et al. identified a significant correlation between follicular fluid Hcy levels and clinical pregnancy rates [ 36 ], while Boyama et al. further supported these findings by evaluating Hcy levels in embryo culture waste media [ 37 ]. Akamine et al. reported that follicular fluid Hcy levels below 4.9 µmol/mL were positively associated with improved pregnancy outcomes [ 38 ]. Previous studies have shown that, in patients with PCOS, endocrine abnormalities such as higher luteinizing hormone levels and hyperandrogenism after ovulation induction are frequently associated with implantation failure and increased miscarriage rates, which may be partially due to higher Hcy levels [ 39 ]. Elevated Hcy may disrupt endometrial receptivity by inducing oxidative stress in vascular endothelial cells, promoting platelet activation, impairing blood flow, stimulating vascular smooth muscle proliferation, and activating the UPR, ultimately leading to endothelial cell apoptosis. These changes may adversely affect embryo implantation and increase the risk of pregnancy loss [ 39 ]. At the molecular level, RNA sequencing of peripheral blood mononuclear cells from PCOS patients has identified 186 differentially expressed genes (69 upregulated and 117 downregulated genes). Among these, AQP9 , PROK2 , S100A12 , and TGM3 showed significant diagnostic potential for distinguishing PCOS patients from healthy controls [ 40 ]. Furthermore, a systematic review and meta-analysis found that polymorphisms in the vitamin D receptor (VDR) gene, including ApaI , BsmI , Cdx2 , and TaqI , are associated with an increased risk of PCOS [ 41 ]. Higher Hcy levels have also been linked to adverse obstetric complications, including hypertensive disorders of pregnancy, intrauterine growth restriction, placental abruption, and preterm delivery, primarily through vascular dysfunction, abnormal uterine contractility, and oxidative stress [ 39 , 42 , 43 ]. Moreover, the interaction between Hcy and insulin resistance increases the risk of gestational diabetes mellitus [ 44 ]. Folate metabolism is a key regulatory pathway for Hcy homeostasis. Genetic variants in the MTHFR gene reduce enzymatic activity, impair folate metabolism, and elevate Hcy levels, increasing the risk of fetal neural tube defects and adverse obstetric outcomes. Due to the retrospective design of this study, prospective control of confounding variables (dietary intake and nutritional status) was not feasible, potentially influencing the results. To minimize potential bias, all clinical data, including Hcy measurements, were obtained from a standardized electronic medical record system, and both PCOS diagnostic criteria and IVF protocols were uniformly used. However, serum folate levels, vitamin B12 levels, and MTHFR gene polymorphisms were not examined or adjusted for, representing a limitation that should be addressed in future studies. The lack of association between Hcy levels and pregnancy outcomes observed in this study may be explained by several factors. First, no significant differences were observed among groups in endometrial thickness on the day of embryo transfer, nor in the number and quality of transferred embryos. Second, the timing of sample inclusion resulted in incomplete follow-up for some patients, potentially introducing bias and limiting the strength of conclusions about offspring outcomes. Despite these findings, managing Hcy levels remains clinically important. Each 1 µmol/L increase in Hcy has been associated with a 1.4% reduction in the rate of high-quality embryos. Although this effect may be modest in a single treatment cycle, it may be clinically relevant in specific populations, such as patients with poor prognosis or recurrent implantation failure, particularly when cumulative outcomes are considered. Embryo morphology is an early developmental indicator and does not necessarily correlate directly with clinical pregnancy or live birth, both of which are influenced by multiple downstream factors, including endometrial receptivity and immune regulation. Therefore, the potential adverse effects of elevated Hcy on early embryonic development may be attenuated or masked during later stages of implantation and gestation. The results primarily indicate an association between Hcy levels and early embryonic developmental competence, whereas the impact of Hcy on final pregnancy outcomes remains inconclusive and needs further validation through large-scale prospective studies. The Hcy level range used in the in vitro experiments is in good agreement with a previous study investigating its effects on ovarian granulosa cell function, allowing for the observation of significant biological responses within a controlled timeframe [ 45 ]. The use of the KGN granulosa cell line is a limitation of this study. A comprehensive understanding of Hcy’s role in reproductive outcomes requires further investigation using endometrial cell models. Future studies should incorporate endometrial epithelial and stromal cells to directly evaluate the effects of Hcy on key markers of endometrial receptivity, including integrin αvβ3, leukemia inhibitory factor (LIF), and HOXA10, as well as its impact on cell adhesion and decidualization. Such approaches would provide direct mechanistic evidence linking Hcy dysregulation to impaired embryo implantation.

Homocysteine (Hcy), a non-protein, sulfhydryl-containing amino acid involved in one-carbon metabolism and generated during methionine demethylation, has cytotoxic properties and contributes to multisystem tissue damage [ 1 ]. Higher Hcy levels may arise from aging, genetic factors, deficiencies of folate and B vitamins, excessive methionine intake, hypothyroidism, chronic alcohol consumption, renal impairment, and malignancies (breast, ovarian, and pancreatic cancers) [ 2 ]. Hyperhomocysteinemia (HHcy) is defined by increased plasma Hcy levels, with normal levels ranging from 5 to 15 µmol/L. Based on severity, HHcy is categorized as mild (15–30 µmol/L), moderate (30–100 µmol/L), and severe (> 100 µmol/L) [ 3 ]. In female reproduction, HHcy induces oxidative stress, promotes platelet activation, disrupts hemodynamics, and stimulates vascular smooth muscle cell proliferation. It activates the unfolded protein response (UPR), leading to endothelial apoptosis through excessive generation of reactive oxygen species (ROS). These changes impair endometrial perfusion and vascular integrity, contributing to obstetric complications, i.e., recurrent miscarriage, gestational diabetes mellitus, hypertensive disorders of pregnancy, intrauterine growth restriction, placental abruption, and preterm birth, ultimately resulting in adverse pregnancy outcomes [ 4 , 5 ]. Higher Hcy levels are also associated with reduced fertilization rates, impaired embryo quality, decreased implantation and clinical pregnancy rates, and increased risk of miscarriage [ 6 ]. HHcy disrupts folliculogenesis, oocyte competence, embryo development, and endometrial receptivity, thus increasing the possibility of implantation failure and pregnancy-related complications. In assisted reproductive technologies (ART), HHcy and variants of 5,10-methylenetetrahydrofolate reductase (MTHFR) are associated with reduced oocyte yield and compromised embryo quality; however, folate and B vitamin supplementation may partially ameliorate these effects [ 7 ]. Proteomic analyses have shown that HHcy induces mitochondrial dysfunction and downregulates zona pellucida proteins (ZP1, ZP2, and ZP3) in oocytes. Ultrastructural evaluation by transmission electron microscopy (TEM) has revealed aberrant zona pellucida formation and microvillar architecture in oocytes from HHcy models. Moreover, in vitro fertilization studies have shown reduced rates of two-cell embryo formation under HHcy conditions, indicating impaired early embryonic development and a diminished reproductive lifespan due to compromised follicular development and oocyte quality [ 8 ]. Polycystic ovary syndrome (PCOS) is a multifactorial disease that affects the endocrine, reproductive, metabolic, and cardiovascular systems. It is clinically characterized by hyperandrogenism, ovulatory dysfunction (anovulation or oligo-ovulation), infertility, and recurrent pregnancy loss [ 9 ]. Based on the Rotterdam criteria established in 2003 by the European Society of Human Reproduction and Embryology and the American Society for Reproductive Medicine, the global prevalence of PCOS ranges from 6 to 21%, with reported rates of 5.6–11.2% in China [ 10 , 11 ]. PCOS is a major cause of female infertility, induced by ovulatory dysfunction, impaired oocyte development, dysregulated gonadotropin secretion, hyperandrogenism, and abnormalities in ovarian growth factors and their binding proteins [ 12 ]. Clinical evidence indicates that patients with PCOS have considerably higher Hcy levels than those without PCOS [ 13 ]. Moreover, PCOS is strongly associated with metabolic disturbances, including dysregulated glucose and lipid metabolism, which increase the risk of cardiovascular disease [ 14 ]. HHcy has been identified as an independent risk factor for cardiovascular disease and is involved in the long-term cardiovascular complications of PCOS. It is also correlated with key clinical features of PCOS, including obesity, dyslipidemia, insulin resistance, hyperandrogenism, and ovulatory dysfunction [ 15 ]. A well-regulated microenvironment is essential for optimal oocyte maturation and embryo development [ 16 ]. Follicular fluid and ovarian granulosa cells provide oocytes with critical nutrients, hormones, and enzymatic support, thus establishing the preovulatory niche required for follicular growth and ovulation [ 17 ]. Changes in follicular fluid composition can significantly affect oocyte maturation and embryogenesis [ 18 ]. The KGN cell line, a widely used human granulosa cell model in reproductive biology, retains responsiveness to follicle-stimulating hormone (FSH) and steroidogenic activity, and is extensively used to investigate granulosa cell function and ovarian pathophysiology [ 19 – 22 ]. Current research has predominantly focused on the association between Hcy and PCOS; however, the underlying molecular targets linking these factors remain poorly characterized, particularly using integrative approaches such as network pharmacology. Moreover, the direct effects of Hcy on ovarian granulosa cell function have not been comprehensively elucidated. Therefore, this study evaluates the effects of different serum Hcy levels on early embryonic development in patients with PCOS undergoing ART. In vitro experiments are conducted to investigate the effects of varying Hcy levels on the viability and autophagic activity of KGN cells, identifying potential molecular targets linking Hcy and PCOS.