Recurrent pregnancy loss (RPL)is a distressing pregnancy disorder experienced by ~2.5% of women trying to conceive (1). Defined as the failure of two or more clinically recognized pregnancies before 20–24 weeks of gestation, and includes embryonic and fetal losses (1). The etiology of RPL is multifactorial, involving genetic anomalies, endocrine disorders, anatomical abnormalities, infectious diseases, thrombophilic disorders, and immune dysregulation. However, nearly 50% of cases remain unexplained even after extensive clinical evaluation (2). Unexplained RPL (uRPL) presents a major clinical challenge, as the lack of identifiable causes hampers the development of targeted treatments. Emerging evidence emphasizes the crucial role of dysregulated maternal-fetal immune interactions and abnormal inflammatory responses in pregnancy maintenance and loss (3). Among the immune mechanisms implicated, neutrophil extracellular traps (NETs)—web-like structures composed of DNA, histones, and antimicrobial proteins released by activated neutrophils—have recently emerged as key players in both physiological defense and pathological inflammation (4, 5). While NETs are essential for combating infections, their excessive or dysregulated formation, termed NETosis, has been linked to autoimmune diseases, thrombotic disorders, and obstetric complications such as preeclampsia and preterm birth (6, 7). However, the contribution of NETs to uRPL, particularly their interplay with decidual immune cell dynamics and molecular pathways, remains poorly characterized. This study addresses this gap by integrating multi-omics and immune profiling to elucidate the role of NETs in uRPL pathogenesis.

Neutrophils, the most abundant leukocytes in human blood, rapidly respond to inflammatory stimuli by releasing NETs, which immobilize pathogens but also exacerbate tissue damage and inflammation (4). In pregnancy, neutrophils infiltrate the decidua and participate in immune tolerance and placental development (8). However, dysregulated NETosis has been implicated in adverse pregnancy outcomes. For instance, elevated NET biomarkers, such as myeloperoxidase-DNA (MPO-DNA) complexes, Neutrophil Elastase (NE), and citrullinated histone H3 (citH3), are observed in preeclampsia and spontaneous preterm birth, correlating with placental inflammation and vascular dysfunction (6, 9, 10). NETs may impair trophoblast invasion, activate complement pathways, and promote thromboinflammatory cascades, all of which could disrupt pregnancy (11). Despite these advances, the role of NETs in uRPL—a condition characterized by recurrent, often idiopathic losses—remains underexplored.

RNA sequencing (RNA-seq) has emerged as a pivotal tool for investigating disease pathogenesis due to its high sensitivity in resolving complex transcriptional regulatory networks (12). Methodological advancements in RNA-seq have significantly enhanced the accuracy of biomarker discovery, while the translational value of transcriptome analysis in therapeutic development is increasingly recognized (13, 14), This technology enables the detection of subtle, biologically critical transcriptomic alterations underlying complex diseases (15, 16).

Here, we investigated the contribution of aberrant NETs formation to the pathogenesis of uRPL by disrupting maternal-fetal immune homeostasis. Utilizing a multidimensional strategy integrating histopathological, transcriptomic, bioinformatic, and clinical data, we delineated the role of NETs in uRPL.

uRPL poses significant clinical and psychological challenges, underscoring the urgent need to elucidate its underlying mechanisms. This study integrated clinical, histopathological, transcriptomic, and bioinformatic approaches to systematically explore the association between NETs and uRPL and their potential mechanisms. We first observed elevated serum NETs markers in uRPL patients, indicating aberrant neutrophil activation and NETs release, consistent with prior research linking NETs to pregnancy complication (10, 28). Histological analyses revealed increased decidual neutrophil infiltration, supported by immunohistochemistry, implicating neutrophil hyperactivation and NETs formation in decidual microenvironment imbalance. Previous studies have shown that NETs play a significant role in autoimmune diseases and thrombosis (29), and excessive release of NETs may disrupt maternal-fetal immune tolerance, leading to embryo loss (5). NETs-derived cfDNA exacerbates inflammation by inducing tumor necrosis factor-α (TNF-α) mRNA, while histones trigger apoptosis and act as damage-associated molecular patterns (DAMPs), promoting proinflammatory cytokine release, cytotoxicity, and ROS-mediated endothelial dysfunction, impairing embryo implantation and placental development, ultimately contributing to adverse pregnancy outcomes (7, 30).

Subsequent transcriptomic analysis of decidual tissues revealed significant upregulation of DE-NRGs. GO, KEGG, and GSVA enrichment analyses demonstrated that DE-NRGs were primarily involved in neutrophil chemotaxis, migration, activation pathways, degranulation regulation, Neutrophil extracellular trap formation, complement and coagulation cascades, and TNF signaling pathways, collectively contributing to RPL pathogenesis ( Figure 3 ). Machine learning algorithms identified four hub DE-NRGs: C3AR1, ITGAM, ITGB2, and LYZ ( Figure 4 ). These genes encode proteins involved in complement activation, leukocyte adhesion, and microbial defense (31–33), suggesting their potential roles in amplifying NETs-driven inflammation and immune dysregulation. Western blot confirmed elevated protein expression of hub DE-NRGs in uRPL decidual tissues, solidifying their association with disease pathology ( Table 2 ).

C3AR1, a complement receptor central to the complement system, regulates C3a signaling to promote neutrophil chemotaxis and phagocytosis, influencing NETs formation (34, 35). Recent research links reduced placental C3AR1 expression in preeclampsia to maternal blood pressure, implicating the janus kinase–signal transducer and activator of transcription (JAK-STAT), transforming growth factor-β (TGF-β), and hypoxia-inducible factor-1 (HIF-1) pathways, as well as NK cell/M1 macrophage activity (36). However, our findings reveal upregulated C3AR1 in uRPL decidua, warranting further mechanistic exploration.

ITGAM and ITGB2 encode integrin αM (CD11b) and β2 (CD18) subunits, forming the CD11b/CD18 heterodimer (also known as aMb2, Mac-1, and CR3), which mediates myeloid cell adhesion, migration, and phagocytosis by binding ligands like ICAM-1 and fibrinogen, contributing to inflammation, immune defense, and thrombosis (37, 38). Dysregulation may promote proinflammatory immune responses, disrupting early pregnancy immune homeostasis and triggering adverse outcomes (39). Turunen et al. reported elevated CD11b in neutrophils/monocytes of infants born to early-onset preeclampsia, correlating with postpartum systemic inflammation (40). Yun-Long Zhang et al. demonstrated CD11b/CD18-mediated monocyte adhesion and macrophage polarization in cardiac remodeling (41). Our study further revealed positive correlations of ITGAM and ITGB2 with Macrophages M2 and negative correlations with monocytes in decidual tissues, but their mechanistic impact on endometrial decidualization remains unclear.

LYZ, a key antimicrobial enzyme of the innate immune system, plays vital roles in embryonic and neonatal immune defense (33). Natalia et al. reported upregulated serum LYZ levels in early-onset preeclampsia (42). In this study, marked LYZ elevation was observed in RPL decidual tissues, positively correlated with neutrophil infiltration. While no direct evidence links LYZ to pregnancy loss, these findings suggest LYZ-mediated decidual inflammation may contribute to uRPL pathogenesis.

Notably, ROC curves of the four hub DE-NRGs exhibited AUC values >0.85 ( Figures 5A-D ), indicating strong diagnostic potential for uRPL. Utilizing these markers, we constructed a nomogram model ( Figure 5E ), with the calibration curve demonstrating excellent agreement between predicted and observed probabilities (mean absolute error = 0.003). Decision curve and clinical impact curve analyses further validated its clinical utility ( Figures 5G, H ). Future efforts will initially expand the cohort at our center (n=100 uRPL patients), subsequently validate the model in multi-center cohorts, and ultimately integrate it with established RPL diagnostic indicators (e.g., maternal age) to construct a composite scoring system. This will be followed by developing a cloud-based platform for automated NETs-related risk assessment.

A key advancement of this study lies in integrating immune cell infiltration analysis with NETs-related molecular networks. Using CIBERSORT, we identified altered infiltration patterns of six immune subsets in uRPL: increased CD8+ T cells, M2 macrophages, and neutrophils, alongside decreased CD4+ activated memory T cells, follicular helper T cells, and monocytes. These findings contrast with prior reports. For instance, Changqiang Wei et al. observed reduced M1/M2 macrophage infiltration in RPL via GEO data analysis (43), while Yujia Luo et al. reported elevated monocytes and reduced T cells in RPL (44). Hui Hu et al. identified differences in eosinophils, monocytes, NK cells, and Tregs, likely reflecting cohort heterogeneity (45). The results of this study show that neutrophil infiltration in the decidual tissue of uRPL patients is increased, which is consistent with the HE staining results of the decidual tissue, and further validates the abnormal NETs formation in uRPL.

Macrophages, as key regulators of both innate and adaptive immune responses, can polarize into pro-inflammatory M1 macrophages or anti-inflammatory M2 macrophages under different microenvironmental conditions (46). M1 macrophages can be activated by lipopolysaccharide (LPS) and interferon-γ (IFN-γ), secreting inflammatory factors like TNF-α, interleukin (IL)-1β, and IL-6, to kill invading pathogens, perform phagocytosis, and clear aged or damaged cells (47, 48). M2 macrophages, activated by IL-4 and IL-13, secrete anti-inflammatory cytokines such as IL-10, IL-4, and transforming growth factor-β (TGF-β), inhibit T cell proliferation and activation, and participate in Th2 immune responses, aiding tissue repair and angiogenesis (48, 49). Therefore, the M1/M2 balance is crucial for tissue homeostasis. Our findings revealed increased M2 macrophage infiltration in uRPL, suggesting unresolved inflammation potentially exacerbated by aberrant NETs formation.

Accumulating evidence indicates that NETs promote inflammatory responses by interacting with various cells in the immune system. Exposure of macrophages to NETs triggers the activation of the NLRP3 inflammasome, facilitating the release of IL-1β and IL-18 (50). NETs also directly induce the secretion of other pro-inflammatory cytokines, including IL-8, IL-6, and TNF-α (51, 52). Research by Wang Y et al. demonstrated that neutrophils can promote the production of inflammatory cytokines in macrophages by triggering p65 nuclear translocation via activation of the TGF-β1/Smad signaling pathway (53). Integrative multi-omics analysis by Zhao YF et al. revealed that macrophages can coordinate NETs generation through the CXCL3/CXCR2 axis, thereby contributing to the progression of related diseases (54). Furthermore, studies by Kuang L et al. showed that macrophages, upon apoptosis, can transfer mitochondria to neutrophils via microvesicles; this process induces mitochondrial dysfunction and triggers NETs formation through the mitochondrial reactive oxygen species (mtROS)/Gasdermin D (GSDMD) axis (55). Collectively, these findings suggest a close relationship and robust crosstalk between NETs and macrophages.

Additionally, four hub DE-NRGs are positively correlated with M2 macrophages and negatively correlated with T cells, follicular helper and monocytes, while LYZ shows a unique positive correlation with neutrophils. We hypothesize that NETs-related genes regulate immune cell interactions, leading to macrophage polarization to the M2 phenotype, suppressing adaptive immunity, and negatively impacting inflammatory responses and immune tolerance during pregnancy, thereby increasing the risk of pregnancy loss.

Finally, this study has limitations. The modest cohort size restricted subgroup stratification; larger cohorts are needed to validate NETs dynamics across uRPL subtypes. Second, the limited transcriptomic samples (6 uRPL vs. 5 controls) may introduce bias despite using stringent thresholds (|logFC|>0.585, adjusted p<0.05) and validating key genes via WB. Third, we focused solely on transcriptional dysregulation of four NETs-related genes without exploring genetic variations or underlying mechanisms. Future investigations will address these limitations through a phased approach: We will first expand clinical cohorts to 50 uRPL and 50 controls, quantifying decidual DE-NRG expression (qPCR/WB/IHC) while locating NETs-gene co-expression via immunofluorescence, and correlating serum DE-NRG/NETs markers (MPO-DNA/citH3) with miscarriage history and gestational age. Subsequently, an in vitro decidual stromal cell-neutrophil coculture system will assess NETosis, apoptosis, and decidualization markers post-siRNA knockdown, followed by phosphoproteomic pathway screening. In vivo studies will administer NETosis inhibitors to classic RPL mice (CBA/J♀×DBA/2♂) to evaluate NETs dynamics and pregnancy outcomes. Finally, we will integrate spatial transcriptomics with single-cell epigenomics to map NETs microenvironment regulation while actively establishing a multicenter uRPL-specific cohort for nomogram validation and integrated genomic/methylation analyses to elucidate NETs’ mechanistic role in uRPL pathogenesis and clinical management.

This study demonstrated elevated NETs in serum and decidua of uRPL patients. Transcriptomic sequencing identified four hub DE-NRGs (C3AR1, ITGAM, ITGB2, LYZ), validated as protein-level markers. Decidual immune dysregulation was associated with these genes, characterized by altered neutrophil and macrophage infiltration. These findings reveal novel molecular mechanisms of NETs in uRPL pathogenesis, proposing NETs components and associated genes as potential biomarkers for early uRPL screening.