Fig 2: Downregulation of PTN in pathological pregnancy conditions, including RIF and RPL. (A) Uniform Manifold Approximation and Projection (UMAP) showing expression of PTN in RIF single-cell data from GSE183837. (B) Expression of PTN in different cell clusters in violin plots. (C) Expression levels of PTN in hESCs isolated from endometrium of RIF and RPL patients via Western blotting, which followed densitometric analysis of Western blotting band (n = 3). (D, E) Expression of PTN and Vimentin in RIF (D) and RPL (E) clinical samples by immunofluorescence, and quantitative analysis of immunofluorescence was performed (n = 6). Data are presented as mean ± SEM and analyzed using t-test (*p < 0.05, **p < 0.01, ****p < 0.0001). RIF, recurrent implantation failure; RPL, recurrent pregnancy loss; hESCs, human endometrial stromal cells.

Fig 3: PTN regulates endometrial stromal cell decidualization through the IGF-2 signaling axis. (A) PPI network of PTN, IGF-2, IGFBP1, LIF, PRL, and other decidualization-related genes. (B) The expression levels of PTN and its interacting partners (IGF-2, WNT4, PRL, IGFBP1, and LIF) in the RIF (left) and RPL (right) datasets. (C) Expressions of decidualization-related genes (including IGFBP1, LIF, PRL, and WNT4) in NC or siPTN hESCs after treatment with control vehicle, cAMP, and IGF-2 via RT-qPCR (n = 9). (D) Immunoblotting for PTN, IGFBP1, and PRL expression levels with control vehicle, cAMP, and IGF-2. (E) ELISA for IGF-2 levels with control vehicle, cAMP, and IGF-2. (F) Schematic diagram of the experiments performed using the different treatments of hESCs. Data are presented as mean ± SEM and analyzed using t-test or one-way ANOVA test. * compared with NC treated with control vehicle, # compared with siPTN treated with control vehicle, $ compared with siPTN treated with control cAMP (NS, no significant difference; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ####p < 0.0001, $p < 0.05, $$p < 0.01, $$$p < 0.001, $$$$p< 0.0001). PPI, protein–protein interaction; RIF, recurrent implantation failure; RPL, recurrent pregnancy loss; hESCs, human endometrial stromal cells; cAMP, cyclic adenosine monophosphate.

Fig 4: Identification of PTN as a shared hub gene with high diagnostic performance in RIF and RPL. (A) PPI network constructed with the co-expressed DEGs in both RIF and RPL. MCODE was used to identify the significant module from the PPI network with a score of ≥5.0. Different node colors represent different functions. (B, C) The mean decrease Gini score rank of the top 20 hub genes in RIF (B) and RPL (C) was calculated using Random Forest analysis. (D) Venn diagram showing the common hub genes calculated using Random Forest analysis of RIF and RPL, and Cytoscape analysis (Supplementary Figure 2). (E) ROC analysis confirmed the diagnostic efficacy of the top four biomarkers (the common hub genes identified between RIF and Hubba Gene) for RIF in training datasets. (F) ROC analysis confirmed the diagnostic efficacy of the four biomarkers (the common hub genes identified between RPL and Hubba Gene) for RPL in training datasets. (G) ROC analysis of the diagnostic efficacy of PTN for RIF (test data: GSE58144) and RPL (test data: GSE71835) in validation datasets. RIF, recurrent implantation failure; RPL, recurrent pregnancy loss; PPI, protein–protein interaction; DEGs, differentially expressed genes; ROC, receiver operating characteristic.

Fig 5: PTN deficiency induces endometrial immune imbalance and leads to adverse pregnancy outcomes, which can be partially reversed by IGF-2. (A) PPI network of PTN, IGF-2, CXCR4, CD4, CD8, CD56, and other immunoregulators. (B, C) Flow cytometry analysis (B) and statistical quantification (C) of cell populations of NK cells and T cells, and the expression levels of CD16, CXCR4, and GZMB in NK cell form the endometrial tissues of NC and siPTN mice treated with control vehicle or IGF-2 (50 ng/mL) (n = 6). (D) The flowchart depicts the steps involved in establishing a mouse model of intrauterine perfusion with siRNA-mediated knockdown of Ptn and then supplemented with IGF-2 (50 ng/mL) through tail vein injection for 2 days; then, these female mice were mated with fertile male mice; analysis of fertility, including pregnancy rate, IF, and fluorescence-activated cell sorting (FACS). were performed at gestational day 13.5. (E) Representative images of uteri from pregnant mice in the NC and siPTN mice treated with control vehicle or IGF-2 (50 ng/mL) at gestational day 13.5 (n = 6) (arrow shows the absorption site). (F) Pregnancy rate (%) in NC and siPTN mice treated with control vehicle or IGF-2 (50 ng/mL) (n = 15). (G) Quantification of embryo numbers (left) and absorption rate (right) per mouse in NC (n = 13) and siPTN mice treated with control vehicle (n = 6) or IGF-2 (50 ng/mL) (n = 11) at gestational day 13.5. Data are presented as mean ± SEM and analyzed using one-way ANOVA test or χ2 test. * Compared with NC treated with control vehicle, # compared with siPTN treated with control vehicle (NS, no significant difference; *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001, ## p < 0.01, ### p < 0.001, #### p < 0.0001). IF, immunofluorescence.