The human monocytic cell line THP-1 was purchased from ATCC. The cells were cultured in RPMI-1640 (Gibco, Thermo Fisher Scientific, United States) containing 10% heat-inactivated FBS (Gibco) and 0.055 mM 2-mercaptoethanol (Gibco) in a 5% CO₂ atmosphere at 37°C. THP-1 cells were induced to differentiate into Mϕs as previously described (Chanput et al., 2014b; Huang et al., 2019b) with modifications. Briefly, the cells were treated with 200 ng/ml phorbol 12-myristate 13-acetate (PMA, Sigma, United States) followed by 24 h of incubation in RPMI-1640 medium to obtain a macrophage-like M0 state. Then, the M0-Mϕs were treated with fresh medium supplemented with 20 ng/ml IL-4 and 20 ng/ml IL-13 (PeproTech, Inc. United States) to induce differentiation into the M2 phenotype. siRNA specifically against SEC5 (SEC5 siRNA: 5′-GGUCGGAAAGACAAGGCAGdTdT-3′) and negative control siRNA (NC: 5′-UUC​UCC​GAA​CGU​GUC​ACG​UTT-3′) were synthesized by Shanghai GenePharma Co., Ltd. Transfections were performed using Lipofectamine 2000 according to the manufacturer’s instructions (Invitrogen, Thermo Fisher Scientific). For generation of polyclonal cell lines stably overexpressing SEC5 or a negative control, a lentivirus constructed and packaged by Shanghai Genechem Co., Ltd. was used to infect THP-1 cells. Cells expressing the Ubi-MCS-3FLAG-CBh-gcGFP vector (V) or expressing the recombinant SEC5 gene (SEC5) were grown in medium supplemented with puromycin at 1 μg/ml for approximately 14 days to eliminate uninfected cells. Quantitative real-time PCR and western blotting analyses were employed to determine the SEC5 expression level. Then, the selected cells continued to be grown in medium supplemented with puromycin at 0.5 μg/ml.

Human decidual tissues from 12 RSA patients (RSA, 6–10 weeks of gestation) and 15 healthy pregnant women in the first trimester (control, 6–9 weeks of gestation) were collected at the Department of Gynecology and Obstetrics, the Second Hospital of Tianjin Medical University (Tianjin, China). These collected decidual tissues were immediately washed several times with sterile, glucose-free PBS solution until there were no obvious blood clots. Then, the tissues were immersed in ice-cold RPMI-1640 medium. Cells were harvested from the tissues or the tissues were fixed within 3 h. Current pregnancy losses of the RSA patients were objectively confirmed by transvaginal ultrasound examination. Patients with classical risk factors, including abnormal parental karyotypes, uterine anatomical abnormalities, infectious diseases, luteal phase defects, diabetes mellitus, thyroid dysfunction, and hyperprolactinemia, were excluded from this study. In parallel, control women who had no history of miscarriage and were undergoing legal, voluntary terminations of early pregnancy were enrolled and evaluated for classical risk factors for early pregnancy loss. The sample collection for this study was approved by the Medical Ethics Committees of The Second Hospital of Tianjin Medical University (KY 2017K002) and Shanghai Institute for Biomedical and Pharmaceutical Technologies (Ref # PJ 2018-17). All samples were collected after informed consent was obtained. No significant differences in the average age or gestational week at sampling were observed between the RSA patients and the control women (Supplementary Table S1).

DMϕs were isolated as previously described (Houser et al., 2011; Jiang et al., 2018; Zhen et al., 2021). Briefly, tissues were washed and crushed into small pieces with a tissue crusher (Gentle MACS Dissociator, Miltenyi Biotec, Germany). Pieces of decidual tissues were digested in 3 mg/ml collagenase Type IV (Gibco, United States), 100 IU/ml DNase I (Sigma–Aldrich, United States), 100 IU/ml penicillin and 100 IU/ml streptomycin (Gibco) at 37°C for 30 min. Subsequently, the decidual cells that were released were filtered through 100-, 200-, and 400-mesh sieves (Corning, United States). To exclude any remaining red blood cells, the filtered cells were incubated with red blood cell lysis buffer (BD Biosciences, United States). DMϕs were obtained with anti-CD14 antibodies conjugated to magnetic beads (Miltenyi Biotec). The purity of the (CD45⁺ CD14⁺) dMϕs, which was detected by flow cytometry, was approximately 90%.

Primary BMDMs of C57BL/6 mice were prepared as previously described (Jia et al., 2014). In brief, bone marrow cells were aseptically collected from female mice aged 6–8 weeks (purchased from SIPPR/BK Laboratory Animal Company, Shanghai, China). The mice were sacrificed, and the collection area was sterilized with 75% ethanol. Then, the cells were isolated by flushing the mouse leg bones with phosphate-buffered saline (PBS). The cells were incubated in red blood cell lysis buffer (Solarbio, Beijing, China) before centrifugation. The cells were then cultured for 7 days in DMEM containing 20% FBS, 0.055 mM 2-mercaptoethanol (Gibco), antibiotic-antimycotic (Gibco), and 30% conditioned medium from L929 cells expressing macrophage colony-stimulating factor (M-CSF). Nonadherent cells were removed.

Total RNA was extracted with TRIzol Reagent (Invitrogen, MA, United States) and then reverse transcribed into cDNA (TaKaRa Bio, Inc. Japan) according to the manufacturer’s instructions. The synthesized cDNA was amplified with specific primers (Supplementary Table S2) and SYBR Green (TaKaRa Bio, Inc.) using a LightCycler 480 II real-time fluorescence quantitative PCR system (Roche, Basel, Switzerland). Triplicate samples were examined for each condition. The relative mRNA expression level was calculated using the 2^−ΔΔCt method.

For antibody staining, THP-1 cells were treated as indicated, washed, and incubated with CD206-PE and CD11b-FITC antibodies (BioLegend, United States) on ice for 30 min. Flow cytometric analysis was performed on a BD LSRFortessa system (BD Biosciences, United States), and the data were analyzed with FlowJo version 7.6.1 software. Inflammatory cytokines in the lysate of mouse uterine tissues were measured with the LEGENDplex Mouse Inflammation Panel (BioLegend, United States) according to the manufacturer’s instructions, and the data were analyzed with Qognit software.

THP-1 cells were differentiated as mentioned above and placed on coverslips (Fisher Scientific). The cells were fixed using 4% PFA in DPBS, followed by blocking and permeabilization with 0.1% Igepal (Sigma–Aldrich, Inc.) in DPBS with 2% BSA (Amresco, OH, United States). Primary antibodies diluted in DPBS with 2% BSA (Supplementary Table S3) were applied overnight at 4°C. The cells were subsequently washed four times with DPBS before being incubated with the appropriate secondary antibodies (Invitrogen, Thermo Fisher Scientific) diluted in 2% BSA in DPBS. The coverslips were washed four times with DPBS before being mounted on slides using mounting solution (Thermo Fisher Scientific). Confocal images were acquired with a Nikon A1R confocal system. For the decidual tissue slides, immunofluorescence staining of formalin-fixed and paraffin-embedded 5 μm sections was performed. After deparaffinization and rehydration, heat-induced antigen retrieval was performed by microwaving in 0.1 M sodium citrate (pH 6.0), and then, the slides were allowed to cool to room temperature. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide for 10 min. The slides were blocked with 10% normal donkey serum in PBS-T (PBS +0.1% Triton X-100) for 1 h at room temperature and incubated with primary antibodies diluted in 5% BSA/PBS-T overnight at 4°C. After being washed twice with PBS-T for 10 min and then PBS for 10 min, the slides were incubated with Cy3-conjugated anti-rabbit (2.5 μg/ml) or Alexa Fluor 488-conjugated anti-mouse (2 μg/ml) secondary antibody at 37°C for 30 min. Next, the slides were washed, stained with Hoechst 33,342 (Invitrogen, Thermo Fisher Scientific), and mounted.

THP-1 cells in the M0 or M2 state of differentiation as mentioned above were lysed using 1 × DPBS containing 1% Triton X-100. Rabbit SEC5 or STAT6 antibody (10 µg) or an equivalent amount of rabbit immunoglobulin G (IgG) isotype control (Sigma; #18140) was conjugated with 30 μL of a 50% slurry of protein G agarose resin according to the manufacturer’s instructions (Yeasen; #36405ES08). Then, the antibody-conjugated agarose resin was incubated with protein lysates (1 mg of protein) with gentle agitation for 2 h. The beads were retained after three washes with PBS. Next, 1× loading buffer was added to dissociate immunoprecipitates, and western blotting was performed accordingly.

For preparation of whole-cell extracts, cells were washed with ice-cold PBS, incubated with RIPA lysis buffer (Sangon Biotech, Shanghai, China) containing 1 mM PMSF and protease inhibitor cocktail (Selleck, Shanghai, China) on ice, and then homogenized with an ultrasonic cell disruptor. The supernatant of the cell and tissue lysates was centrifuged at 12,000 × g for 15 min at 4°C. Cell nuclear protein was extracted using a Nucleoprotein Extraction Kit (Sangon Biotech) according to the manufacturer’s instructions. Approximately 50 μg of protein from each sample was subjected to SDS–PAGE, and the separated proteins were transferred to nitrocellulose membranes (Merck Millipore, Darmstadt, Germany). Blots were incubated with the appropriate primary antibodies diluted in TBST (containing 0.1% Tween 20 and 2% BSA) for 1 h at room temperature. Then, the blots were washed and incubated with appropriate secondary antibodies and detected using an Odyssey CLx Imaging System (LI-COR, Nebraska, United States).

After coculturing with THP1-derived Mϕs for 48 h, the alteration in morphology of HTR-8/SVneo cells was observed, and cells with more than two pseudopodia were counted as morphological changes. Cells in a total of 3 independent fields of view were counted in each group, for a total of at least 300 cells. HTR-8/SVneo cell invasion assay was performed using a BD BioCoat™ Matrigel™ Invasion Chamber (BD Biosciences, New Jersey, United States) according to the manufacturer’s instructions. A total of 1 ×10⁵ cells were seeded into the upper compartment of the prepared inserts, and medium with 25% FBS was added to the lower compartment to induce migration. After 24 h of incubation at 37°C with 5% CO₂, the cells remaining inside the inserts were removed using a cotton swab. The membranes were then fixed with 4% paraformaldehyde, stained with 0.1% crystal violet (Sangon Biotech, Co., Ltd. Shanghai, China), and washed with ddH₂O. After air-drying the samples, ten independent fields for each group were captured with a Nikon inverted microscope. Furthermore, the whole membranes containing all the stained cells were dissolved in methanol at 4°C for 10 min and mixed, and the absorbance at a wavelength of 560 nm was measured using a UV spectrophotometer. The number of invaded cells was determined by comparing the OD560 values. The experiments were repeated three times for each group under the same conditions.

Adult male and female C57BL/6 mice (6–8 weeks old) were purchased from SIPPR/BK Laboratory Animal Company (Shanghai, China). Heterozygous SEC5-deficient (SEC5^−/+) mice, constructed using CRISPR/Cas9, were purchased from Cyagen Biosciences (Suzhou, China). In short, SEC5 knockout was generated by pronuclear injection of C57BL/6 zygotes with 20 ng/μL Cas9 mRNA and two sgRNAs (10 ng/μL each; sgRNA 1 (5′-AAG​GTT​GTA​TAC​CTA​GAG​AC -3′) and sgRNA2 (5′-AAT​TCT​AGA​ACT​TTG​CCC​GC-3′)). SEC5^−/+ mice were bred and genotyped. Genotypes were confirmed by PCR using the following primers: 5′-AAG​TCA​GGG​GAG​TAA​AGT​ACA​CAC-3′ and 5′-CTC​GTT​ATC​TTT​CAC​TGC AGTATCT-3’. All experiments were carried out in accordance with standard laboratory animal care protocols that were approved by the Institutional Animal Care Committee of Shanghai Institute for Biomedical and Pharmaceutical Technologies (#2018-14). Female mice were mated in natural cycling with males. Mice were inspected every morning for vaginal plugs. The day of vaginal plug detection was designated GD 0.5. Pregnant females were intraperitoneally injected with 250 μg/kg LPS (Sigma–Aldrich, St. Louis, MO, United States) at GD 7.5 to induce abortion. The control group was administered 100 μL of sterile saline solution. All mice were sacrificed at 24 h or 48 h after LPS treatment.

At least three biological replicates were performed for all experiments unless otherwise indicated. Student’s t test was used for statistical analyses of paired observations. Differences between means were accepted as statistically significant at the 95% level (p < 0.05).

CD14⁺ dMϕs were isolated from decidual tissues of RSA patients and healthy pregnant women using immunomagnetic beads. The SEC5 expression level in dMϕs was subsequently determined by qPCR and western blot analyses. The results showed that the expression of SEC5 was significantly decreased in dMϕs of RSA patients (RSA) compared to the expression level in healthy pregnant women (control) (Figures 1A−C). Immunofluorescence costaining showed that SEC5 was widely expressed in human decidual tissues in early pregnancy and abundantly localized in M2-subtype (CD206⁺) macrophages (Figure 1D) and (Pan-CK⁺) trophoblasts (Supplementary Figure S1). As expected, CD206⁺ macrophage cells were significantly reduced in the decidua of RSA patients. In addition, the average fluorescence intensity of SEC5 (red) expressed in CD206⁺ macrophage cells (green) was obviously weaker in the decidua of RSA patients than in that of control women (Figure 1D).

The immortalized human peripheral blood monocyte cell line THP-1 was used to establish a human M0-type Mϕ model (M0) via treatment with PMA, and the M0-type Mϕs were treated with IL-4 and IL-13 to generate M2-type Mϕs (M2) in vitro (Chanput et al., 2014a; Huang et al., 2019a). SEC5 expression in THP1-derived Mϕs was down- or upregulated by transfection with a specific siRNA (siSEC5) or recombinant pCDNA3.0-GFP-Flag-SEC5 plasmid (SEC5-M0/SEC5-M2), respectively. The expression levels of the M2 phenotype markers CD206, CCL22, and TGFβ were detected to evaluate the effect of SEC5 expression on M2 polarization of human Mϕs in vitro. The results showed that the SEC5 expression level in THP1-derived Mϕs was effectively knocked down by siSEC5, and the reduction in SEC5 expression was accompanied by significantly decreased CD206, CCL22, and TGFβ expression levels (Figure 2A), indicating that the downregulation of SEC5 expression had an inhibitory effect on M2 polarization. To verify this finding, we measured the M2 macrophage cell surface marker CD206 via flow cytometry and found that SEC5 knockdown significantly reduced the cell surface expression of CD206 on CD11b⁺ THP-1 cells (Figure 2B). Moreover, the SEC5 expression level in THP-1-derived Mϕs transfected with the pCDNA3.0-GFP-Flag-SEC5 plasmid was obviously increased (Figure 2C), and upregulation of SEC5 expression led to increased CD206 and CCL22 expression levels (Figure 2D), suggesting a stimulatory effect of SEC5 on M2 polarization. Consistently, it was also observed that the M2 polarization of primary BMDMs derived from SEC5^−/+ mice was significantly reduced compared to that of BMDMs from WT mice (Figure 2E), suggesting an inhibitory effect of decreased SEC5 expression on the M2 polarization of mouse Mϕs in vitro. Furthermore, SEC5 expression in mouse BMDMs was significantly reduced by transfection with siSEC5, and downregulation of SEC5 expression led to significantly decreased ARG1 and CD206 expression, two phenotypic markers of M2-type Mϕs in mice (Supplementary Figure S2).

Given that IL-4/IL-13-induced M2 polarization is mediated by STAT6 signaling (Gordon and Martinez, 2010), we investigated the effect of SEC5 on the activation of STAT6 in the M2 polarization of THP-1-derived human Mϕs. The amount of STAT6 translocated to the nucleus was obviously increased after M2 polarization of the THP-1-derived Mϕs but significantly reduced in the SEC5 knockdown Mϕs (Figure 3A). Although the cytoplasmic STAT6 protein expression level in THP-1-derived Mϕs was not affected by M2 polarization or SEC5 knockdown, the level of STAT6 phosphorylated at tyrosine 641 (Y641) (pSTAT6) among the total cytoplasmic proteins was significantly increased after M2 polarization but dramatically reduced in SEC5 knockdown M2 cells (Figure 3B). In contrast, the pSTAT6 level in SEC5-overexpressing THP-1-derived Mϕs was significantly increased after treatment with IL-4 and IL-13 (Figure 3C). However, neither total JAK1 protein nor phosphorylated JAK1 (pJAK1, Tyr1034/1035) was affected by downregulation of SEC5 expression (Figure 3D). In mouse BMDMs, the increased pSTAT6 level caused by M2 polarization was also markedly reduced by SEC5 knockdown (Figure 3E), indicating an inhibitory effect of decreased SEC5 expression on the phosphorylation of STAT6 protein in human and mouse Mϕs. Interestingly, it was found that the pSTAT6 level in dMϕs of RSA patients was also significantly decreased (Figure 3F).

As SEC5 regulated the activation of STAT6, we wondered whether SEC5 directly interacts with STAT6 in Mϕs. The results of coimmunoprecipitation (co-IP) assays showed that the SEC5 antibody precipitated STAT6 protein (Figure 4A) and the STAT6 antibody precipitated SEC5 protein (Figure 4B) in the THP-1-derived Mϕs, and the binding capacity was enhanced after M2 polarization induced by IL-4 and IL-13 (Figures 4A,B), indicating an interaction between SEC5 and STAT6 in human Mϕs. To further confirm this interaction, we determined the spatial distribution of SEC5 and STAT6 in THP-1-derived Mϕs via immunofluorescence staining assays. SEC5 colocalized with STAT6 in both the nucleus and cytoplasm of M0-Mϕs but predominantly in the nucleus of M2-Mϕs, as STAT6 was activated and translocated to the nucleus (Figure 4C). Furthermore, after M2 polarization, phosphorylated STAT6 (pSTAT6) protein signals were obviously enhanced and predominantly localized in the nucleus of Mϕs, and SEC5 protein signals were also colocalized with pSTAT6 (Figure 4D).

It has been reported that M2-Mϕs promote the epithelial-to-mesenchymal transition (EMT), migration and invasion of extravillous trophoblasts (EVTs) by secreting granulocyte-colony stimulating factor (G-CSF) (Ding et al., 2021b). Thus, we evaluated whether downregulation of SEC5 expression would interfere with the effect of M2-Mϕs on EVT activities by using an in vitro coculture model (Figure 5A). After coculture with THP-1-derived M2-Mϕs, the percentage of HTR-8/SVneo cells, an immortalized human first trimester EVT line, with multiple pseudopodia was significantly increased (Figure 5B, black arrows). However, HTR-8/SVneo cells cocultured with SEC5 knockdown Mϕs were contracted and had elongated tentacles, and the percentage of cells with multiple pseudopodia was significantly decreased (Figure 5B). Consistent with the reported data, cell counting and Transwell assay results showed that the proliferation and invasion of HTR-8/SVneo cells cocultured with M2-Mϕs were significantly enhanced compared to those of HTR-8/SVneo cells cocultured with M0-Mϕs, and interestingly, such stimulatory effects were effectively weakened by downregulation of SEC5 expression in Mϕs (Figures 5C,D).

It has been reported that administration of LPS to pregnant mice at an early gestational stage can cause pregnancy loss by destroying immune tolerance at the maternal-fetal interface (Aisemberg et al., 2007; Aisemberg et al., 2013; Zhou et al., 2017). According to previously described methods (Triggianese et al., 2016), we established an early pregnancy loss mouse model via a single intraperitoneal injection of LPS (250 μg/kg) on GD 6.5 (GD 0.5 = day of vaginal plug), and 100% fetal death was observed 24 h after LPS injection (GD7.5), followed by embryonic resorption 48 h after injection (GD8.5) (Figure 6A). qPCR results showed that the expression levels of CD206, IL-6 and TNFα were significantly increased in LPS-treated mice on GD 7.5, whereas the SEC5 expression level was distinctly decreased at both implantation sites (ISs) and nonimplantation sites (NISs) compared with the control GD7.5 mice (Figure 6B). Immunofluorescence costaining assays showed that the number of M2-Mϕs (CD206⁺, green) and the intensity of SEC5 signals (red) in uterine tissues in LPS-treated mice (LPS) seemed to be reduced on GD 8.5 compared to these parameters in normal mice (control) (Figure 6C). Moreover, EVTs (pancytokeratin⁺, green) invading uterine tissues were observed and were localized near Mϕs (Figure 6C). To investigate the role of SEC5 in LPS-induced early pregnancy loss, SEC5 gene deletion mice were constructed using CRISPR/Cas9 technology (Supplementary Figure S2). We intraperitoneally injected wild-type and SEC5^−/+ mice with low doses of LPS (25 μg/kg) on GD 7.5. Uterine tissue from the pregnant SEC5^−/+ mice showed embryonic death on GD 8.5, while the wild-type mice did not (Figure 6D). Furthermore, the expression levels of the proinflammatory cytokines IL-1α, TNFα, IL-1β, and IL-6 were significantly increased at the implantation sites in SEC5^−/+ pregnant mice (Figure 6E).