Ten to eleven weeks old C57BL/6 mice were purchased from the experimental animal center of the Minhang campus of East China Normal University. The animals were exposed to a 12 h dark/12 h light cycle at a fixed humidity level (50–60%) and temperature (21 ± 2 °C). The mice were caged according to a female to male ratio of 2:1, while pregnancy confirmation was determined by obvious visual inspection of the vaginal plug and was marked as day 0 of pregnancy (Day 0). Afterward, the pregnant mice were randomly assigned into 2 groups: intraperitoneal injection of PBS group (control group) and intraperitoneal injection of LPS (PTB group). At 15.5 days of gestation, pregnant mice were intraperitoneally injected with LPS (ultrapure-LPS, Dakewe Biotech Co. Ltd. USA) at a dose of 50 μg/ kg. Following the first PBS or LPS injection, the mice were monitored hourly for any signs of labor (decreased movement, vaginal bleeding, and preterm delivery). The beginning of preterm delivery was defined as the delivery of the first pup. The mice started giving birth 10–18 h post-injection (mainly between 10–12 h). At the initiation of labor in PTB mice, all mice were sacrificed, then blood samples and cervical tissues were collected. All mice were at the same gestational age. The animal study was reviewed and approved by The Ethics Committee of Animal Experiments at Shanghai Tongren Hospital, Shanghai Jiao Tong University School of Medicine.

After labeling the cervical tissues, the specimens were cut into serial macroscopic slices. Some cervical tissues were stained with hematoxylin and eosin (H&E), and the remaining were used for immunohistochemical staining. The slides were observed under a light microscope for cellular changes, and photographs were taken digitally with a NanoZoomer S60 (Hamamatsu, Japan).

H&E staining: The samples were fixed in 10% formalin saline for 24 h, then transferred into 70% ethanol and processed to paraffin-embedded blocks to produce 5 μm thick sections. The samples were then deparaffinized in xylene, rehydrated in descending concentrations of alcohol, and lastly stained with hematoxylin and eosin.

Immunohistochemical staining: Following dehydration in graded ethanol and transfer to xylene, the tissues were embedded in paraffin. After the slices were microwaved in a citrate–phosphate buffer (pH 6.0) to retrieve antigens, they were treated with 3% hydrogen peroxide followed by 10% normal goat serum blocking at room temperature for 30 min. Next, the slices were incubated with primary antibodies (α-smooth muscle actin [α-SMA], smooth muscle heavy chain 22 [SM-22], calponin, cyclo-oxygenase 2 [COX-2], oxytocin receptor [OR], and connexin-43) diluted in PBS for 24 h at 4 °C. The sections were then incubated with secondary antibodies for 30 min at room temperature following several washes in PBS. Finally, the signals were detected using a biotin-streptomycin-hydroxide system using diamorphine as the chromosome. Negative controls were performed with the primary antibodies replaced by PBS.

Immunofluorescence staining: The CSMC were cultured on slides and treated with metabolites, then fixed in 4% paraformaldehyde. The sections were rinsed in PBS and incubated in a non-immune blocking solution for 2 h at room temperature (Triton 100 × and bovine serum protein were dissolved in PBS). Incubation with primary antibodies was performed for 18 h at 4℃. Then, the sections were washed with PBS and incubated with a secondary antibody for 1 h at room temperature way from light. Finally, the sections were again rinsed with PBS, mounted with fluoroshield and DAPI mounting medium (Cat# ab104139, Abcam), and photographed with a Confocal Laser Scanning Microscope (Leica, Germany).

The information of the primary antibodies was as follows: α-SMA polyclonal antibody (Cat No.14395–1-AP, Proteintech, USA), SM22 polyclonal antibody (Cat No.10493–1-AP, Proteintech, USA), calponin polyclonal antibody (Cat No. 13938–1-AP, Proteintech, USA), COX-2 monoclonal antibody (Cat No. 66351–1-Ig, Proteintech, USA), connexin-43 polyclonal antibody (Cat No. 26980–1-AP, Proteintech, USA), and OR polyclonal Antibody (Cat No. 23045–1-AP, Proteintech, USA).

The information of the secondary antibodies was as follows: CoraLite488-conjugated Goat Anti-Mouse IgG(H + L) (Cat No.SA00013-1, Proteintech, USA) and CoraLite488-conjugated Goat Anti-Rabbit IgG(H + L) (Cat No.SA00013-2, Proteintech, USA).

This study was approved by the Ethics Committee of Shanghai Tongren Hospital, Shanghai Jiao Tong University School of Medicine (No.2020–035-02). Cervical tissues were obtained from women undergoing a total hysterectomy for benign indications. The cervical tissues were quickly passed through 75% alcohol and washed twice with PBS, then placed in a 6 cm dish and chopped as much as possible. The chopped tissues were digested using 200 U/mL collagenases (Invitrogen, USA) and 25 kU/mL trypsin (Invitrogen, USA) for 2 h under 37℃ conditions. After observing under the microscope to confirm the tissues became loose, the large tissue pieces were filtered out with a 100-mesh cell sieve (BD, USA)and transferred to a 15 ml centrifuge tube centrifuged at 1500 rpm for 5 min before discarding the supernatant. Afterwards, 4 ml of culture base was resuspended before dividing into Petri dishes. The resuspended culture base was maintained in DMEM (Gibco, USA) with 10% FBS (Gibco, USA), 100 U/ml penicillin, and 100 μg/ml streptomycin in a humidified atmosphere at 37 °C with 5% CO2. The growth medium was changed every 2–3 days. In order to maintain the viability and phenotype of primary muscle cells, continuous plating was carried out, and the culture was limited to 5 generations.

RNA from tissues and cells was extracted using the Trizol reagent (Invitrogen, USA). Then, total RNA was converted into complementary DNA (cDNA) using a Reverse Transcription Kit (Invitrogen, USA). The qPCR was carried out using SYBR Green qPCR SuperMix (Invitrogen, USA) and the ABI PRISM 8000 Sequence Detection System according to the manufacturer’s protocols. The PCR primers are depicted in Table 1. The following conditions were used for amplification: 50℃ for 2 min, 95℃ for 2 min, 40 cycles of 95℃ for 15 s and 60℃ for 60 s. The fluorescence intensities of the probes were plotted against the PCR cycle numbers. GAPDH served as a control gene for gene expression normalization using the 2–ΔΔCt method.

At the onset of PTB mice, all mice were sacrificed, blood samples were collected. Venous blood (1 ml per mouse) was drawn from the tail vein of mice. The blood was centrifuged at 3500 rpm for 10 min at 4℃. Next, the supernatant was transferred to a 1.5 mL centrifuge tube as a serum sample for LC–MS detection. Each frozen serum sample was thawed separately at room temperature. Then, 100 µL serum was added to a 300 µL methanol solution containing 5 µg/ mL L-2-chlorophenylalanine as the internal standard and rotated for 2 min. Centrifugation was done at 13,000 rpm at 4℃ for 10 min. Finally, 200 µL supernatant was extracted. The same volume of serum was extracted from all samples and evenly mixed to prepare quality control (QC) samples.

Ultra-High Performance Liquid chromatography (Ultimate 3000, USA) combined with the thermo-Orbitrap Elite mass spectrometer was utilized for the LC–MS analysis. The system was equipped with an electrospray ionizationsource and operated in either positive or negative ionizationmode using a mass resolution of 70, 000 at an m/z of 200. Data-dependent (dd-MS2, Top N = 10) MS/MS mode with a full scanmass resolution of 17, 500 at an m/z of 200 was used. The scanrange was 100–1, 500. Metabolic profiles in electrospray ionization (ESI) positive and ESI negative ion modes were performed using an ACQUITY UPLC I-Class system (Waters Corporation, USA) coupled with an AB SCIEX Triple TOF 5600 System (AB SCIEX, USA). The binary gradient elution systems consisted of water containing 0.1% formic acid, v/v (A), and acetonitrile containing 0.1% formic acid, v/v, (B). 20% B for 2 min; 60% B for 4 min; 100% B for 11 min; 100% B for 13 min % B for 13.5 min and finally, 5% B for 14.5 minutesnthe above steps, separation was achieved. The chromatographic conditions were as follows: injection volume was 2 μl; column temperature was 25 °C; flowrate was 0.35 ml/min. Data were acquired in centroid mode using the Thermo Excalibur 2.2 software (Thermo Fisher Scientific, USA).

Statistical analyses were performed by the SPSS Statistics 22.0 software (IBM, USA) and GraphPad Prism 9.0 software (GraphPad Software Inc, USA). Statistical significance was determined according to the sample distribution and homogeneity of variance, while statistical comparisons between two groups were determined by the t-test. The metabolic data were acquired using the Thermo Xcalibur 2.2 software (Thermo Scientific, USA). Peak alignment and extraction were performed using the Compound Discoverer software (Thermo Scientific, USA). Next, a data table containing information regarding the retention time, m/z, and peak area was obtained. The edited data matrix was imported into the SIMCA-P 11.0 software (Umetrics, Sweden) for multivariate statistical analysis, principal components analysis (PCA), and partial least squares discrimination analysis (PLS-DA). The unsupervised PCA analysis assessed the overall trend of segregation between the samples, while a supervised PLS-DA analysis model screened for significantly different metabolites between the PTB and control groups. The ion peaks were normalized and Pareto-scaled. According to PLS-DA model, the variables with variable importance in the projection (VIP) value > 1.0 were selected and p < 0.05 was considered as statistically significant. Bonferroni correction was used for multiple testing adjustment. In order to identify these potential biomarkers, the accurate ion mass was input into the human metabolome database (HMDB, https://hmdb.ca) databases to match the exact molecular weight, and MS1/MS2 fragment ions were automatically searched. Finally, in order to confirm the structure of the compound, we used our internal standard metabolite library, matching the exact mass,fragment ion mass, and retention time.The ingenuity pathway analysis (IPA) from the Kyoto Encyclopedia of Genes and Genomes (KEGG) online database was applied to understand the functions and interactions of genes and metabolites. * p < 0.05, ** p < 0.01 and *** p < 0.001.