Six patients with PE and 6 healthy pregnant control patients were recruited from June 2022 to December 2023. The plasma from three PE patients and three healthy controls was used for high-throughput sequencing of tRF and tiRNA, while the plasma and placental trophoblast tissues of three other PE patients and three healthy controls were used to verify the consistency of tsRNA expression levels. To assess the potential of tsRNA as a predictive biomarker for PE, plasma samples were also collected from 150 patients with suspected PE, of which 28 did not give birth in our hospital. Consequently, the study analyzed 122 patients with suspected pre-eclampsia (PE). Given the documented 20% incidence of PE in women who are suspected of having this condition [18], the tiRNA-Gln-CTG biomarker was anticipated to reach 80% sensitivity and specificity, with Z = 1.96. The minimum sample size required was therefore calculated as: n = 1.96² $\times$ 0.2 $\times$ (1 – 0.2)/0.2², corresponding to a minimum sample size of 62. PE was diagnosed according to the 2020 ACOG Clinical Guidelines [19], and suspected PE according to previously reported criteria [18, 20].

For high-throughput sequencing, plasma total RNA was first extracted using TRIzol reagent (15596026CN, Invitrogen, Carlsbad, CA, USA) as per the manufacturer’s guidelines. It was then quantified using a NanoDrop-1000 (ND-1000, ThermoFisher Scientific, Waltham, MA, USA) instrument. The rtStarTM tRF & tiRNA Preprocessing Kit (AS-FS-005, Arraystar, Rockville, MD, USA) was used to eliminate RNA modifications that hinder the formation of small RNA sequencing libraries, including the removal of 3^′-amino acids and 3^′-cyclic phosphate for 3^′-adaptor ligation, phosphorylation of 5^′-OH to 5^′-P for 5^′-adaptor ligation, and demethylation of m1A and m3C. Sequencing libraries were prepared from pretreated total RNA in the following steps: (1) ligation of 3^′-adaptors; (2) ligation of 5^′-adaptors; (3) synthesis of complementary deoxyribonucleic acid (cDNA); (4) amplification via polymerase chain reaction (PCR); and (5) identification of PCR-amplified fragments within the size range of 134–160 bp. The amplified cDNA libraries were quantitatively analyzed using an Agilent 2100 Bioanalyzer (G2939A, Agilent, Santa Clara, CA, USA), and tsRNA high-throughput sequencing was performed with the Illumina NextSeq 500 system (NEXTSEQ 500, Illumina, San Diego, CA, USA). Following the generation of raw sequencing data, mature tRNA libraries were prepared by removing intronic sequences and appending ‘CCA’ to the 3^′ end. FastQC was used to assess sequencing quality. 5^′- and 3^′-adapter bases were trimmed by Cutadapt through Illumina quality filters. Trimmed reads were then aligned, allowing only one mismatch with the mature tRNA sequence. Unmapped reads were subsequently aligned to the precursor tRNA sequence using Bowtie software (Version 1.3.0, The Johns Hopkins University, Baltimore, MD, USA), with only one mismatch permitted. The abundance of tRF and tiRNA was assessed using their sequencing counts, which were normalized to counts per million (CPM) aligned total reads [21].

The HTR-8/SVneo cell line was obtained from Servicebio Technology Co., Ltd. (Wuhan, Hubei, China) and cultured in RPMI-1640 (G4535, Servicebio, Wuhan, Hubei, China) with 10% fetal bovine serum (FBS, Gibco, Rockville, MD, USA) at 37 °C and in 5% CO₂. Cell lines underwent short tandem repeat (STR) profiling for validation and tested negative for mycoplasma. The HTR-8/SVneo PE-like hypoxia model was developed using 100 µM cobalt chloride hexahydrate (CoCl₂$\cdot$6H₂O, C106772, Aladdin, Shanghai, China) as described previously [22]. Whether the construction of the hypoxia model of HTR-8/SVneo cells was successful was verified by detecting hypoxia-inducible factor-1$\alpha$ (HIF-1$\alpha$) through Western blot (WB) assay.

RNA was extracted from cells, plasma, and placenta using TRIzol reagent, and its quality and quantity assessed using the NanoDrop-1000. Quantitative real-time PCR (qRT-PCR) was performed using the StepOnePlus Real-Time PCR System from Applied Biosystems (ThermoFisher Scientific, Waltham, MA, USA). The tsRNA was synthesized using the non-human-derived spike-in microRNA (miRNA) cel-miR-39-3p (miRA0000010-1, Ribobio, Guangzhou, Guangdong, China) as an exogenous reference for normalization in plasma, and RNU6B (U6, Ribobio, China) as an endogenous reference for normalization in placenta. Standardization of the messenger RNA (mRNA) level in cells was achieved using $\beta$-actin. The relative RNA expression was calculated by 2^{-Δ⁢Δ⁢CT}. The primer sequences used for human THBS2 were: forward, 5^′-GACACGCTGGATCTCACCTAC-3^′ and reverse, 5^′-GAAGCTGTCTATGAGGTCGCA-3^′. The primer sequences for human $\beta$-actin were: forward, 5^′-CATGTACGTTGCTATCCAGGC-3^′ and reverse, 5^′-CTCCTTAATGTCACGCACGAT-3^′. Primers for tsRNA reverse transcription and PCR were provided by Ribobio, Guangdong, China. Detailed information and samples can be provided by e-mail at marketing@ribobio.com.

HTR-8/SVneo cells were seeded into 6-well plates and incubated in complete medium for 24 h until reaching 60–70% confluence. They were then transfected with a negative control (NC, sequence: GUUGUAGCUCCGCAAUUGCCGUUCAAUGUUGGGA), a tiRNA-Gln-CTG mimic (sequence: GGUUCCAUGGUGUAAUGGUUAGCACUCUGGACUC), or a tiRNA-Gln-CTG inhibitor (sequence: GAGUCCAGAGUGCUAACCAUUACACCAUGGAACC). All sequences were obtained from Ribobio (China) and were added with 200 µL Opti-MEM (31985070, ThermoFisher Scientific, Waltham, MA, USA) and 5 µL Lipofectamine 3000 (L3000015, ThermoFisher Scientific, Waltham, MA, USA) to transiently transfect HTR-8/SVneo cells. Subsequent experiments were performed 24–48 h after cell transfection. Detailed information and samples can be requested from Robibio by e-mail (marketing@ribobio.com). Negative control (NC, UUCUCCGAACGUGUCACGUTT), THBS2 small interfering (siRNA, GUACCUGCAAGAAAUUUAATT), and Y-box-binding protein 1 (YBX1) siRNA (GGAGGCAGCAAAUGUUACATT) were obtained from GenePharma (Shanghai, China) and transfected into HTR-8/SVneo cells using Lipofectamine 3000 as per the manufacturer’s guidelines.

TargetScan (https://www.targetscan.org) and miRanda (https://anaconda.org/bioconda/miranda) [23] were employed to predict target genes for the selected tRFs/tiRNAs. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses were conducted using Metascape bioinformatics resources (https://metascape.org) [24], followed by an investigation of gene/protein interactions through the String database (https://string-db.org) [25].

At 24 h after transfection, HTR-8/SVneo cells were seeded into 96-well plates at a density of 3000 cells per well. After incubating for 12, 24, and 48 h, 10 µL of CCK-8 reagent (CK04, Dojindo, Kumamoto Prefecture, Japan) was added to each well. Following 2 h of incubation, the optical density at 450 nm was analyzed using a multifunctional enzyme marker (Synergy™ 4, Biotek, Winooski, VT, USA).

At 24 h after transfection, HTR-8/SVneo cells were seeded at 10,000 per well in 96-well plates and cultured for 24 h. The EdU proliferation assay was carried out as per the manufacturer’s guidelines (G1601, Servicebio, Wuhan, Hubei, China), and the proportion of cells that incorporated EdU was ascertained by fluorescence microscopy (Zeiss, Oberkochen, BW, Germany). The cell proliferation rate (%) = the number of positive cells (green fluorescence) / the number of total cells (blue fluorescence).

Analysis of cell apoptosis. Cells were transfected, digested with EDTA-free trypsin, washed twice with phosphate-buffered saline (PBS), and then processed as per the manufacturer’s guidelines (40310ES60, YEASON, Shanghai, China). Subsequently, apoptosis was assessed using flow cytometry (Attune NxT, Thermo Fisher Scientific, USA).

Analysis of the cell cycle. Transfected cells were digested, washed twice with PBS, and then processed as per the manufacturer’s guidelines (40301ES60, YEASON, Shanghai, China). At the completion of treatment, flow cytometry was used to analyze the cell cycle (Attune NxT, Thermo Fisher Scientific, Waltham, MA, USA).

Transfected HTR-8/SVneo cells were cultured in 6-well plates until they were fully confluent. Scratch wounds were then created using a 200 µL pipette tip, and three points were randomly marked. After washing the 6-well plates with PBS to eliminate dead cells, 2 mL of RPMI 1640 medium without FBS was added. Images were taken at designated sites at 0, 24, and 48 h post-incubation and analyzed with Image J software (Version 1.8.0, National Institutes of Health, Bethesda, MD, USA) to determine the wound area and calculate the healing rate. The wound healing rate was determined by subtracting the wound area at 24 or 48 h from the initial wound area (0 h), then dividing by the initial wound area.

Cell migration assay. A total of 30,000 HTR-8/SVneo cells and 200 µL of FBS-free RPMI 1640 medium were introduced into the upper chamber of a Transwell (3422, Corning, Steuben County, New York, USA). Simultaneously, 700 µL of RPMI 1640 medium with 10% FBS was added to the lower chamber. Following removal of the culture medium from the upper chamber, samples were fixed with 700 µL of 4% paraformaldehyde for 30 minutes. The cells were then stained in the dark for 30 minutes using a 0.5% crystal violet dye solution (C0121, Beyotime, Shanghai, China). The cells and excess dye solution remaining on the polycarbonate membrane in the upper chamber were removed with cotton swabs, and the chamber was subsequently dried. Cells that penetrated the polycarbonate membrane were imaged using a Zeiss inverted microscope, and their number quantified using Image J software.

Transwell invasion assay. Before the experiment, RPMI 1640 culture medium was mixed with Matrigel matrix glue (354234, Corning, Steuben County, New York, USA) in a 1:7 volume ratio. Subsequently, 60 µL of this mixture was placed in the upper chamber of the Transwell and left to incubate overnight. The remaining experimental procedures were the same as those used in the cell migration assay.

A total of 400 µL of HTR-8/SVneo cellular protein solution was divided equally into 4 groups (Sense, Antisense, Input and Beads). Biotin-labeled tiRNA-Gln-CTG and its control probe were incubated with cytoplasmic protein solution at 4 °C overnight (no probe was added to the beads group) to form RNA-protein complexes. Agarose beads (30 µL) were added to three 1.5 mL enzyme-free EP tubes (151440, MedChemExpress, Monmouth Junction, NJ, USA) and centrifuged for 30 seconds. The upper liquid layer was discarded and the beads were resuspended in 200 µL of PBS and centrifuged at room temperature (3000 g, 1 minute). The supernatant was discarded and the procedure was repeated five times. The RNA-protein complexes were then combined with the agarose beads and incubated at 4 °C for 1 h with rotation. Following centrifugation (3000 g, 1 minute), the supernatant was discarded and the pellet was resuspended in 200 µL of PBS. This step was repeated nine times. Finally, after adding 40 µL of SDS-PAGE sample buffer and incubating in a 100 °C water bath for 10 minutes, the samples underwent 10% SDS-PAGE electrophoresis followed by rapid silver staining (P0017S, Beyotime, Shanghai, China).

The SDS gel was cut into 1 mm pellets with a scalpel and placed into 2 mL centrifuge tubes. Decolorizing solution (25 mM ammonium bicarbonate and acetonitrile) was added and the tubes were evenly shaken. This was repeated several times until the destaining was complete. Ethyl ether was added to dehydrate the pellet until it turned white, followed by vacuum drying. Dithiothreitol (10 mM, D9779, Sigma-Aldrich, St. Louis, MO, USA) was then added and incubated for 60 min at 56 °C. An iodoacetamide solution (55 mM, RPN6302, GE Healthcare, Chicago, IL, USA) was added and reacted for 45 min in the dark. A trypsin solution (10 ng/µL, V5111, Promega, Madison, WI, USA) was then added to fully cover the pellet and left at 37 °C overnight. The next day, the extraction solution (50% acetonitrile + 3% trifluoroacetic acid) and 100% acetonitrile were added and the sample centrifuged. The resulting supernatant with enzyme solution was collected in a new centrifuge tube, vacuum dried and then stored at –20 °C. Peptides were analyzed via elution using an EASY-nLC 1200 UHPLC system (EASY-nLC 1200, Thermo Fisher Scientific, Waltham, MA, USA), and using a Thermo Q Exactive HF-X mass spectrometer (Q Exactive HF-X, Thermo Fisher Scientific, Waltham, MA, USA) following elution and ionization.

The 3^′UTR of the THBS2 sequence containing the tiRNA-Gln-CTG binding site was inserted into the SacI/XhoI restriction site of the pmirGLO vector (GenePharma, Shanghai, China). It contained either a wild-type (WT) or mutant (MUT) sequence. HEK293T cells were transfected with a luciferase-reporter vector and either a tiRNA-Gln-CTG mimic or control using Lipofectamine 3000. After transfection for 48 h, firefly and renilla luciferase activities were measured using a luciferase assay system (Glomax 96, Promega, Madison, WI, USA) as per the manufacturer’s guidelines (G1701, Servicebio, Wuhan, Hubei, China).

HTR-8/SVneo cells underwent two PBS washes before the addition of 60 µL of a radioimmunoprecipitation (RIPA) assay lysate mixture containing phenylmethanesulfonyl fluoride and a protease inhibitor (all from Beyotine, Shanghai, China). The cells were thoroughly lysed on ice and the protein concentration was then measured with a BCA protein assay kit (P0012, Beyotine, China). Although 20 µg of protein is typically utilized for WB experiments, we loaded 5–10 µg of extracellular vesicle (EV)-derived protein for exosome identification, and 10 µg for cell-derived proteins. The sample volume used for the EV-free group was 15 µL. The ratio of cell supernatant without EVs to loading buffer was 4:1. Following separation by 10% polyacrylamide gel electrophoresis, proteins were transferred to a polyvinylidene fluoride membrane. After blocking with 5% skim milk or bovine serum albumin for 2 h, membranes were incubated overnight on a shaker with primary antibody. They were then washed three times for 10 minutes each with Tris Buffered Saline containing Tween-20 (TBST) on a shaker, followed by incubation with the secondary antibody for 1 h at room temperature. After three washes with TBST, protein bands were visualized with a chemiluminescent reagent (WBKLS0500, Darmstadt, Germany). Image J software was used to semi-quantify the grayscale values of the immunoblots. Information on the antibodies used in WB is presented in Supplementary File 1.

HTR-8/SVneo cells were incubated for 36 h in a medium containing EV-free FBS (C380109199, VivaCell Bioscience, Shanghai, China). The medium was then collected and EVs wereisolated using the VEX Exosome Isolation Reagent (R601, Vazyme, Nanjing, Jiangsu, China) as per the manufacturer’s instructions. Cell culture medium was filtered through a 0.22 µM pore filter to remove cell debris, followed by centrifugation at 5600 rpm for 20 minutes at 4 °C. Subsequently, Exosome Isolation Reagent was added to the supernatant at a 1:3 ratio and mixed thoroughly. After refrigerating for 16 h at 4 °C, the supernatant was removed after centrifugation (10200 rpm, 30 minutes, 4 °C) and the pelletresuspended in PBS to obtain the exosome suspension. The size and shape of EVs were observed using transmission electron microscopy (HT7800, Hitachi, Chiyoda-ku, Tokyo, Japan), and the expression of cluster of differentiation 63 (CD63), cluster of differentiation 81 (CD81), ALG-2-interacting protein X/apoptosis-linked gene 2-interacting protein 1 (ALIX), tumor susceptibility gene 101 (TSG101), heat shock protein 70 (HSP70) and golgi matrix protein 130 (GM130) proteins was evaluated by WB assay.

HTR-8/SVneo cells were cultured in 24-well plates with cell coverslips. They were then fixed in 4% formaldehyde and permeabilized with 0.5% Triton X-100. A FAM-labeled tiRNA-Gln-CTG probe was designed and synthesized by GenePharma (Shanghai, China). The FISH assay was performed as recommended bythe reagent supplier (C10910, RiboBio, China). A suitable field of view was chosen and imaged using a Zeiss fluorescence microscope (Oberkochen, Germany) to examine the localization of tiRNA-Gln-CTG in HTR-8/SVneo cells.

Data analysis was conducted using SPSS (Version 20.0, International Business Machines Corporation, Armonk, New York, USA) and GraphPad Prism (Version 8.0, GraphPad software, San Diego, CA, USA). Study results were expressed as the mean $\pm$ standard deviation (SD). Statistical significance between two data groups was determined using a two-tailed Student’s t-test. One-way ANOVA followed by Tukey’s post hoc test was used for comparisons among multiple data groups. The Shapiro-Wilk test evaluated the normality of tsRNA level distributions in plasma, as well as the baseline characteristics of women with suspected PE. Normal distributions were analyzed using the unpaired Student’s t-test. Non-normal distributions were assessed with the Mann-Whitney U test, with data presented as the median and interquartile range. A p-value of 0.05 or lower was considered statistically significant. Individual p-values were noted when appropriate. The predictive diagnostic efficacy of serum tiRNA-Gln-CTG for suspected PE cases was evaluated using a receiver operating characteristic (ROC) curve and the area under the curve (AUC) Each experiment was conducted independently a minimum of three times. The data of qRT-PCR, EdU, wound healing, transwell and WB experiments were all presented after normalization. The level of tiRNA in the plasma of patients with suspected PE was presented in the form of $\mathrm{\Delta }$CT.

High-throughput sequencing of tsRNA was performed to compare the level of tsRNA in plasma from PE and healthy control women (Fig. 1A). Different subclasses of tsRNAs were distributed according to their size and length (Fig. 1B,C). Results above indicated that significant differences existed in the composition of the tsRNA profile and the expression level of tsRNA between the PE and healthy control groups. Identifying the differential tsRNAs is the primary step for their roles in the development of PE and their potential as predictive markers for PE. A total of 1532 tsRNAs showed elevated counts per million (CPM) mapped read values of $>$1.5-fold in the PE plasma compared to healthy controls, while 814 showed decreased values (Fig. 1D). Of these, 28 were significantly up-regulated (unadjusted p 0.05), while 59 were significantly down-regulated (Fig. 1E). Fig. 1F shows the categorization of the differential tsRNAs.

Fig. 1.

Expression profiles of tsRNA in the plasma of pregnant women with PE (n = 3) and healthy pregnant women (n = 3). (A) Heat map of differential tsRNAs. (B) Distribution of tsRNA species in the plasma of healthy pregnant women. (C) Distribution of tsRNA species in the plasma of women with PE. (D) tsRNAs with $>$1.5-fold differential expression. (E) Volcano plot of differential tsRNAs. (F) Categorization of the differential tsRNAs. PE, pre-eclampsia; tsRNA, transfer ribonucleic acid-derived small RNA; tiRNA, transfer ribonucleic acid halves; tRF, transfer ribonucleic acid-derived fragment.

tsRNAs with CPM $>$1.5 and p 0.05 were selected for potential biological function analysis using the TargetScan, miRanda, and Metascape bioinformatics tools. Ultimately, 6 tsRNAs were chosen for further experiments based on GO and KEGG analysis results suggesting they were potentially involved in the development of PE (Fig. 2A). These tsRNAs were subsequently analyzed in the placenta and plasma samples of both PE patients and healthy controls. The expression levels of the 6 tsRNAs in plasma aligned closely with the high-throughput sequencing results (Fig. 2B). High-throughput tRF and tiRNA sequencing revealed that tRF-SeC-TCA expression levels were significantly lower in the plasma of patients with PE compared to healthy controls. The plasma verification results showed that the levels of tRF-SeC-TCA in patients with PE did not decrease significantly. The expression of three of these tsRNAs (tRF-Ser-TGA, tiRNA-Gln-CTG and tRF-Cys-GCA) in the placenta was consistent with that found in plasma (Fig. 2C). tiRNA-Gln-CTG was chosen for further analysis in subsequent experiments because the biological process of its possible 3^′UTR mRNA binding targets were involved in the cellular response to oxidative stress, cell migration, and cell proliferation (Fig. 3A). Oxidative stress and altered cellular functions in the placenta are key factors in the pathogenesis of PE. Moreover, the cellular component involved the cell leading edge, cell body, and perinuclear region of cytoplasm (Fig. 3B). The molecular function included salt transmembrane-transporter activity, while protein molecular function encompassed salt transmembrane transporter activity, protein serine/threonine kinase activity, and acyltransferase activity (Fig. 3C). KEGG pathway analysis also indicated that tiRNA-Gln-CTG may play a role in classical signaling pathways, including mitogen-activated protein kinase (MAPK), forkhead box, sub-group O (FoxO), and phosphatidylinositol 3-kinase/protein kinase B (PI3K-AKT, Fig. 3D). Protein-protein interaction (PPI) network maps were drawn to analyze the potential core target genes of tiRNA-Gln-CTG (Fig. 3E). The structure of tiRNA-Gln-CTG is shown in Fig. 3F.

Fig. 2.

Expression levels of 6 selected tsRNAs in the plasma and placenta of women with pre-eclampsia and healthy pregnant controls. Data are presented as the mean $\pm$ SD. (A) Basic information about the 6 tsRNAs. (B) Expression levels in plasma for the PE and control groups. (C) Expression levels in placenta for the two groups.

Fig. 3.

Analysis of the potential biological functions of tiRNA-Gln-CTG. (A) Biological process categories. (B) Cellular component categories. (C) Molecular function categories. (D) Canonical signaling pathways. (E) PPI network analysis. (F) Structure of tiRNA-Gln-CTG. PPI, protein-protein interaction.

FISH experiments revealed that tiRNA-Gln-CTG was expressed in both the cytoplasm and nucleus of HTR-8/SVneo cells (Fig. 4A). Hypoxia models of HTR-8/SVneo cells were created using varying CoCl₂ concentrations to mimic the placental hypoxic conditions in PE [26, 27, 28]. The expression level of HIF-1$\alpha$ in the cells was evaluated, leading to the selection of 100 µM CoCl₂ for subsequent experiments (Fig. 4B). The concentration of CoCl₂ used previously in the literature ranges mainly from 50 to 250 µM. Our experiments revealed that concentrations between 50–250 µM significantly increased the HIF-1$\alpha$ protein level. However, HTR-8/SVneo cells underwent significant apoptosis when the CoCl₂ concentration exceeded 100 µM, as observed under the microscope. Therefore, the concentration of 100 µM was selected for subsequent experiments. EVs were isolated from the cell culture medium and their structure analyzed by transmission electron microscopy (Fig. 4C). The expression of EV signature proteins was also detected using WB (Fig. 4D). Following CoCl₂-induced hypoxia, the expression of tiRNA-Gln-CTG was significantly reduced in HTR-8/SVneo cells, in the cell culture medium, and in EVs present in the culture medium (Fig. 4E–G).

Fig. 4.

Localization of tiRNA-Gln-CTG in HTR-8/SVneo cells, and its expression and secretion in a CoCl₂-induced HTR-8/SVneo cell model of PE. Data are presented as the mean $\pm$ SD. (A) Localization of tiRNA-Gln-CTG in HTR-8/SVneo cells (scale bar = 50 µm). (B) HIF-1$\alpha$ expression levels in the PE trophoblast cell model under different concentrations of CoCl₂. For subsequent experiments, 100 µM CoCl₂ was selected to induce hypoxia. (C) Transmission electron microscopy showing the morphology of EVs extracted from HTR-8/SVneo cell culture medium (scale bar = 10 µm, left; scale bar = 2000 nm, right). (D) Western blot identification of the signature proteins for EVs. (E) Expression level of tiRNA-Gln-CTG in HTR-8/SVneo cells from the PE trophoblast cell model, n = 4 per group. (F) Expression level of tiRNA-Gln-CTG in EVs from the culture medium of the PE trophoblast cell model, n = 4 per group. (G) Expression level of tiRNA-Gln-CTG in culture medium from the PE trophoblast cell model, n = 4 per group. EV, extracellular vesicle; NC, normal control; CD63, cluster of differentiation 63; CD81, cluster of differentiation 81; ALIX, ALG-2-interacting protein X/apoptosis-linked gene 2-interacting protein 1; TGS101, tumor susceptibility gene 101; HSP70, heat shock protein 70; GM130, golgi matrix protein 130. NC, negative control.

Transfection with the tiRNA-Gln-CTG inhibitor did not significantly alter the expression of tiRNA-Gln-CTG in HTR-8/SVneo cells compared to the control group (Fig. 5A). However, the cell viability (Fig. 5B), proliferation (Fig. 5C), wound healing (Fig. 5D), and cell migration and invasion (Fig. 5E) were all significantly reduced 48 h after transfection with the inhibitor. Moreover, cells transfected with the tiRNA-Gln-CTG inhibitor exhibited a lower proportion of cells in the S-phase, and a significantly higher proportion in the G0/G1 phase (Fig. 5F). No significant difference in the apoptosis level (Fig. 5G) was observed between the two groups.

Fig. 5.

Effect of transfection with the tiRNA-Gln-CTG inhibitor on HTR-8/SVneo cell function. Data are presented as the mean $\pm$ SD. (A) The tiRNA-Gln-CTG level after transfection with tiRNA-Gln-CTG inhibitor, n = 4 per group. (B) CCK-8 assay, n = 6 per group. (C) EdU assay, n = 5 per group (scale bar = 100 µm). (D) Wound healing assay, n = 3 per group (scale bar = 100 µm). (E) Transwell cell migration and invasion assay, n = 5 per group (scale bar = 100 µm). (F) Cell cycle, n = 3 per group. (G) Cell apoptosis, n = 3 per group. EdU, 5-ethynyl-2’-deoxyuridine.

Transfection with the tiRNA-Gln-CTG mimic led to a significant 55-fold increase in tiRNA-Gln-CTG expression in HTR-8/SVneo cells (Fig. 6A). Cell viability (Fig. 6B), cell proliferation (Fig. 6C), wound healing (Fig. 6D), and cell migration and invasion (Fig. 6E) were all significantly increased 48 h after transfection with the tiRNA-Gln-CTG mimic. The mimic group also exhibited a significantly higher proportion of cells in the S-phase, and a reduced proportion in the G0/G1 phase compared to the mimic-NC group (Fig. 6F). No significant difference in the apoptosis level was observed between the two groups (Fig. 6G).

Fig. 6.

Effect of transfection with the tiRNA-Gln-CTG mimic on HTR-8/SVneo cell function. Data are presented as the mean $\pm$ SD. (A) The tiRNA-Gln-CTG level after transfection with the tiRNA-Gln-CTG mimic, n = 4 per group. (B) CCK-8 assay, n = 6 per group. (C) EdU assay, n = 6 per group (scale bar = 100 µm). (D) Wound healing assay, n = 3 per group (scale bar = 100 µm). (E) Transwell cell migration and invasion assay, n = 5 per group (scale bar = 100 µm). (F) Cell cycle, n = 3 per group. (G) Cell apoptosis, n = 3 per group.

The TargetScan and miRanda websites predicted that tiRNA-Gln-CTG binds to the THBS2 mRNA 3^′UTR region as 8mer-1u (Fig. 7A). The 3^′UTR of THBS2 containing WT or MUT sequence for the tiRNA-Gln-CTG binding site was inserted into the SacI/XhoI restriction site of the pmirGLO Dual-Luciferase miRNA Target Expression Vector (Fig. 7B). Dual luciferase assay showed the relative fluorescence intensity was elevated when HEK293T cells were simultaneously transfected with tiRNA-Gln-CTG inhibitor and GP-miRGLO-THBS2-WT, whereas the relative fluorescence intensity was significantly decreased when they were simultaneously transfected with tiRNA-Gln-CTG mimic and GP-miRGLO-THBS2-WT (Fig. 7C). Transfection of HTR-8/SVneo cells with a tiRNA-Gln-CTG inhibitor significantly increased THBS2 mRNA and protein levels, while transfection with a tiRNA-Gln-CTG mimic significantly reduced these levels (Fig. 7D,E). Moreover, THBS2 protein expression was significantly higher in PE placentas (Fig. 7F), and the THBS2 protein level was also significantly increased in CoCl₂-induced HTR-8/SVneo cells (Fig. 7G). These findings indicate that THBS2 significantly influences the biological functions of tiRNA-Gln-CTG.

Fig. 7.

tiRNA-Gln-CTG is involved in the development of PE by mediating THBS2 mRNA degradation through binding to its 3^′UTR region. Data are presented as the mean $\pm$ SD. (A) Pairing site of tiRNA-Gln-CTG with THBS2 mRNA. (B) The 3^′UTR of THBS2 containing a WT or MUT sequence for the tiRNA-Gln-CTG binding site was inserted into the SacI/XhoI restriction site of the pmirGLO Dual-Luciferase miRNA Target Expression Vector. (C) Dual luciferase reporter gene assay was used to verify the targeting between tiRNA-Gln-CTG and THBS2 mRNA, n = 6 per group. (D) THBS2 mRNA expression level after transfection of HTR-8/SVneo cells with tiRNA-Gln-CTG inhibitor or mimic, n = 4 per group. (E) THBS2 protein expression level after transfection of HTR-8/SVneo cells with tiRNA-Gln-CTG inhibitor or mimic, n = 3 per group. (F) Western blot detection of THBS2 protein expression in placental trophoblasts, n = 3 per group. (G) Western blot detection of THBS2 protein expression in the PE trophoblast cell model induced by CoCl₂, n = 3 per group. THBS2, thrombospondin-2; WT, wild type; MUT, mutant; mRNA, messenger ribonucleic acid.

Co-transfection of HTR-8/SVneo cells with a tiRNA-Gln-CTG inhibitor and THBS2-siRNA significantly reduced THBS2 mRNA (Fig. 8A) and protein (Fig. 8B) expression levels in the cells. Reduction of THBS2 expression improved the viability (Fig. 8C), proliferation (Fig. 8D), wound healing (Fig. 8E), and migratory and invasive abilities (Fig. 8F). Reduced THBS2 expression also lowered the proportion of cells in the G0/G1 phase, while increasing the proportion in the proliferative phase (Fig. 8G), however the apoptosis level remained unchanged (Fig. 8H).

Fig. 8.

Cellular functions of HTR-8/SVneo cells after co-transfection with tiRNA-Gln-CTG inhibitor and siRNA-THBS2. Data are presented as the mean $\pm$ SD. (A) THBS2 mRNA expression level after co-transfection of tiRNA-Gln-CTG inhibitor with siRNA-THBS2, n = 4 per group. (B) Western blot detection of THBS2 protein expression after co-transfection of tiRNA-Gln-CTG inhibitor with siRNA-THBS2, n = 3 per group. (C) CCK-8 assay, n = 6 per group. (D) EdU assay, n = 4 per group (scale bar = 100 µm). (E) Wound healing assay, n = 4 per group (scale bar = 100 µm). (F) Transwell cell migration and invasion assay, n = 4 per group (scale bar = 100 µm). (G) Cell cycle, n = 3 per group. (H) Cell apoptosis, n = 3 per group.

tiRNA-Gln-CTG sense and antisense molecules were identified by nucleic acid electrophoresis (Fig. 9A). RNA pull-down assay (Fig. 9B) and mass spectrometry revealed that tiRNA-Gln-CTG could specifically bind to 45 proteins (Fig. 9C). YBX1 was chosen for further validation due to its involvement in the regulation of gene expression and in non-coding RNA incorporation into the exosome. In vitro experiments confirmed that tiRNA-Gln-CTG specifically binds to YBX1 (Fig. 9D). The YBX1 expression level was reduced after transfection of YBX1-siRNA (Fig. 9E). The expression of tiRNA-Gln-CTG in HTR-8/SVneo cells remained unchanged following transfection of YBX1-siRNA (Fig. 9F), but its expression was significantly reduced in EVs present in the culture medium (Fig. 9G), as well as in the culture medium itself (Fig. 9H). AKT occupies a central position in the PPI interaction network analysis. The present results suggest that tiRNA-Gln-CTG exerts its biological function possibly through its regulation of AKT phosphorylation (Fig. 9I). This study also identified several signature proteins associated with the biological function of tiRNA-Gln-CTG, notably changes in the expression levels of cyclin D1, proliferating cell nuclear antigen (PCNA), vimentin, N-cadherin, and E-cadherin, whereas the expression levels of B-cell lymphoma/leukemia-2 (Bcl2) and Bcl-2 associated X (Bax) remained unchanged (Fig. 9J,K).

Fig. 9.

Preliminary investigation of the mechanism by which tiRNA-Gln-CTG is sorted into EVs and mediates their biological functions. Data are presented as the mean $\pm$ SD. (A) Identification of sense and antisense tiRNA-Gln-CTG, n = 3 per group. (B) RNA pull-down assay used to probe the proteins bound to tiRNA-Gln-CTG. (C) Mass spectrometry analysis used to probe differential proteins in sense and antisense tiRNA-Gln-CTG. (D) In vitro validation of the binding of tiRNA-Gln-CTG to YBX1. (E) YBX1 mRNA expression levels after transfection of YBX1-small interfering RNA (siRNA), n = 4 per group. (F) tiRNA-Gln-CTG expression level in HTR-8/SVneo cells after YBX1 silencing, n = 4 per group. (G) tiRNA-Gln-CTG expression level in EVs secreted by HTR-8/SVneo cells after YBX1 silencing, n = 4 per group. (H) tiRNA-Gln-CTG expression level in the HTR-8/SVneo cell culture medium after YBX1 silencing, n = 4 per group. (I) tiRNA-Gln-CTG may degrade THBS2 mRNA and mediate AKT phosphorylation to carry out its biological functions, n = 3 per group. (J–K) tiRNA-Gln-CTG has important effects on the expression of some functional proteins that regulate HTR-8/SVneo cells, n = 3 per group. YBX1, Y-box-binding protein 1; AKT, protein kinase B.

A total of 122 women with a suspected diagnosis of PE were enrolled in the study (Table 1), of whom 28 ultimately developed PE and 94 did not (non-PE group). Women who developed PE had a lower body mass index (BMI) (p = 0.036), delivered earlier (p 0.001), and were more likely to have a cesarean delivery (p 0.001) than the non-PE group. Women who were suspected of having PE due to elevated blood pressure had a higher proportion of ultimately developing PE (p = 0.002). In this study of 122 patients with suspected PE, plasma tiRNA-Gln-CTG levels were used to predict the likelihood of developing PE within the following week (Fig. 10A), the following 4 weeks (Fig. 10B), and before the termination of pregnancy (Fig. 10C). The sensitivity of reduced tiRNA-Gln-CTG for predicting the onset of PE at these times was 91.7%, 85.7%, and 89.3%, respectively, while the specificity was 84.5%, 79.2%, and 73.4% (Fig. 10D).

Fig. 10.

Receiver operating curves were employed to analyze the predictive ability of plasma tiRNA-Gln-CTG for the future occurrence of PE in women with suspected PE. Data are presented as the mean $\pm$ SD. (A) Predictive efficacy of PE occurring within the next 1 week, n_control = 110, n_{pre-eclampsia} = 12. (B) Predictive efficacy of PE occurring within the next 4 weeks, n_control = 101, n_{pre-eclampsia} = 21. (C) Predictive efficacy of PE occurring before the termination of pregnancy, n _control = 94, n_{pre-eclampsia} = 28. (D) Sensitivity and specificity for the prediction of PE within the next 1 week after suspected diagnosis of PE, within the next 4 weeks, and before delivery.