Maternal Serum tRNA-Derived Fragments (tRFs) as Potential Candidates for Diagnosis of Fetal Congenital Heart Disease

Background: Congenital heart disease (CHD) is one of the most predominant birth defects that causes infant death worldwide. The timely and successful surgical treatment of CHD on newborns after delivery requires accurate detection and reliable diagnosis during pregnancy. However, there are no biomarkers that can serve as an early diagnostic factor for CHD patients. tRNA-derived fragments (tRFs) have been reported to play an important role in the occurrence and progression of numerous diseases, but their roles in CHD remains unknown. Methods: High-throughput sequencing was performed on the peripheral blood of pregnant women with an abnormal fetal heart and a normal fetal heart, and 728 differentially expressed tRFs/tiRNAs were identified, among which the top 18 tRFs/tiRNAs were selected as predictive biomarkers of CHD. Then, a quantitative reverse transcriptase polymerase chain reaction verified the expression of tRFs/tiRNAs in more clinical samples, and the correlation between tRFs/tiRNAs abnormalities and CHD was analyzed. Results: tRF-58:74-Gly-GCC-1 and tiRNA-1:35-Leu-CAG-1-M2 may be promising biomarkers. Through further bioinformatics analysis, we predicted that TRF-58:744-GLy-GCC-1 could induce CHD by influencing biological metabolic processes. Conclusions: Our results provide a theoretical basis for the abnormally expressed tRF-58:74-Gly-GCC-1 in maternal peripheral blood as a new potential biomarker for the accurate diagnosis of CHD during pregnancy.


Introduction
Congenital heart disease (CHD), defined as a general structural abnormality of the heart or large blood vessels present before and/or at birth [1], is the most frequently occurring congenital disorder in newborns and the most common cause of infant death due to birth defects [2,3]. The number of reported births with CHD varies widely by country and region, with the generally accepted best estimate being 6-13 per 1000 live births, and mortality is approximately 40% of prenatal deaths and 20% of deaths in the first year of life [4][5][6][7][8][9][10]. However, the morbidity and mortality of CHD have been greatly reduced with the significant advancement in prenatal and postnatal diagnosis and surgical treatment, such as four-dimensional ultrasonography, fetal cardiac ultrasonography, cardiac MRI, cardiopulmonary bypass technology, etc. However, due to differences in the healthcare system environment, screening professional skills, and available resources, CHD-related examination procedures are often cumbersome, inefficient, and problematic. Moreover, the use of repeat ultrasonography and the application of diagnostic imaging to the potential

Clinical Specimen Collection and Processing
The studies involving human participants were reviewed and approved by the Ethics Committee of Women's Hospital of Nanjing Medical University. All experiments were conducted according to the principles of the Declaration of Helsinki. All patients in this study provided written informed consent. A total of 41 blood samples (N = 41, age = 30.98 ± 3.70) of pregnant women whose fetuses were diagnosed with CHD and 54 blood samples (N = 54, age = 31.62 ± 4.36) of pregnant women with healthy fetal hearts were collected. All CHD diagnoses were verified by fetal echocardiography. Subjects with healthy fetal hearts were recruited from pregnant women participating in prenatal screening. Serum was collected in accordance with Early Detection Research Network operating procedures after obtaining informed consent from patients and healthy individuals. The basic clinical features of the study subjects are shown in Tables 1 and 2. Whole blood was collected at the outpatient clinic and placed in yellow-capped tubes containing citrate dextrose. It was incubated for 15 min at room temperature and after centrifugation at 3000× g for 10 min, serum was transferred to new RNase free tubes. The entire sample processing was performed on ice. Finally, we froze the serum at −80 • C for further experiments [30].  C28  34  G2P1  23W + 6D  NO  YES  C29  28  G1P0  24W + 1D  NO  YES  C30  31  G2P0  23W + 6D  NO  YES  C31  28  G1P0  23W + 6D  NO  YES  C32  27  G1P0  24W + 0D  NO  YES  C35  28  G1P0  23W + 6D  NO  YES   Samples were used for follow-up qRT-PCR   B1  34  --NO  YES  B2  35  --NO  YES  B3  34  --NO  YES  B4  36  --NO  YES  B5  33  --NO  YES  B6  28  --NO  YES  B7  26  --NO  YES  B8  36  --NO  YES  B9  34  --NO  YES  B10 36

RNA Extraction and Preservation
A fixed volume of serum (200 µL per sample) was used for quantitative reverse transcriptase polymerase chain reaction (qRT-PCR) and tRFs/tiRNAs sequencing detection. Following the manufacturer's protocol, total RNA was isolated from patient serum with Trizol LS (Life Technologies, Carlsbad, CA, USA) [31]. Then, we confirmed that all OD 260/280 absorbance ratios for NanoDrop ND-1000 were between 1.8 and 2.0. Finally, store the extracted RNA in a −80 • C freezer for subsequent experiments.

Library Preparation and Sequencing
First, the total RNA samples of 6 pregnant women whose fetal diagnosis of CHD (Table 1) and 6 pregnant women with normal fetal heart (Table 2) were preprocessed. In order to achieve efficient reverse transcription, we removed the RNA modifications that interfered with the construction of the small RNA sequence library as follows: 3 -cP (2 ,3 -cyclic phosphate) removal to 3 -OH for 3 adaptor ligation; 3 -aminoacyl (charged) diacylation to 3 -OH for 3 adaptor ligation; 5 -OH (hydroxyl group) phosphorylation to 5 -P for 5 -adaptor ligation and m1A and m3C demethylation. The synthesized cDNA was amplified with RT primers and amplification primers from Illumina. Then, 135-160 bp PCR-amplified fragments (corresponding to the 15-40 nt small RNA size range [32]) were extracted and purified from PAGE gel. Finally, the prepared library was quantified with an Agilent BioAnalyzer 2100 and sequenced with an Illumina NextSeq 500 [33,34].

Bioinformatic Analysis of tRFs/tiRNAs
Image analysis and base calling were performed using Solexa pipeline v1.8 (Off-Line base Caller software, v1.8). Sequencing quality was investigated by FastQC and trimmed reads were aligned to mature tRNA sequences [34]. We filtered the reads for length < 14 nt or length > 40 nt using cutadapt software [35]. The abundance of tRFs/tiRNAs was assessed by sequencing counts and normalized to counts per million aligned reads (CPM). Next, the expression profiles and differential expression of tRFs/tiRNAs were analyzed by R package EdgeR [36], and tRFs/tiRNAs fold change >1.5 and p < 0.05 were defined as significant differential expression. R software (v4.1.2) was used to construct heat maps, volcano plots, pie charts, scatter plots, and hierarchical clustering of differentially expressed tRFs/tiRNAs for visual identification [37].

Quantitative Real-Time PCR (qRT-PCR) for tRFs/tiRNA
The sequencing results were verified by qRT-PCR, and the expression levels of tRFs/tiRNAs in serum samples of different pregnant women were determined by the Bulge-loop™ miRNA qRT-PCR Starter Kit (Ribo Bio, Guangzhou, China). The specific experimental method was as follows: First, according to the manufacturer's instructions, we mixed RT Primer (1 µL), 5X Reverse-Transcription Buffer (2 µL), RNA template (1 µL), RTase Mix (2 µL), and RNase-free H2O (4 µL). Then, the mixture was incubated at 42 • C for 60 min and then at 70 • C for 10 min to obtain the cDNA. Next, cDNA (1 µL) was added to the qPCR along with RNase-free H2O (3.2 µL), Bulge-loop™ miRNA forward primer (0.4 µL), Bulge-loop™ reverse primer (0.4 µL), and 2X SYBR Green mixture (5 µL). The reactions were conducted in the Roche Light Cycler 480 Real Time PCR System (Roche Light Cycler ® 480 system, Basel, Switzerland), including incubation at 95 • C for 10 min, followed by 40 cycles of 95 • C for 2 s, 60 • C for 20 s, and 70 • C for 20 s. miDETECT™ miRNA external control (cel-miR-39-3p) was the external control. The 2 −∆∆Ct method was used to calculate the expression level of tRFs/tiRNAs in each sample. Details of the tRFs/tiRNAs selected for validation are presented in Table S1.

Receiver Operating Characteristic Analysis
Receiver operating characteristic (ROC) was analyzed using GraphPad Prism 8.0 (GraphPad Software, California, CA, USA) software. The expression level of each selected tRFs/tiRNAs was ranked, showing the patient's status (1 for CHD, 0 for Healthy). Generate an ROC curve by calculating the area under the curve (AUC) using a binomial exact confidence interval [38].

Statistical Analysis
qRT-PCR validation data are presented as mean standard deviation. Data management and analysis were performed using GraphPad Prism 8.0 (GraphPad Software, California, CA, USA). Statistical differences were measured by unpaired Student's t-test. A p-value < 0.05 was considered statistically significant. All experiments were repeated three times to ensure the accuracy of the results.

Catalogue of tRFs/tiRNAs Expressed in Pregnant Women with Fetal Normal Heart and Fetal CHD
High-throughput sequencing technology was used to analyze the expression levels of tRFs/tiRNAs in the disease and control groups. High-throughput sequencing data has been uploaded to the Gene Expression Omnibus (GEO) database, and the accession number is GSE221349 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE221349 (uploaded on 17 December 2022 and acquired accession number on 19 December 2022)). To examine the sequencing quality, the quality (Q) score plot of each sample was plotted ( Figure S1). Q30 means the incorrect base calling probability to be 0.001 or 99.9% base calling accuracy. A Q-score > 30 was considered high-quality data. The correlation coefficient of samples is an important evaluation criterion to measure the reliability and reasoning ability of sample selection. The closer the correlation coefficient is to one sample, the more similar the two comparison samples are. Based on the expression level of each sample, the correlation coefficient between any two samples in all samples was calculated ( Figure 1a). The results showed that the correlation coefficients within the disease group and within the control group were both high, while the correlation coefficients between the groups were both low, indicating good biological replication of the samples. Principal Component Analysis (PCA) is a statistical method for unsupervised analysis to reduce the dimensionality of large datasets and is a useful tool for exploring sample classes based on expressions. The results of PCA showed that the expression profiles of tRFs/tiRNAs were significantly different between samples (Figure 1b). We used bar graphs to represent the read length distribution for each group sequence (Figure 1c,d). As the tRNA decoder has the same anti-codon, a tRFs/tiRNAs with the same sequence may be the cleavage product of different tRNA genes. 728 tRFs/tiRNAs were identified and mapped to 460 unique tRNAs in the two groups (Table S2). The stacked bar graph is used to compare the number of all tRFs/tiRNAs subtypes with the tRNA decoder, and the frequency of subtypes was compared with the length of tRFs/tiRNAs (Figure 1e

Differentially Expressed tRFs/tiRNAs between Pregnant Women with Fetal Normal Heart and Fetal CHD
In this study, out of a total of 728 differentially regulated tRFs/tiRNAs detected, 88 could be found in tRFdb (http://genome.bioch.virginia.edu/trfdb/ (accessed on 10 September 2022)), and the rest were unknown (Figure 2c). The commonly expressed tRFs/tiRNAs represent the CPM values which were more than 20 in both two groups, and the specifically expressed tRFs/tiRNAs represent the CPM values which were more than 20 in one group while less than 20 in the other group. There were 61 tRFs/tiRNAs specifically expressed in group disease and 130 tRFs/tiRNAs specifically expressed in group control (Figure 2d). The pie chart shows the percentage of expression of each isoform of tRFs/tiRNAs in the two groups of samples; we found that the CHD group mainly increased the expression of tRF-5a, tRF-5b and tRF-1, and decreased the expression of tiRNA-5 and tRF-3b and tRF-5c (Figure 2e,f). Then, we analyzed the characteristics of tRFs/tiRNAs expression by drawing heatmap and volcano plots. A heatmap suggested that there were significant differences in the expression of tRFs/tiRNAs between the two groups ( Figure 2a). Under log2FoldChange >1.5 and p value < 0.05 criteria, 66 tRFs/tiRNAs were reported to be dysregulated in group disease compared with group control, of which 28 were up-regulated and 38 were down-regulated. The remaining 662 were excluded because they did not meet the criteria (Figure 2b).  Subsequently, we looked for up-and down-regulated tRFs/tiRNAs in the sequencing results in MINTBASE, and the results showed that five up-regulated tRFs/tiRNAs and 13 down-regulated tRFs/tiRNAs could be looked up in MINTBASE (Table S1). We verified the expression levels of these tRFs/tiRNAs in the peripheral serum of pregnant women (12 disease, 12 control) by qRT-PCR. The qRT-PCR results showed that in the disease group, tRF-58:74-Gly-GCC-1 was up-regulated and tiRNA-1:35-Leu-CAG-1-M2 was downregulated. We were unable to detect statistically significant differences in the expression of other tRFs/tiRNAs (Figure 3c). Then, we proceeded to expand the sample size (35 disease, 48 control) for further validation of tRF-58:74-Gly-GCC-1 and tiRNA-1:35-Leu-CAG-1-M2. The qRT-PCR results showed that tRF-58:74-Gly-GCC-1 was significantly up-regulated and tiRNA-1:35-Leu-CAG-1 M2 was significantly down-regulated (Figure 3d,e); both tRFs/tiRNAs expression levels were consistent with the sequencing data, and their positions on the secondary structure of tRNA are shown in Figure 3a,b. We subdivided the collected samples into different types of CHD, and then compared them separately with those of the healthy group. Interestingly, the qRT-PCR results showed that tRF-58:74-Gly-GCC-1 was significantly upregulated in the VSD group compared with the control group (Figure 3f,g). However, we did not find tRFs/tiRNAs specific expression in other types of congenital heart disease.  Figure 3a,b. We subdivided the collected samples into different types of CHD, and then compared them separately with those of the healthy group. Interestingly, the qRT-PCR results showed that tRF-58:74-Gly-GCC-1 was significantly upregulated in the VSD group compared with the control group (Figure 3f,g). However, we did not find tRFs/tiRNAs specific expression in other types of congenital heart disease.

Potential Targets and Functional Analysis of Target Genes for Predicting Differential Expression of tRFs/tiRNAs
The potential targets for tRF-58:74-Gly-GCC-1 and tiRNA-1:35-Leu-CAG-1-M2 were identified using the TargetScan, miRanda, miRDB, and tRForest. Then, in order to explore the possible function and mechanism of tRF-58:74-Gly-GCC-1 and tiRNA-1:35-Leu-CAG-1-M2 in the occurrence of CHD, we a performed GO and KEGG pathway analysis of target genes. We classified the target genes of tRFs/tiRNAs based on cellular components, molecular functions, and biological processes. The GO annotation showed that the target genes of tRF-58:74-Gly-GCC-1 were abundant in the organophosphate ester transport and cellular modified amino acid metabolic process (Figure 4c). KEGG annotation showed that the target gene of tRF-58:74-Gly-GCC-1 was highly expressed in adrenergic signaling in cardiomyocytes (Figure 4d). Regarding tiRNA-1:35-Leu-CAG-1-M2, GO annotation indicated that its target gene was highly expressed on positive regulation of catabolic process, positive regulation of cellular catabolic process, cell-substrate junction and ubiquitin-like protein transferase activity (Figure 4e). KEGG annotation showed that the target gene of tRF-58:74-Gly-GCC-1 was highly expressed in the regulation of actin cytoskeleton (Figure 4f). These results may explain their role in fetal heart development and the formation of congenital heart defects.

Protein-Protein Interaction (PPI) Network Analysis
To better understand the underlying mechanisms of tRFs/tiRNAs, we analyzed predicted target gene interactions. First, we analyzed the proteins encoded by 101 target genes of tRF-58:74-Gly-GCC-1 and identified the most frequently interacting proteins H2AFX (Figure 5a). H2AFX, a member of the H2A histone family, is essential for checkpoint mediated cell cycle arrest, DNA repair after DNA double strand breaks, in addition to playing a role in DNA damage repair, DNA fragmentation during apoptosis, and is phosphorylated by different kinases in response to apoptotic signals [40][41][42][43][44]. Thus, we speculate that tRF-58:74-Gly-GCC-1 may play a role in CHD pathogenesis by regulating the activation of the H2AFX pathway. Then, by analyzing 76 genes of tiRNA-1:35-Leu-CAG-1-M2, we found that HLTF was the central protein ( Figure 5b). HLTF has chromatin remodeling activity, resulting in the displacement of DNA-binding proteins on stalled replication forks and promoting DNA damage repair. HLTF was specifically expressed in the heart during early embryonic development [45]. Therefore, we speculate that tiRNA-1:35-Leu-CAG-1-M2 and tiRNA-1:35-Leu-CAG-1-M2, may play a role in CHD pathogenesis by regulating the activation of the HLTF pathway.  (Figure 4d). Regarding tiRNA-1:35-Leu-CAG-1-M2, GO annotation indicated that its target gene was highly expressed on positive regulation of catabolic process, positive regulation of cellular catabolic process, cell-substrate junction and ubiquitin-like protein transferase activity (Figure 4e). KEGG annotation showed that the target gene of tRF-58:74-Gly-GCC-1 was highly expressed in the regulation of actin cytoskeleton ( Figure  4f). These results may explain their role in fetal heart development and the formation of congenital heart defects.

Protein-Protein Interaction (PPI) Network Analysis
To better understand the underlying mechanisms of tRFs/tiRNAs, we analyzed predicted target gene interactions. First, we analyzed the proteins encoded by 101 target genes of tRF-58:74-Gly-GCC-1 and identified the most frequently interacting proteins H2AFX (Figure 5a). H2AFX, a member of the H2A histone family, is essential for checkpoint mediated cell cycle arrest, DNA repair after DNA double strand breaks, in addition to playing a role in DNA damage repair, DNA fragmentation during apoptosis, and is phosphorylated by different kinases in response to apoptotic signals [40][41][42][43][44]. Thus, we speculate that tRF-58:74-Gly-GCC-1 may play a role in CHD pathogenesis by regulating the activation of the H2AFX pathway. Then, by analyzing 76 genes of tiRNA-1:35-Leu-CAG-1-M2, we found that HLTF was the central protein ( Figure 5b). HLTF has chromatin remodeling activity, resulting in the displacement of DNA-binding proteins on stalled replication forks and promoting DNA damage repair. HLTF was specifically expressed in the heart during early embryonic development [45]. Therefore, we speculate that tiRNA-1:35-Leu-CAG-1-M2 and tiRNA-1:35-Leu-CAG-1-M2, may play a role in CHD pathogenesis by regulating the activation of the HLTF pathway.

Discussion
As the most common congenital defective disorder in newborns, the pathogenesis of CHD is not yet clear [46]. With the continuous improvement of medical conditions and prenatal diagnostic techniques, the vast majority of CHDs can be detected during pregnancy, but it cannot be excluded that there are still some CHDs that cannot be detected throughout pregnancy. In the clinic, some simple types of CHD, such as atrial septal defects, patent ductus arteriosus, and ventricular septal defects, can be treated surgically after birth and have a good prognosis [47]. However, if there is no early detection and

Discussion
As the most common congenital defective disorder in newborns, the pathogenesis of CHD is not yet clear [46]. With the continuous improvement of medical conditions and prenatal diagnostic techniques, the vast majority of CHDs can be detected during pregnancy, but it cannot be excluded that there are still some CHDs that cannot be detected throughout pregnancy. In the clinic, some simple types of CHD, such as atrial septal defects, patent ductus arteriosus, and ventricular septal defects, can be treated surgically after birth and have a good prognosis [47]. However, if there is no early detection and treatment of some complex types of CHD, such as hypoplastic left heart syndrome, even after multiple surgeries, the prognosis is poor [48]. Therefore, if we had a serum biomarker for more comprehensive screening and diagnosis of CHD from a different perspective based on routine cardiac ultrasonography, this would largely reduce the prevalence and mortality of CHD. Although many efforts have been made to identify biomarkers for CHD, there are currently no clear biomarkers in the clinic.
With the development of sequencing technology, a new class of ncRNA-tRFs has entered the scene of diagnostic and prognostic biomarkers. tRFs/tiRNAs are an important class of regulatory molecules that play an integral role in cellular functions, such as proliferation, differentiation, apoptosis, and biological processes, such as RNA modification, DNA damage repair, and regulation of gene expression, as well as in the development of many diseases [17,20,[49][50][51]. tRFs/tiRNAs are also involved in pathological processes such as viral infection, tumorigenesis, and neurodegeneration. Given that tRFs/tiRNAs are widespread in nature and their expression is closely related to functions such as cell proliferation, tRFs/tiRNAs may be useful biomarkers in some cases. Therefore, we examined the expression of tRFs/tiRNAs in the serum of pregnant women and performed a biological analysis to verify whether tRFs/tiRNAs in the serum of pregnant women could be used as potential biomarkers for CHD.
In this study, we first performed high-throughput sequencing of peripheral serum from pregnant women with fetal heart defects and normal fetal hearts. Based on the sequencing results, we selected the top 18 abnormally expressed tRFs/tiRNAs for qRT-PCR to verify the authenticity of the data. Then, the most up-or down-regulated and statistically significant tRFs/tiRNAs in serum samples from two groups of pregnant women were tRF-58:74-Gly-GCC-1 and tiRNA-1:35-Leu-CAG-1-M2, respectively. The qRT-PCR results showed that tRF-58:74-Gly-GCC-1 was significantly up-regulated in the disease group and tiRNA-1:35-Leu-CAG-1-M2 was down-regulated in the disease group compared to the control group, which was consistent with the sequencing results. Moreover, we unexpectedly found that tRF-58:74-Gly-GCC-1 was specifically expressed in the VSD group. However, abnormal tRFs/tiRNAs expression was not found in other types of CHD in clinical samples. Therefore, we conclude that tRF-58:74-Gly-GCC-1 plays a crucial role in the occurrence and development of ventricular septal defect. The possible reason for our failure to find differentially expressed in tRFs/tiRNAs other types of CHD is the small number of clinical samples. To further understand the underlying mechanism of tRFs/tiRNAs in the development of CHD, we performed GO and KEGG pathway analysis. The results showed that tRF-58:74-Gly-GCC-1 was enriched for target genes in biological processes and adrenergic signaling in cardiomyocytes signaling pathways in cardiomyocytes. Currently, a large body of evidence suggests that CHD occurs as a result of the accumulation of environmental risk factors and an individual's factors, and that exobiotic metabolic processes such as the elimination of environmental teratogens play an important role in the development and progression of CHD [48,52,53]. Therefore, tRF-58:74-Gly-GCC-1 may be involved in the biological processes leading to the development of CHD. For tiRNA-1:35-Leu-CAG-1-M2, we found that its target genes were mainly enriched in positive regulation of catabolic and positive regulation of cellular catabolic processes, and therefore, tiRNA-1:35-Leu-CAG-1-M2 plays an important role in normal cell growth and development. Since these differential tRFs/tiRNAs biological processes are closely related to the development of CHD, it is particularly important to further explore the relevant molecular mechanisms. Finally, we constructed a PPI network with all target genes predicted by biological analysis in order to further explain the potential mechanisms of tRFs/tiRNAs involvement in CHD development. The results showed that H2AFX and HLTF were the central proteins of tRF-58:74-Gly-GCC-1 and tiRNA-1:35-Leu-CAG-1-M2, respectively. H2AFX plays an important role in DNA damage repair and DNA fragmentation during apoptosis. HLTF also promotes DNA damage repair. Therefore, we conclude that tRF-58:74-Gly-GCC-1 and tiRNA-1:35-Leu-CAG-1-M2 could play a crucial role in DNA damage repair and cell growth and development. In conclusion, our findings suggest that tRF-58:74-Gly-GCC-1 and tiRNA-1:35-Leu-CAG-1-M2 in the peripheral serum of pregnant women are most likely to serve as promising biomarker candidates.
The onset and progress of CHD, mainly in early pregnancy, is a complex and relatively long-term process involving many genetic and epigenetic alterations. To date, there are quite a few types of CHD. Therefore, we need to continue to expand the sample size and collect more detailed pathological features to further identify and analyze different types of CHD. The relationship between the change of tRFs/tiRNAs expression level and the time of pregnancy, the causal relationship between tRFs/tiRNAs expression level and the occurrence of CHD, as well as how the potential downstream targets predicted in this study regulate the biological process also need to be further studied.

Conclusions
In summary, our study has performed a comprehensive analysis of tRFs/tiRNAs in the sera of pregnant women with fetal heart defects and those with healthy fetal hearts, suggesting that tRF-58:74-Gly-GCC-1, tiRNA-1:35-Leu-CAG-1-M2 may play a significant role in fetal heart development. After further bioinformatics analysis, we could identify tRF-58:74-Gly-GCC-1 as a potential biomarker for CHD. We believe our study could provide insight to further investigate the mechanism of tRFs/tiRNAs in CHD and utilize serum factor for predictive and preventive clinical diagnosis.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jcdd10020078/s1. Table S1: Details of the tRFs/tiRNAs selected for validation. Table S2: Results of high-throughput sequencing. Figure S1: The quality (Q) score plot of each sample.

Institutional Review Board Statement:
The study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of Women's Hospital of Nanjing Medical University (protocol code: 2020KY072 and date of approval: 26 November 2020).

Informed Consent Statement:
Informed consent was obtained from all involved in the study.

Data Availability Statement:
High-throughput sequencing data has been uploaded to the GEO database, and the accession number is GSE221349 (https://www.ncbi.nlm.nih.gov/geo/query/ acc.cgi?acc=GSE221349 (uploaded on 17 December 2022 and acquired accession number on 19 December 2022)).