Abstract Down syndrome (DS), a typical chromosomal disease caused by all or part of an extra genomic copy of chromosome 21. Few reports have investigated roles of competing endogenous RNA (ceRNA)-mediated regulatory mechanisms in DS pathogenesis. RNA from PBMCs of cord blood from DS and non-DS fetuses was used for RNA-Seq to profile lncRNAs, miRNAs, and mRNAs. Bioinformatics revealed DS-associated differential gene expression. Predicted miRNA-mRNA/lncRNA interactions were used to build and visualize ceRNA networks in Cytoscape. A total of 216 DEmiRNAs, 651 DElncRNA and 15,789 DEmRNA transcripts were identified in umbilical cord blood PBMC RNA preparations from DS and non-DS control subjects. KEGG pathway enrichment analysis showed that DEmRNAs were involved in pathways such as Huntington’s disease, Alzheimer’s disease, Parkinson’s disease, etc., which are closely related to DS. The 11 mRNAs corresponding to the highest degree PPI network nodes (RPS27A, UBA52, UBC, RPL11, RPS27, MRPS7, RPL23, RPL9, NFKB1, RBX1, and RELA) may play important roles in expression of DS-associated phenotypic characteristics. Finally, we constructed upregulated and downregulated lncRNA-miRNA-mRNA ceRNA sub-networks and found several pairs of ceRNAs that might be involved in DS. MIAT might serve as a ceRNA to sponge hsa-miR-378c and ultimately regulate the expression of RBX1 to affect the cell cycle and lead to DS occurrence. In this study, we comprehensively analysed gene regulatory mechanisms associated with DS progression. The lncRNA-associated ceRNAs identified here may contribute to DS diagnosis and treatment. Supplementary Information The online version contains supplementary material available at 10.1038/s41598-025-20187-3. Keywords: Down syndrome, Competing endogenous RNA (ceRNA), Transcriptome, Bioinformatics analysis, Cell cycle, Umbilical cord blood Subject terms: Computational biology and bioinformatics, Diseases, Genetics, Molecular biology, Neuroscience Introduction Down syndrome (DS) is a classic chromosomal disorder that about 95% of DS subjects possess an extra chromosome 21 and 5% have partially duplicated chromosome segments, and DS occurring with an overall frequency of approximately 1 in 800 births worldwide^[50]1,[51]2. Individuals with DS face various health issues, including intellectual disability, Alzheimer’s disease (AD), leukemia, congenital heart diseases (CHD), etc^[52]3–[53]6. It has been reported that overexpression of genes within the D21S55 region of the proximal region of 21q22.3 plays a major role in DS pathogenesis^[54]7, while genes outside this region also influence DS phenotypic characteristics^[55]8. Many studies have indicated that noncoding RNAs, including long non-coding RNA (lncRNA) and microRNA (miRNA) play important roles in physiological and pathological DS characteristics^[56]9–[57]12. LncRNAs are non-coding RNAs, usually long, that play key roles in physiology and disease, and therefore have great diagnostic and therapeutic potential^[58]13. Indeed, accumulating research shows that lncRNAs can act as molecular sponges of miRNAs to regulate gene expression. In addition, lncRNAs have potential uses as prognostic biomarkers and therapeutic targets for human diseases^[59]14–[60]16. Another class of noncoding RNAs, miRNAs, are small single-stranded non-coding RNAs. They are evolutionally conserved and widely distributed among all species of animals and plants and can suppress gene expression by binding to target mRNAs^[61]17–[62]19. In fact, a single miRNA may have hundreds of potential mRNA targets that regulate multiple biological processes. However, dysregulated miRNA expression can lead to altered expression of hundreds of mRNAs and proteins^[63]20,[64]21, thus, miRNAs may be intimately involved in DS occurrence and progression. Importantly, overexpression of miRNAs in patients with DS leads to a decrease in the expression of specific target proteins, which contributes in part to the DS phenotype^[65]22. For example, miRNAs encoded by genes located on chromosome 21, such as let-7c, mir-99a, mir-155, mir-125b, mir-802, etc., are known to regulate more than 3000 genes. Abnormal expression of these miRNAs may cause multiple DS-like complications, such as Alzheimer’s disease, congenital heart disease and various types of cancer^[66]23. To our knowledge, no studies focusing on DS associations with lncRNA-related ceRNAs have been reported. To further investigate expression of lncRNAs, miRNAs and mRNAs associated with DS phenotypes, expression profiles of these RNAs as expressed in umbilical cord blood PBMCs were compared between DS and non-DS subjects using deep RNA-Seq. After differentially expressed RNAs were identified, RNAs were annotated via bioinformatics analysis in order to assign them to key functional categories. Next, miRNA-mRNA and miRNA-lncRNA interactions were predicted and used to construct co-expression networks. Finally, up-regulated and down-regulated lncRNA-miRNA-mRNA sub-networks were constructed consist of 197 lncRNAs, 5 miRNAs and 4 mRNAs. The identified ceRNA networks may provide a potentially novel understanding for DS mechanisms that could helpful in the diagnosis and future treatment of DS. Materials and methods Cell samples The workflow diagram of this study is shown in Fig. [67]1. The fetus was diagnosed using chromosome G banding (Fig. [68]2). Umbilical cord blood of DS fetus (Karyotyping, 47, XY, + 21) and non-DS fetus (Karyotyping, 46, XY) served as sources of PBMCs for the experimental group and normal control group, respectively. PBMCs were isolated from umbilical cord blood, immediately frozen in liquid nitrogen, then were stored for future use. Subject clinical information and groupings (one DS fetus and one non-DS fetus) are shown in Table [69]1. This study was approved by Ethics Committee of the Shenzhen People’s Hospital, China (LL-KY2019362). All samples were collected from subjects seeking care at Shenzhen People’s Hospital and all participants signed the written informed consent form. Fig. 1. [70]Fig. 1 [71]Open in a new tab Flowchart of the study design. Fig. 2. [72]Fig. 2 [73]Open in a new tab Chromosome G banding of fetus. Non-DS fetus (A) and DS fetus (B). Table 1. Clinical diagnostic information and grouping of all participants. Group Karyotyping Diagnosis Gestational age Maternal age NC 46, XY Normal 16 + 5 40 DS 47, XY, + 21 Down syndrome 17 + 5 37 [74]Open in a new tab RNA isolation, library preparation and sequencing Total RNA from PBMCs isolated from cord blood was extracted using Trizol reagent. miRNA sequencing libraries and lncRNA/mRNA sequencing libraries were constructed, and Assessment of the quality of the repository. Clustering and sequencing were then performed. miRNA library preparation and sequencing produced 50-bp single-end reads, and LncRNA/mRNA library preparation and sequencing produced 150-bp paired-end reads. Methods details are described in Fig. [75]S1. Differential expression analysis The expression levels of miRNAs were quantified using transcripts per million (TPM)^[76]24, as TPM provides an appropriate normalization for short RNA transcripts such as miRNAs. For lncRNAs and mRNAs, expression levels were calculated as fragments per kilobase per million fragments mapped (FPKM)^[77]25, which accounts for transcript length variations typical of long RNAs. Differential expression analysis was performed using the DEGseq R package (version 1.2.2) for miRNAs and the edgeR package (version 3.2.4) for lncRNAs and mRNAs. Significantly differentially expressed RNAs were identified using a false discovery rate (FDR)-corrected p-value threshold of < 0.01 and |log[2].fold change)| > 1. Functional enrichment analysis Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analysis of differentially expressed genes was performed using KOBAS software (version 2.0)^[78]26. Gene Ontology (GO) enrichment analysis was performed using the GO seq R package (Release2.12)^[79]27,[80]28 with annotations from the Ensembl database. Each GO and KEGG pathway term with the corrected p-value < 0.05 was defined as significantly enriched. PPI network analysis Differentially expressed mRNAs were analyzed using the STRING database (version 11.0) with the highest-confidence threshold (0.900) to generate a protein-protein interaction (PPI) network. The resulting network was visualized using Cytoscape (version 3.5.0). CeRNA network analysis With the presence of the same miRNA binding site shared by a set of lncRNAs and mRNAs indicating a potential lncRNA-miRNA-mRNA interaction. First, miRanda (version 3.3a; minimum free energy ≤ -10 kcal/mol), PITA and RNAhybrid (version 2.0) were used to predict miRNA-mRNA pairs. Only the miRNA-mRNA interactions that existed in all databases were included in the ceRNA network. Next, miRanda was used to predict lncRNA-miRNA pairs. According to the ceRNA hypothesis, lncRNA-miRNA and miRNA-mRNA pairs with negative correlations of expression and lncRNAs and mRNAs with positive correlations of expression were selected, with the presence of the same miRNA binding site shared by a set of lncRNAs and mRNAs indicating a potential lncRNA-miRNA-mRNA interaction. Next, we incorporated DEmiRNA-DEmRNA and DEmiRNA-DElncRNA pairs into the ceRNA network and visualised the network using Cytoscape (v. 3.5.0). Results Differential expression analysis To analyse expression profiles of lncRNAs, miRNAs and mRNAs in PBMCs of DS subjects, RNA-sequencing was performed. A total of 216 (88 up-regulated and 128 down-regulated) differentially expressed miRNAs (DEmiRNAs), 651 (154 up-regulated and 497 down-regulated) differentially expressed lncRNA transcripts (DElncRNAs) and 11,733 (3915 up-regulated and 7818 down-regulated) differentially expressed mRNA transcripts (DEmRNAs) were identified (Fig. [81]3). Tables [82]2, [83]3 and [84]4 lists the 10 RNAs with the highest differential expression. Fig. 3. [85]Fig. 3 [86]Open in a new tab Volcano plot for differentially expressed RNA in umbilical cord blood. miRNAs (A), lncRNA (B) and mRNA (C). Table 2. The top 10 differentially expressed mRNAs in umbilical cord blood. miRNA log2.Fold_change. p.value q.value Up-regulated hsa-miR-629-5p 8.7597 2.14E-48 1.28E-47 hsa-miR-4467 6.8387 1.34E-05 3.25E-05 hsa-miR-16-2-3p 6.8031 5.73E-119 4.96E-118 hsa-miR-6747-3p 6.1204 5.31E-04 1.08E-03 hsa-miR-6780a-5p 5.9211 1.17E-03 2.23E-03 hsa-miR-203a-3p 5.0242 7.40E-04 1.47E-03 hsa-miR-203b-5p 5.0242 7.40E-04 1.47E-03 hsa-miR-6735-5p 4.9211 1.15E-03 2.19E-03 hsa-miR-5010-5p 4.8667 9.41E-05 2.17E-04 hsa-miR-6855-5p 4.5925 4.55E-05 1.08E-04 Down-regulated hsa-miR-548ar-3p − 8.7763 1.18E-25 5.69E-25 hsa-miR-1307-5p − 7.6539 8.61E-14 3.15E-13 hsa-miR-376a-3p − 5.6282 1.58E-04 3.51E-04 hsa-miR-136-3p − 5.3398 2.32E-24 1.11E-23 hsa-miR-548d-3p − 5.2261 1.15E-03 2.19E-03 hsa-miR-376c-3p − 4.9423 5.73E-09 1.76E-08 hsa-miR-889-3p − 4.9159 3.11E-74 2.12E-73 hsa-miR-29c-3p − 4.8561 4.89E-03 8.55E-03 hsa-miR-660-3p − 4.811 5.71E-03 9.84E-03 hsa-miR-199b-5p − 4.5912 5.20E-31 2.63E-30 [87]Open in a new tab Table 3. The top 10 differentially expressed LncRNAs in umbilical cord blood. lncRNA log2.Fold_change. p.value q.value Up-regulated LNC_000749 11.16076 8.07E-24 5.50E-22 ENST00000504675.5 10.46388 2.83E-19 1.04E-17 LNC_001005 9.842943 1.57E-15 3.72E-14 ENST00000423311.1 9.358145 7.79E-13 1.27E-11 LNC_001004 8.980293 5.84E-11 7.51E-10 LNC_000579 8.951205 7.60E-11 9.55E-10 LNC_000430 8.796257 5.21E-10 5.88E-09 LNC_000087 8.338042 4.11E-08 3.54E-07 LNC_000723 8.292382 5.79E-08 4.89E-07 ENST00000522525.5 8.292382 5.79E-08 4.89E-07 Down-regulated LNC_000648 − 13.8476 1.16E-41 9.15E-39 ENST00000478845.2 − 11.3148 1.04E-23 6.88E-22 ENST00000424415.1 − 10.9149 4.45E-21 2.09E-19 ENST00000432045.6 − 10.637 2.44E-19 9.12E-18 ENST00000381108.3 − 10.6067 3.75E-19 1.36E-17 LNC_000748 − 10.3524 1.35E-17 4.05E-16 ENST00000423414.5 − 10.3303 1.97E-17 5.79E-16 ENST00000420195.1 − 10.3078 2.54E-17 7.38E-16 ENST00000563635.5 − 10.3003 2.89E-17 8.31E-16 ENST00000451837.5 − 10.0525 7.71E-16 1.88E-14 [88]Open in a new tab Table 4. The top 10 differentially expressed mRNAs in umbilical cord blood. mRNA log2.Fold_change. p.value q.value Up-regulated RPL21 17.29404 8.66E-69 1.57E-64 DDX3X 16.43885 2.28E-62 2.77E-58 HBG1 16.18769 1.75E-60 1.59E-56 CA1 15.67178 1.29E-56 7.79E-53 SDCBP 15.26324 1.46E-53 5.90E-50 TFDP2 15.20514 3.97E-53 1.44E-49 DYNLL1 14.85474 1.64E-50 3.72E-47 UBE2V1 14.82297 2.83E-50 6.04E-47 HMGB1 14.61092 1.08E-48 1.96E-45 RPLP0 14.42177 2.76E-47 4.00E-44 Down-regulated MBNL3 − 18.0817 2.69E-73 9.76E-69 MBNL1 − 15.6343 6.16E-55 2.80E-51 ATF4 − 15.2221 7.38E-52 2.44E-48 TMEM63B − 15.1595 2.16E-51 5.89E-48 OPA1 − 15.1566 2.27E-51 5.89E-48 ESPN − 15.0554 1.29E-50 3.13E-47 PLEC − 14.845 4.75E-49 9.09E-46 PRRC2C − 14.7624 1.95E-48 3.08E-45 WBP2 − 14.4202 6.78E-46 7.95E-43 MACF1 − 14.032 5.06E-43 4.97E-40 [89]Open in a new tab Functional enrichment analysis of DEmRNAs Enrichment analysis of DEmRNAs showed that up-regulated mRNAs were associated with 839 biological processes (BPs) such as cellular process, metabolic process and biological regulation; 182 cellular components (CCs) such as cell part, intracellular part and organelle part; and 124 molecular functions (MFs) such as protein binding, nucleic acid binding and DNA binding (Fig. [90]4A). Meanwhile, the down-regulated mRNAs were associated with 1015 BPs in cellular processes, cellular process regulation and biological process regulation; 158 CCs such as cell part, intracellular part and intracellular organelle; and 189 MFs such as organic cyclic compound binding, nucleic acid binding and small molecule binding (Fig. [91]4B). Fig. 4. [92]Fig. 4 [93]Open in a new tab Functional enrichment analysis of differentially expressed mRNAs. Gene ontology analysis of up-regulated mRNAs (A) and down-regulated mRNAs (B); KEGG pathway enrichment of up-regulated mRNAs (C) and down-regulated mRNAs (D). (We have obtained copyright license from Kanehisa laboratories: [94]https://www.kegg.jp/kegg/kegg1.html) KEGG pathway enrichment analysis of DEmRNAs showed that up-regulated mRNAs were associated with pathways such as Alzheimer’s disease, Parkinson’s disease, Huntington’s disease, Transcriptional misregulation in cancer, Ribosome, Oxidative phosphorylation (Fig. [95]4C). Down-regulated mRNAs were associated with pathways such as NF-kappa B signaling pathway, Osteoclast differentiation, T cell receptor signaling pathway, B cell receptor signaling pathway and Apoptosis (Fig. [96]4D). Down syndrome PPI network analysis DEmRNAs identified using KEGG pathway analysis with corrected p-values of < 0.05 were selected for inclusion in the PPI network that was generated using the STRING database network analysis tool. The network was subsequently visualised using Cytoscape v 3.5.0 (Fig. [97]5A). PPIs with interaction scores of > 0.9 (highest confidence) were selected. Topological parameters of the network are depicted in Table [98]5 and topological characteristics of nodes of the PPI network are illustrated in Table SI. Overall, 11 hub mRNAs (RPS27A, UBA52, UBC, RPL11, RPS27, MRPS7, RPL23, RPL9, NFKB1, RBX1 and RELA) corresponding to the highest degree nodes of the PPI network may play important roles in DS progression (Fig. [99]5B). Using the plug-in MCODE for network analysis led to detection of 4 mRNA clusters (Fig. [100]5C-F). KEGG analysis results showed that cluster 1 mRNAs were associated with Ribosome (Fig. [101]6A), cluster 2 mRNAs were associated with Oocyte meiosis, Cell cycle, Ubiquitin mediated proteolysis, Renal cell carcinoma, etc. (Fig. [102]6B), cluster 3 mRNAs were associated with Retrograde endocannabinoid signaling, Oxidative phosphorylation, Alzheimer’s disease, Parkinson’s disease, Huntington’s disease, etc. (Fig. [103]6C), cluster 4 mRNAs were associated with NF-kappa B signaling pathway, Apoptosis, NOD-like receptor signaling pathway, TNF signaling pathway, IL-17 signaling pathway, etc. (Fig. [104]6D). Fig. 5. [105]Fig. 5 [106]Open in a new tab Protein–protein interaction network analysis of DEmRNAs. (A) PPI network complex analysis of DEmRNAs, with sizes of nodes based on degree scores of DEmRNAs. (B) Top 11 mRNAs in PPI network based on node degree status. (C–F) Significant modules in the PPI network. Cluster 1 consisting of 47 nodes and 1077 edges with an MCODE score of 46.826 (C); cluster 2 consisting of 34 nodes and 561 edges with an MCODE score of 34 (D); cluster 3 consisting of 23 nodes and 249 edges with an MCODE score of 22.636 (E); cluster 4 consisting of 11 nodes and 50 edges with an MCODE score of 10 (F). Circled nodes represent hub mRNAs. Table 5. Topological parameters for the PPI network. Topological parameters Network stats number of nodes 341 number of edges 3463 average node degree 20.3 avg. local clustering coefficient 0.662 expected number of edges 1953 PPI enrichment p-value < 1.0e-16 [107]Open in a new tab Fig. 6. [108]Fig. 6 [109]Open in a new tab KEGG pathway enrichment analysis of MCODE_clusters mRNAs. (A–D) respectively represent KEGG pathway enrichment of clusters 1–4 mRNAs. Prediction of LncRNA–miRNA–mRNA CeRNA network in down syndrome The ceRNA hypothesis suggests that mRNAs and lncRNAs can regulate each other by competitively binding to the same miRNA response elements (MREs) and thus^[110]29. In this study, a total of 13,634 miRNA-lncRNA pairs and 1027 miRNA-mRNA pairs were predicted then used to construct a network based on a total of 93,108 lncRNA-miRNA-mRNA interactions (Table SII). Network analysis revealed that lncRNAs (ENST00000616213.4, ENST00000429829.5 and ENST00000604430.1) regulate multiple miRNAs/mRNAs and thus can be considered key nodes in the ceRNA network. Finally, we constructed up-regulated and down-regulated lncRNA-miRNA-mRNA ceRNA sub-networks that included 197 lncRNAs, 5 miRNAs (hsa-miR-505-5p, hsa-miR-378c, hsa-miR-503-5p, hsa-miR-25-5p and hsa-miR-3184-5p) and 4 mRNAs (RELA, RBX1, RPL18 and RPLP0) (Fig. [111]7). Fig. 7. [112]Fig. 7 [113]Open in a new tab CeRNA sub-networks associated with Down syndrome. Up-regulated (A) and down-regulated (B) lncRNA-miRNA-mRNA ceRNA sub-networks. Nodes denoted with triangle, ellipse and rectangle symbols represent lncRNA, miRNA, miRNA, respectively; red and green colours represent up-regulated and down-regulated RNAs, respectively. Discussion RNA sequencing (RNA-Seq) has become an effective method for systematically identifying all genes and their associated regulatory roles in congenital heart diseases, cancer, Alzheimer’s disease, and other diseases^[114]30–[115]33. The clinical phenotype of DS, clinical phenotype is complex and overlaps with phenotypes associated with various diseases, including neurodegenerative diseases (e.g., Alzheimer’s disease), congenital heart defects, various cancers (e.g., leukemia), etc^[116]23. To better understand DS pathogenesis, in this study high-throughput sequencing of RNA samples from PBMCs in umbilical cord blood of DS and non-DS infants was conducted in order to identify differentially expressed lncRNAs, miRNAs, and mRNAs for the first time. It was worth noting that the DEmRNAs were enriched in pathways such as Alzheimer’s disease, Parkinson’s disease, Huntington’s disease NF-kappa B signaling pathway, Cell cycle, Transcriptional misregulation in cancer, Non-alcoholic fatty liver disease (NAFLD), Oxidative phosphorylation, Ribosome and Ubiquitin mediated proteolysis, which were closely related to Down syndrome. For instance, it has been reported that miRNAs (miRNA-99a, miRNA-125b and miRNA-155) play highly important roles in multiple DS phenotypes under transcriptional control of NF-κB that may participate in the down-regulation of important innate immune regulatory and inflammatory signaling genes^[117]23,[118]34. Moreover, oxidative stress and neuroinflammation are pathological features of neurodegenerative disorders (e.g., Alzheimer’s and Parkinson’s diseases) and DS, with mitochondrial dysfunction possibly playing important roles in progression of all of these diseases^[119]35. Cell cycle regulation may hold the key to understanding DS pathogenesis. In this study we conducted KEGG analysis to determine functions of differentially expressed mRNAs compared between DS and non-DS subjects then constructed a PPI network. Within the network, 4 mRNA clusters and 11 hub mRNAs (RPS27A, UBA52, RPL11, RPS27, MRPS7, RPL23, RPL9, UBC, RBX1, NFKB1 and RELA) were identified and these genes may participate in progression of DS-associated disorders. For example, persistent microglial activation may contribute to neurodegeneration, whereby RPS27A might act as a controller of microglial activation to trigger neurodegenerative disorders^[120]35,[121]36. RPS27A can also promote cell proliferation and regulate the cell cycle during S and G2/M phases^[122]37. Meanwhile, UBA52 regulates ubiquitination of ribosomes and plays an essential role in embryonic development, while UBA52 deficiency leads to reduced protein synthesis that can cause cell cycle arrest^[123]38. RBX1, a component of the E3 ligase complex, is involved in mitochondrial ubiquitination^[124]39 as well as in control over chromosomal stability, where it can restrict the homologous recombination pathway of DNA double-strand break repair during the G1 phase by promoting EXO1 degradation^[125]40,[126]41. Moreover, E3 ligase-mediated ubiquitination of CENP-A lysine 124 can induce CENP-A chromosomal deposition, thus ensuring precise chromosomal segregation^[127]42. Notably, RBX1 silencing would trigger the DNA damage response, G2/M arrest, DNA double-strand breaks, and finally aneuploidy. RELA is an important member of the NF-κB family and plays an important role in DNA binding, in which the DNA binding activity of the NF-κB RELA subunit is dependent on the cofactor protein, since RELA is unable to bind to κB DNA sites in the absence of cofactors such as RPS3 and p53^[128]43,[129]44. Interestingly, expression levels of RELA, TP53 and RPS3 are down-regulated in DS. Importantly, other reports have shown that p53 and its associated regulatory pathways play a critical role in limiting aneuploid reproduction and maintaining the nature of human diploid cells, while TP53 family members and their associated regulatory factors maintain spindle stability during meiosis and embryonic development^[130]45,[131]46. Analysis of a systematically constructed ceRNA network can provide novel insights into mechanisms underlying DS pathology. DS phenotypes are diverse due to complexities associated with disease pathogenesis in that the role of a single gene is very limited. Consequently, we constructed a sophisticated network encompassing a total of 93,108 lncRNA-miRNA-mRNA interactions, including interactions between a single mRNA or lncRNA target and multiple miRNAs, as well as between a single miRNA target and multiple lncRNAs and mRNAs. Analysis of this network revealed that it should be possible to concurrently repress activities of multiple targets to regulate various biological functions associated with DS phenotypes. For example, some LncRNAs in the network (ENST00000616213.4, ENST00000429829.5 and ENST00000604430.1) were found to regulate most DS-associated miRNAs/mRNAs, thus indicating that these LncRNAs could potentially serve as DS biomarkers or therapeutic targets. Specifically, PPI network analysis led to extraction of up-regulated and down-regulated lncRNA-associated ceRNA sub-networks containing 197 lncRNAs, 5 miRNAs and 4 mRNAs. ENST00000616213.4 (MIAT) plays important roles in neurogenic commitment and neuronal survival and has been implicated in cell fate change of progenitors and survival of newborn neurons^[132]47, as well as in neurovascular dysfunction associated with Alzheimer’s disease^[133]48. Moreover, MITA can regulate cell cycle progression by influencing gene expression during the S phase of DNA replication, with MIAT-depleted cells exhibiting reduced regulatory control over numerous genes with direct roles in DNA replication^[134]49. ENST00000429829.5 (XIST) is closely related to X chromosome inactivation such that chromosome silencing induced by this gene can mitigate DS developmental pathology by correcting both DS-associated over-production of megakaryocytes and erythrocytes and the inability of trisomic neural stem cells to differentiate to form neurons^[135]50,[136]51. In summary, our study elucidated lncRNA-associated-ceRNAs in PBMCs isolated from umbilical cord blood of fetus with or without DS using RNA-seq analysis. Our results demonstrated that ceRNAs play important roles in DS development. For instance, MIAT might serve as a ceRNA that sponge hsa-miR-378c to ultimately regulate the expression of RBX1 to affect the cell cycle and lead to DS occurrence. Our study identified several important lncRNAs, miRNAs and mRNAs that may have value as DS biomarkers. Moreover, these biomarkers identified from umbilical cord blood samples should make it easier to translate to the prenatal diagnosis of DS. And we can further explore the expression of these markers in the peripheral blood of pregnant women to evaluate the feasibility of these markers as a non-invasive prenatal diagnosis in further studies. The current study had several limitations. First, our study only investigated putative interactions among lncRNAs, miRNAs and mRNAs yet the lncRNA-miRNA-mRNA ceRNA networks in DS are highly complex as well. Moreover, the samples in this study was insufficient, due to the scarcity of samples. Therefore, it will be necessary to further confirm results reported here by conducting a larger study using a greater number of samples. Conclusion Through RNA-seq and bioinformatics analysis of differentially expressed genes of PBMCs isolated from umbilical cord blood of fetus with and without DS, we identified expression profiles of DEmRNAs, DElncRNAs and DEmiRNAs. Then comprehensively analysis of gene regulation associated with DS progression was performed through analysis of the lncRNA-miRNA-mRNA ceRNA network. Our results further demonstrated that ceRNA regulatory networks participate in DS pathophysiology, with network analysis leading to identification of several important lncRNAs (ENST00000616213.4, ENST00000429829.5 and ENST00000454129.5), miRNAs (hsa-miR-378c, hsa-miR-503-5p, hsa-miR-505-5p, hsa-miR-25-5p and hsa-miR-3184-5p) and mRNAs (RPS27A, UBA52, UBC, RPL11, RPS27, MRPS7, RPL23, RPL9, NFKB1, RBX1 and RELA) with putative important roles in DS occurrence and progression. In a nutshell, our study found several potential biomarkers that may be contribute to DS screening and diagnosis, while also enhancing our understanding of lncRNAs and their molecular biological functional roles in DS pathogenesis. Supplementary Information Below is the link to the electronic supplementary material. [137]Supplementary Material 1^ (109.5KB, docx) [138]Supplementary Material 2^ (52.5KB, xlsx) [139]Supplementary Material 3^ (13.7MB, xlsx) Abbreviations DS Down syndrome PBMCs Peripheral blood mononuclear cells lncRNA Long non-coding RNA miRNA MicroRNA ceRNA Competing endogenous RNA DEmiRNAs Differentially expressed miRNAs DElncRNAs Differentially expressed lncRNA DEmRNAs Differentially expressed mRNA Author contributions All authors contributed to this present work: Zhipeng Zeng, Fengying Zhou, Yaxin Zheng, Donge Tang designed the study. Wei Shi, Wenlong Hu, Xi Wang, Mei Ye acquired the data. Zhipeng Zeng, Fengying Zhou, Yaxin Zheng drafted the manuscript, Jun Zhou, Pingping Ye, Fang Yuan Yaoshuang Zou, Yong Dai revised the manuscript. I certify that all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript. Funding This study was supported by the Sanming project of medicine in Shenzhen (SZSM201812078), the Natural Science Foundation of Guangdong Province (No. 2017A030310629), the Shenzhen Science and Technology Program (No. JCYJ20220530152015035), and the Science and Technology Planning Project of Guangdong Province (No. 2017B020209001), Taizhou Science and Technology Bureau (NO. 23ywa60). Data availability The raw data generated for this study has been deposited in the Genome Sequence Archive (Genomics, Proteomics & Bioinformatics 2021) in National Genomics Data Center (Nucleic Acids Res 2021), China National Center for Bioinformation / Beijing Institute of Genomics, Chinese Academy of Sciences (GSA-Human: HRA001723) and is publicly accessible at [140]https://ngdc.cncb.ac.cn/gsa-human/browse/HRA001723. Declarations Competing interests The authors declare no competing interests. Ethical approval This study was performed according to the declaration of Helsinki and was approved by Ethics Committee of the Shenzhen People’s Hospital, China (LL-KY2019362). All samples were collected in Shenzhen People’s Hospital and all participants had signed the written informed consent. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Zhipeng Zeng, Fengying Zhou and Yaxin Zheng contributed equally to this work. Contributor Information Qiang Yan, Email: yanqiang1967@sina.com. Yong Dai, Email: daiyong22@aliyun.com. Donge Tang, Email: donge66@126.com. References