DEGs were associated with inflammation and immunity related functionsThere were 142 DEGs1 (59 up-regulated and 83 down-regulated) between 5 HC and 10 VV samples, 3,563 DEGs2 (2,287 up-regulated and 1,276 down-regulated) between 5 HC and 10 SVT samples, and 4,712 DEGs3 (3,012 up-regulated and 1,700 down-regulated) between 10 VV and 10 SVT samples (Fig. 1A–C). Functionally, it was shown that the 142 DEGs1 were mainly enriched to neutrophil chemotaxis, positive regulation of intrinsic apoptotic signaling pathway, oxygen transport, organic acid binding, antioxidant activity and etc. In addition, DEGs1 were associated with Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways such as histidine metabolism, IL-17 signaling pathway, TGF-beta signaling pathway and etc. (Fig. 1D). It was found that the 3,563 DEGs2 were enriched for gene ontology (GO) functions such as positive regulation of cell adhesion and cytokine production, regulation of small GTPase mediated signal transduction and etc. In addition, DEGs2 were associated with KEGG pathways such as cellular senescence, chemokine signaling pathway, NOD-like receptor signaling pathway and etc. (Fig. 1E). The 4,712 DEGs3 were enriched for GO functions such as regulation of T cell activation, leukocyte cell-cell adhesion, leukocyte migration and etc. And DEGs3 were associated with KEGG pathways such as focal adhesion, protein digestion and absorption, PI3K-Akt signaling pathway and etc. (Fig. 1F). It is worth noting that both DEGs2 and DEGs3 were associated with the rap1 signaling pathway, and all of these DEGs were associated with the functions of inflammation and immunity.Fig. 1Identification of VV-related DEGs and the corresponding enrichment analyses. (A–C) The differential expression landscapes between VV and other clinical samples (Control and SVT). (D–F) Go and KEGG biological analyses of DEGs. VV varicose vein, SVT shallow vein thrombosis, DEGs differentially expressed genes.Immune cell infiltration was significantly different between VV and SVTIn this study, the proportion of mast cells was significantly different between HC and SVT groups, and the proportions of B cells, cytotoxic cells, central memory T cells (Tcm), follicular helper T cells (Tfh) were significantly different between VV and SVT groups (Fig. 2A and B). Moreover, the expressions of marker genes were consistent with above results (Fig. 2C and D).Fig. 2The effects of core DEGs on immune cells in VV and SVT. (A) The heatmap of immune cell proportion between different clinical samples. (B) The differences in the infiltration levels of five immune cells between normal, VV and SVT samples. (C) The expressive heatmap of immune markers between different clinical samples. (D) The differences in the expressions of several immune markers between normal, VV and SVT samples. *P<0.05, **P<0.01, ***P<0.001; NS no significance.Totals of 12 genes were associated with the progression and regression of VVTotals of 46 TGs1 that were associated with occurrence of VV, 2,257 TGs2 that were associated with development of VV, and 12 TGs3 were that associated with the progression and regression of VV were screened (Fig. 3A). Moreover, the TGs1 and TGs3 were associated with the functions of metabolic process, biological regulation and etc. The TGs2 were associated with the functions of immune system process, homeostatic process, regulation of the biological processes and etc. (Fig. 3B–D).Fig. 3Identification of core genes in VV progression. (A) The intersection between three types of DEGs. (B–D) Go and KEGG biological analyses of core genes based on Metascape database. VV varicose vein, SVT shallow vein thrombosis, DEGs differentially expressed genes.B cell and Tcm were the key immune cellsPHF21B, IRS2, DUSP15, NUPR1, AC138969.1, OLFML3, POM121, CCDC124, SOX2 and CADM2 were defined as the KGs1, and KIF20A, SELL, MEFV, POU2F2, METTL7B, PTCRA, CTRC, AURKB, BIRC7 and SKA3 were defined as the KGs2 (Fig. 4A and B). Among them, the expressions of CCDC124 and PHF21B were positively correlated with B cell, the expression of SOX2 was positively correlated with Tcm, and the expression of POM121 was negatively correlated with Tcm (p < 0.05, |cor| > 0.3). All KGs2 were positively correlated with B cell, cytotoxic cell, mast cell, Tfh, and negatively correlated with Tcm (Fig. 4C and D).Fig. 4Clinical and immune values of core genes. (A) Diagnostic accuracy of core genes in VV. (B) Diagnostic accuracy of core genes in the process of VV progressing to SVT. (C,D) The expressive correlations between two types of DEGs and five immune cells. (E) Diagnostic accuracy of core genes in VV and SVT. (F,G) The expressive correlations between six critical genes and five immune cells.Six genes, including PLP2, DACT3, LRRC25, PILRA, MSX1 and APOD were screened with AUC value > 0.8 and defined as the KGs3 (Fig. 4E). The correlation analysis results showed that the expressions of LRRC25 and PILRA were significantly negatively associated with Tcm (p < 0.05, |cor| > 0.3), and significantly positively associated with B cell (p < 0.01, |cor| > 0.3) (Fig. 4F and G).The construction of ceRNA regulatory networkThe ceRNA regulatory network was constructed with 6 KGs3, 92 miRNAs, and 37 lncRNAs. The hsa-miR-107 was capable of simultaneously regulating LRRC25, PILRA, and APOD, while hsa-miR-25-3p could simultaneously regulate PLP2, DACT3, and PILRA, and hsa-miR-5010-5p could simultaneously regulate DACT3, APOD, and PLP2. There were four common miRNAs (hsa-miR-125a-5p, hsa-miR-3126-5p, hsa-miR-514a-5p, and hsa-miR-6875-5p) between LRRC25 and MSX1. There were three common miRNAs between LRRC25 and PILRA (hsa-let-7e-5p, hsa-miR-345-3p and hsa-miR-6763-5p), three common miRNAs between DACT3 and MSX1 (hsa-miR-129-1-3p, hsa-miR-129-2-3p, and hsa-miR-6866-3p), and three common miRNAs between DACT3 and LRRC25 (hsa-miR-378b, hsa-miR-513a-5p and hsa-miR-516b-5p). There were two common miRNAs between APOD and LRRC25 (hsa-let-7d-5p and hsa-miR-103a-3p), two common miRNAs between LRRC25 and PLP2 (hsa-miR-1343-3p and hsa-miR-4525), and two common miRNAs between PLP2 and DACT3 (hsa-miR-378a-3p and hsa-miR-378i). Besides, there was only one miRNA (hsa-miR-1296-5p) between PLP2 and PILRA. It was worth noting that XIST could regulate 24 miRNAs at the same time, NEAT1 could regulate 15 miRNAs at the same time, and MALAT1 could regulate 15 miRNAs at the same time (Fig. 5).Fig. 5The ceRNA-miRNA-mRNA regulatory network of six critical genes.Drug predictionIn this study, the targeted drugs of KGs1, KGs2 and KGs3 were predicted. The results indicated that the targeted drugs of KGs1 included Betamethasone, Phosphate, Cimetidine, Samarium (153Sm) lexidronam, Ethylhexyl methoxycrylene, Sodium acetate, Tetradecyl hydrogen sulfate (ester), Girentuximab I-124, and etc. (Fig. 6A). Degarelix, Diamorphine, Cimetidine, Ursodeoxycholic acid, Fluvastatin, Inositol, Flecainide, Caffeine, Leuprolide and others were the targeted drugs of KGs2 (Fig. 6B). Nitrazepam, Cimetidine, Dactinomycin, Calcitriol, Pilocarpine, Ethylhexyl methoxycrylene and others were the targeted drugs of KGs3 (Fig. 6C; Table 1).Fig. 6The drug prediction of core genes based on DGIdb database. (A) The drug prediction of core differential genes between VV and normal samples. (B) The drug prediction of core differential genes between SVT and normal samples. (C) The drug prediction of core differential genes between VV and SVT samples.Table 1 Drug prediction of core genes in VV and SVT.Expression verificationThe results of qRT-PCR showed that the expression of PLP2, APOD and DACT3 were significantly decreased in the process of occurrence and development of VV (p < 0.05) (Fig. 7). Given that the pivotal functions of DACT3 in cell growth, DACT3 was selected for further investigation.Fig. 7Validation the expressions of core genes in the process of occurrence and development of VV. *P<0.05, **P<0.01, ***P<0.001; NS no significance.DACT3 mediates the onset and progression of VVSimilar to PCR results, the protein expressions of DACT3 in VV were significantly downregulated compared to normal samples, as four pairs of clinical samples determined (Fig. 8A). sh-DACT3 and OE-DACT3 can effectively manipulate DACT3 expressions in HVSMCs (Fig. 8BC). CCK8 assays revealed that overexpression of DACT3 inhibited HVSMCs proliferation, whereas DACT3 deletion suppressed this process (Fig. 8D). As expected, overexpression of DACT3 blocked the transition from the G0/G1 phase to the S phase (Fig. 8E). Quantitative data confirmed above observations (Fig. 8F). The proportion of cell at G0/G1 phase in overexpression group was significantly higher than that in control group, whereas that in silencing group was markedly lower (Fig. 8F).Fig. 8The biofunctions of DACT3 in VV. (A) Western blot detections on 4 pairs of clinical samples for confirming the DACT3 expressions in VV; (B,C) The tests of transfection efficiency via PCR and Western blot assays; (D) The effects of DACT3 on the proliferation of VSMCs as CCK8 assay determined; (E) The effects of DACT3 on the cell cycle of VSMCs as flow cytometric analysis determined; (F) Quantitative analysis of flow cytometry; (G) The effects of DACT3 on the migration of VSMCs as transwell assay determined; (H) Cell counting of transwell migration assay; (I) The effects of DACT3 on the vascular phenotype transition; VV varicose vein, sh-DACT3 short-hairpin target DACT3, OE-DACT3 overexpression of DACT3,VSMCs vascular smooth muscle cells; *P<0.05, **P<0.01, ***P<0.001.Moreover, DACT3 also mediated the vascular migration and phenotypic transition. Transwell assay indicated that overexpression of DACT3 significantly inhibited VSMCs migration, but silencing DACT3 promoted the migrative process (Fig. 8G). The similar trends were observed in cell counting (Fig. 8H). Meanwhile, the expression of OPN, a common marker of the extracellular matrix (ECM) synthesis, was significantly increased in overexpression group compared to other groups (Fig. 8I). However, the expression of SM22α, a classical marker of the VSMCs contraction, was significantly increased in overexpression group (Fig. 8I). Clearly, overexpression of DACT3 can promote the transition from synthetic to contractile phenotype. Collectively, DACT3 was closely involved in the proliferation, cell cycle, migration and phenotype transition of VSMCs, in turn regulating the progression of VV.
