Transfer RNA-derived small RNA tRF-Glu-CTC attenuates neointimal formation via inhibition of fibromodulin

Abstract

Neointimal hyperplasia is a pathological vascular remodeling caused by abnormal proliferation and migration of subintimal vascular smooth muscle cells (VSMCs) following intimal injury. There is increasing evidence that tRNA-derived small RNA (tsRNA) plays an important role in vascular remodeling. The purpose of this study is to search for tsRNAs signature of neointima formation and to explore their potential functions. The balloon injury model of rat common carotid artery was replicated to induce intimal hyperplasia, and the differentially expressed tsRNAs (DE-tsRNAs) in arteries with intimal hyperplasia were screened by small RNA sequencing and tsRNA library. A total of 24 DE-tsRNAs were found in the vessels with intimal hyperplasia by small RNA sequencing. In vitro, tRF-Glu-CTC inhibited the expression of fibromodulin (FMOD) in VSMCs, which is a negative modulator of TGF-β1 activity. tRF-Glu-CTC also increased VSMC proliferation and migration. In vivo experiments showed that inhibition of tRF-Glu-CTC expression after balloon injury of rat carotid artery can reduce the neointimal area. In conclusion, tRF-Glu-CTC expression is increased after vascular injury and inhibits FMOD expression in VSMCs, which influences neointima formation. On the other hand, reducing the expression of tRF-Glu-CTC after vascular injury may be a potential approach to prevent vascular stenosis.

Keywords: Migration; Neointimal hyperplasia (NIH); Proliferation; Transfer RNA (tRNA)-derived small RNAs (tsRNAs); Vascular remodeling; Vascular smooth muscle cell.

Fig. 1

Experimental design and rat carotid artery neointimal formation. A Experimental procedure. First, balloon injury was used in the rat carotid artery to replicate the intimal hyperplasia model. Second, the carotid artery was subjected to small RNA sequencing to identify the differentially expressed tsRNAs (DE-tsRNAs). The expression of partial DE-tsRNAs was verified by qPCR. Then, the potential target mRNAs of the DE-tsRNAs were predicted by bioinformatic analysis. Finally, the expression of the DE-tsRNAs was modulated in vitro and in vivo to determine the function. B Intimal injury of the rat carotid artery was induced by a 2.0F Fogarty catheter. 14 days after surgery, the intima of the carotid artery was thickened and the lumen was narrowed. n = 5. Five sections per sample were measured for neointimal hyperplasia calculations

Fig. 2

Small RNA sequencing. A Heatmap of the correlation coefficient of the carotid artery samples. Blue represents the samples with a high correlation coefficient, and white represents the samples with no similarity. B Primary component analysis. The location of the colored dots shows the main character of the sample. The spatial distance represents the similarity of the samples. C The hierarchical clustering heatmap for tsRNAs. The color in the panel represents the relative expression level. Blue represents an expression level below the mean and red represents an expression level above the mean. D The scatter plot between the sham group and the balloon injury group for tsRNAs. Red dots (upregulated) and green dots (downregulated) indicate ≥ 1.5 fold change between the two groups. Gray dots indicate non-differentially expressed genes. E The volcano plot of tsRNAs. Red (up-regulated) and green (down-regulated) dots indicate statistically significant differentially expressed tRFs & tiRNAs with fold change ≥ 1.5 and P-value ≤ 0.05. F Venn diagram showing the number of tsRNAs expressed in the sham group and balloon injury group. G The subtypes of tsRNAs. The X-axis represents the origin of tRNAs and the Y-axis shows the number of tsRNAs. n = 3. Sh sham, BI balloon injury

Fig. 3

Differentially expressed tsRNAs in the hyperplastic intima of rat carotid artery. A The abundance of tsRNAs and miRNAs was evaluated by their sequencing counts and were normalized as counts per million of total aligned reads (CPM). The top 15 up- and down-regulated tsRNAs in hyperplasic intima were listed based on their statistical significance, CPM, and expression change (fold change). B–E The expression differences of tRF-49:69-chrM.Trp-TCA, tRF-1:29-Glu-CTC-1, tRF-1:29-Gly-GCC-2-M2 and tRF-1:28-Gly-CCC-M2 in the sham surgery group and the balloon injury group were verified by qPCR. Their expression was all upregulated in the balloon injury group except for tRF-1:29-Gly-GCC-2-M2 by qPCR. n = 3, data are expressed as mean ± standard error, *P < 0.05, **P < 0.01, ns indicates for not statistically significant

Fig. 4

Bioinformatic analysis of differentially expressed tsRNAs (DE-tsRNAs) and target gene screening. A Potential target genes of DE-tsRNAs were predicted by miRanda and TargetScan. The clustering map of DE-tsRNAs and target genes was generated by Cytoscape. B The process of screening for target genes that may be involved in intimal hyperplasia. First, the GeneCard database was used to select genes associated with intimal hyperplasia from the predicted list of potential target genes. Then, a GEO dataset GSE164050 (n = 4) was used to screen the genes downregulated in the balloon injury rat model, and 4 genes GLP1R, FMOD, HSPB6 (HSP20), and XBP1 were focused. C According to GSE164050, the mRNA expression of FMOD and HSP20 was decreased in the balloon injury, the expression of XBP1 was undetectable, and the expression change of GLP1R was not statistically significant. D Carotid proteomic detection confirmed that the protein levels of FMOD and HSP20 were decreased in the balloon injury group. n = 3, data are expressed as mean ± SD. *P < 0.05, **P < 0.01; ns indicates not statistically significant

Fig. 5

Structures of tRF-1:29-Gly-GCC-2-M2 and tRF-1:29-Glu-CTC-1 and the interactions with putative target genes. A tRF-1:29-Gly-GCC-2-M2 and tRF-1:29-Glu-CTC-1 are derived from the 5′ ends of tRNA-Gly-GCC and tRNA-Glu-CTC, respectively. In this schematic, bases putatively bound to target genes are highlighted with circles. B According to the sequence prediction of tRF-1:29-Gly-GCC-2-M2 and tRF-1:29-Glu-CTC-1, they may bind to the 3′ untranslated region (UTR) of HSP20 or FMOD

Fig. 6

tRF-Glu-CTC inhibited fibromodulin (FMOD) levels in vascular smooth muscle cells (VSMCs) and upregulated VSMC proliferation and migration. A According to bioinformatics analysis, tRF-Glu-CTC may be a negative regulator of fibromodulin (FMOD) in VSMCs. Therefore, polynucleotide analogs of tRF-Glu-CTC were synthesized and transfected into rat thoracic aortic VSMCs by liposomes. B Compared with the transfection of scrambled RNA sequences, the transfection of tRF-Glu-CTC analog inhibited the level of FMOD in VSMCs. C Transfection of tRF-Glu-CTC increased the proliferation of VSMCs in the EdU incorporation assay. D Transfection of tRF-Glu-CTC increased the migration of VSMCs in the wound healing assay. E Overexpression of tRF-Glu-CTC also increased the migration of VSMCs in the wound healing assay. n = 5, data are expressed as mean ± standard error, *P < 0.05, **P < 0.01

Fig. 7

tRF-Glu-CTC affected the transcription of FMOD. A The 3′UTR of FMOD, which is predicted to interact with tRF-Glu-CTC, was ligated into the pmiR-RB-Reporter vector. B Dual-luciferase reporter assay performed in the HEK 293 cell line showed that the analogs of tRF-Glu-CTC reduced the relative luciferase activity of the expression vector constructed from wild-type FMOD mRNA and the firefly luciferase gene but did not affect the relative luciferase activity of the expression vector constructed from the mutated FMOD mRNA sequence. n = 3, data are expressed as mean ± standard error, **P < 0.01, ns indicates for not statistically significant

Fig. 8

Transfection of tRF-Glu-CTC antisense attenuated neointimal formation after carotid artery balloon injury in rats. A Multipoint injection of chemically modified tRF-Glu-CTC antisense around the arteries on the side of surgery and repeated every 3 days until postoperative day 14 after carotid balloon injury in rats. B Detection of tRF-Glu-CTC expression in rat carotid arteries at 14 days. Compared with the control group injected with scrambled sequence, the expression of tRF-Glu-CTC in the carotid artery of the tRF-Glu-CTC antisense injection group decreased. C The cross section of the rat carotid artery showed that the area of carotid intimal hyperplasia was less in the tRF-Glu-CTC antisense treatment group than in the control group. The intima/media ratio was decreased in the tRF-Glu-CTC antisense treatment group. D Immunohistochemical staining for FMOD on rat carotid arteries showed that rats transfected with tRF-Glu-CTC antisense increased the positive staining of FMOD in the vessel wall after carotid artery intima injury compared with the control group transfected with scrambled RNA sequence. E Western blotting showed the increased level of FMOD in the rat carotid artery transfected with tRF-Glu-CTC antisense after balloon injury. F Transfection of tRF-Glu-CTC antisense decreased the level of Smad3, the major effector of TGF-β1. n = 5, data are expressed as mean ± standard error, *P < 0.05, **P < 0.01

Fig. 9

Mechanism of tRF-Glu-CTC promotion of neointimal hyperplasia. After vascular wall injury, TGF-β1 is produced by the recruitment of inflammatory cells (neutrophils/macrophages, etc.), which is a potent agonist of vascular smooth muscle cells and mediates neointimal formation by promoting proliferation and migration of VSMCs via Smad3. Vascular injury induced the expression of tRF-Glu-CTC, which suppressed fibromodulin (FMOD) levels through the RNA silencing mechanism. FMOD is a TGF-β1 antagonist that inhibits the activity of TGF-β1. Thus, tRF-Glu-CTC played a promoting role in the process of neointima formation by indirectly facilitating the activity of TGF-β1