lncITPF Promotes Pulmonary Fibrosis by Targeting hnRNP-L Depending on Its Host Gene ITGBL1

Abstract

The role of long non-coding RNA (lncRNA) in idiopathic pulmonary fibrosis (IPF) is poorly understood. We found a novel lncRNA-ITPF that was upregulated in IPF. Bioinformatics and in vitro translation verified that lncITPF is an actual lncRNA, and its conservation is in evolution. Northern blot and rapid amplification of complementary DNA ends were used to analyze the full-length sequence of lncITPF. RNA fluorescence in situ hybridization and nucleocytoplasmic separation demonstrated that lncITPF was mainly located in the nucleus. RNA sequencing, chromatin immunoprecipitation (ChIP)-qPCR, CRISPR-Cas9 technology, and promoter activity analysis showed that the fibrotic function of lncITPF depends on its host gene integrin β-like 1 (ITGBL1), but they did not share the same promoter and were not co-transcribed. Luciferase activity, pathway inhibitors, and ChIP-qPCR showed that smad2/3 binds to the lncITPF promoter, and TGF-β1-smad2/3 was the upstream inducer of the fibrotic pathway. Furthermore, RNA-protein pull-down, liquid chromatography-mass spectrometry (LC-MS), and protein-RNA immunoprecipitation showed that lncITPF regulated H3 and H4 histone acetylation in the ITGBL1 promoter by targeting heterogeneous nuclear ribonucleoprotein L. Finally, sh-lncITPF was used to evaluate the therapeutic effect of lncITPF. Clinical analysis showed that lncITPF is associated with the clinicopathological features of IPF patients. Our findings provide a therapeutic target or diagnostic biomarker for IPF.

Keywords: CRISPR-Cas9 technology; IPF; ITGBL1; RNA-protein pull-down; RNA-sequencing; TGF-β1-smad2/3; lncITPF; lncRNA.

Figure 1

lncITPF Verification, Expression, and Conservation (A) In vitro translation assay showed that lncITPF did not exhibit translation activities. The black arrow indicates that the positive control translated a 75-kDa protein. (B) The full-length sequence of lncITPF in the human genome was analyzed via RACE. (C) The size of lncITPF was detected via northern blot and was close to the 1,000-bp length of the human genome. (D) lncITPF maps to chromosome 13 and contains introns and exons of the ITGBL1 gene in the human genome. (E) lncITPF was upregulated in MRC-5 cells treated with 5 ng/mL TGF-β1 for 12, 24, 48, and 72 hr. Each bar represents the mean ± SD; n = 6; **p < 0.01.

Figure 2

lncITPF Promoted Pulmonary Fibrosis (A) α-SMA expression was decreased by lncITPF RNAi and increased by lncITPF overexpression (lncITPF RP). Typical α-SMA staining revealed that cytoplasmic fibers partially co-localized with cell skeleton F-actin. F-actin co-localization was used to validate the α-SMA. α-SMA around the nuclear membrane was stained with FITC (green). The nuclei were counterstained with Hoechst 33258 (blue). F-actin was stained red. (B) Western blot showed that lncITPF RNAi decreased α-SMA, collagen, and vimentin expression levels, whereas lncITPF overexpression increased α-SMA, collagen, and vimentin expression levels. (C) Real-time migration analysis system showed that lncITPF RNAi decreased migration, whereas lncITPF overexpression increased migration. GAPDH served as the control. NC indicates a negative control, BP indicates blank plasmid, and RP indicates the recombinant plasmid of the overexpressed lncITPF. Each bar represents the mean ± SD; n = 6; **p < 0.01.

Figure 3

Regulatory Mechanism of lncITPF in Its Host Gene (A) Itgbl1 expression was higher in fibrotic tissue than in normal tissue on the basis of RNA sequencing. (B) Itgbl1 was upregulated in cells treated with 5 ng/mL TGF-β1 for 12, 24, 48, and 72 hr. (C) qRT-PCR analysis showed that Itgbl1 was decreased by lncITPF RNAi and increased by lncITPF overexpression. (D) Western blot showed that ITGBL1 was decreased by lncITPF RNAi and increased by lncITPF overexpression. (E) Single-molecule RNA-FISH detecting the location of lncITPF (red) in cells. U6 and 18S RNA were used as cytoplasmic and nuclear localization markers, respectively. DNA (blue) was stained with DAPI. A representative image is shown. (F) qRT-PCR analysis of RNAs purified from nucleoplasmic (red), chromatin nuclear (gray), and cytosolic (blue) compartments in cells. NC indicates a negative control, BP indicates blank plasmid, and RP indicates the recombinant plasmid of the overexpressed lncITPF. Each bar represents the mean ± SD; n = 6; **p < 0.01 and ***p < 0.001 versus the control group.

Figure 4

Regulatory Mechanism of lncITPF in Its Host Gene ChIP analysis of the level of H3 and H4 histone acetylation of ITGBL1 (A) and GAPDH (B) promoter regions using anti-acetyl-histone H3 and H4. Enrichment was determined relative to input controls. GAPDH was used as the control. (C) Western blot showed that ITGBL1 RNAi decreased α-SMA, vimentin, and collagen expression levels. GAPDH served as the control. (D) Real-time migration system showed that ITGBL1 RNAi decreased the migration ability of cells. Migration was monitored in the xCELLigence DP system. (E) Vimentin and collagen were increased by ITGBL1 overexpression. (F) Migration was increased by ITGBL1 overexpression. (G) Rescue experiment showed that the suppression or induction of lncITPF on migration was partly reversed by ITGBL1. (H) Rescue experiment showed that the suppression or induction of collagen and vimentin expression levels by lncITPF RNAi or lncITPF overexpression was partly reversed by ITGBL1. NC indicates a negative control, BP indicates blank plasmid, and RP indicates the recombinant plasmid of the overexpressed lncITPF or ITGBL1. Each bar represents the mean ± SD; n = 6; *p < 0.05 and **p < 0.01.

Figure 5

Upstream Mechanism of lncITPF (A) TGF-β1 increased the lncITPF promoter activity. Cells transfected with pGL3-lncITPF vectors displayed significantly higher luciferase activity than empty pGL3. Similarly, cells treated with TGF-β1 displayed significantly higher luciferase activity than those without. Cells were transfected with pGL3-lncITPF vectors or an empty vector with or without TGF-β1 treatment for 48 and 72 hr. Luciferase activity was normalized to renilla. (B) Result of Cas9 technology showed that lncITPF had its own promoter located inside the Itgbl1 gene, and they could not be co-transcribed. (C) Signal pathway inhibitors were used to detect the change in lncITPF expression. Only the inhibitor SB431542 of the smad2/3 pathway blocked the expression of lncITPF. (D) ChIP-qPCR analysis of the binding of smad2/3 at the lncITPF promoter region with or without TGF-β1 treatment. ChIP-purified DNA targeting of the indicated genes was analyzed by qPCR. L1 and L2 indicated the different sequences of lncITPF. (1) MRC-5 input; (2) MRC-5 chip DNA; (3) MRC-5 input treated with TGF-β1; (4) MRC-5 chip DNA treated with TGF-β1. (E) The ChIP-qPCR gel images were quantified and analyzed. Each bar represents the mean ± SD; n = 6; *p < 0.05, **p < 0.01, and ***p < 0.001.

Figure 6

Downstream Mechanism of lncITPF (A) SDS-PAGE analysis of proteins purified from in vitro binding assay using biotinylated lncITPF or antisense control RNA. The arrow indicates that the highlighted protein bands were subjected to LC-MS. (B) Protein levels of hnRNP-L in immunoprecipitates with lncITPF RNA were evaluated by western blot. (C) RNA immunoprecipitation assays followed by qRT-PCR analysis of co-purified RNAs with hnRNP-L in cross-linked cells. lncITPF RNA expression levels are presented as fold enrichment values relative to IgG immunoprecipitates. (D) Overexpression of lncITPF increased Itgbl1 mRNA expression, which was impaired in hnRNP-L knockdown. (E) Overexpression of lncITPF increased ITGBL1 protein expression, which was impaired in hnRNP-L knockdown. (F) ChIP-qPCR analysis of hnRNP-L occupancy in ITGBL1 promoter in cells. (G) The ChIP-qPCR gel images were quantified and analyzed. (H) ChIP analysis of the effect of hnRNP-L on H3 and H4 histone acetylation of ITGBL1 promoter regions. NC indicates a negative control, BP indicates blank plasmid, and RP indicates the recombinant plasmid of the overexpressed lncITPF. Each bar represents the mean ± SD; n = 6; *p < 0.05 and **p < 0.01.

Figure 7

Knockdown of lncITPF Suppressed the Process of Lung Fibrosis in Mice (A) The inference of lncITPF in mice was detected via qRT-PCR. (B) Body weight monitoring showed that sh-lncITPF blocked the lost body mass compared with the control mice. (C) The FVC result showed that sh-lncITPF improved the lung function of mice compared with the control group. (D) Histological images of lung paraffin sections stained with H&E. Collagen deposition in the lung tissue was detected by Masson’s trichrome staining. Scale bars, 20 μm. (E) Representative immunohistochemical staining of α-SMA in serial lung sections. The brown color indicates positive staining. (F) HYP content was decreased by sh-lncITPF. Fibrotic markers, including mesenchymal markers and epithelial cell markers, such as collagen and vimentin (G), E-cadherin, SP-C, α-SMA, and Snail (H) in lung tissue tested via western blot. Each bar represents the mean ± SD; n = 6; **p < 0.01.

Figure 8

Assessment of lncITPF Value in IPF Patients and Its Regulatory Mechanism (A) Expression levels of lncITPF increased in 76 patients with IPF, as detected by qRT-PCR. Normal samples corresponding to the sex and age of patients with IPF were selected. (B) Negative correlation between lncITPF and FVC analyzed through Pearson’s correlation. (C) lncITPF expression increased with age. This population comprised a group of healthy volunteers. (D) ROC curve of lncITPF shows that the sensitivity and specificity values are 64.3 and 95, respectively, in IPF patients. (E) Regulatory mechanism of lncITPF. Each bar represents the mean ± SD; *p < 0.05, **p < 0.01, and ***p < 0.001.