hnRNPL-activated circANKRD42 back-splicing and circANKRD42-mediated crosstalk of mechanical stiffness and biochemical signal in lung fibrosis

Figure 1. circANKRD42 promoted pulmonary fibrogenesis (A) qRT-PCR assay showed the upregulated circANKRD42 in the blood of 25 patients with IPF compared with the matched normal blood. (B) qRT-PCR assay showed that circANKRD42 was highly expressed in MRC-5 cells treated with TGF-β1 for 0, 24, 48, and 72 h. (C) Migration images were automatically monitored using the Incucyte S3 Live-Cell Analysis System. Compared with the TGF-β1 group, the circANKRD42 knockdown repressed myofibroblast migration at different time points, and the circANKRD42 overexpression group accelerated migration. (D) A real-time cell growth analysis system reflected that the circANKRD4 knockdown inhibited myofibroblast proliferation, and overexpression accelerated proliferation compared with the TGF-β1 group. (E) A real-time cell growth analysis system reflected that the circANKRD4 knockdown inhibited myofibroblast migration, and overexpression accelerated migration compared with the TGF-β1 group. (F) Western blot results showed that the circANKRD4 knockdown decreased α-SMA, vimentin, and collagen I expression levels, and circANKRD4 increased these fibrotic proteins’ expression compared with the TGF-β1 group. GAPDH served…

Figure 2. hnRNP L function as RNA reverse splicing factor promoted circANKRD42 biogenesis (A) Schematic showed the genomic location of circANKRD42 generated from its host gene ANKRD42. Sanger sequencing after PCR with specified primers confirmed “head-to-tail” splicing of circANKRD42, which had a full length of 797 bp. (B) Agarose gel electrophoresis experiment showed that the divergent primers amplified circANKRD42 from cDNA, but not genomic DNA (gDNA). (C) After RNase R treatment, no expression of GAPDH and ANKRD42 genes were detected, but circANKRD42 can be detected. (D) Actinomycin D assay manifested that circANKRD42 was more stable than ANKRD42. (E) circANKRD42 pull-down proteins were subjected to silver staining. Red arrow shows different bands of hnRNP L between sense and antisense lanes. (F) RIP was performed using MRC-5 cell lysate and either anti-hnRNP L or IgG as immunoprecipitation (IP) antibody. Purified RNA from RIP was analyzed through qRT-PCR by using primers specific for the pre-ANKRD42. hnRNP L was observed in the anti-hnRNP L RIP, but not in the IgG RIP. (G) Silencing hnRNP L decreased circANKRD42 expression and increased ANKRD42 expression. Overexpressed hnRNP L…

Figure 3. circANKRD42 sponged miR-324-5p and miR-136-5p to promote pulmonary fibrogenesis (A) RNA FISH detecting the location of endogenous circANKRD42 (red) in cells. Images showed that circANKRD42 existed in the cytoplasm and a little in the nucleus. si-circANKRD42 did not cause circANKRD42 translocation. U6 and 18S RNA were used as nuclear and cytoplasmic localization markers, respectively. DNA (blue) was stained with DAPI. (B) qRT-PCR analysis of RNAs purified from nuclear (red) and cytosolic (gray) compartments in cells. β-actin served as a cytoplasmic control. U6 served as a nuclear control. (C) Dual-luciferase reporter gene assay demonstrated that miR-324-5p and miR-136-5p mimics decreased the luciferase activity. After circANKRD42 mutation, miR-324-5p and miR-136-5p mimics did not cause any change in the luciferase activity compared with the control. WT indicates wild type, and MUT indicates mutation. (D) The RIP experiment verified the binding relationship between circANKRD42 and miR-324-5p/miR-136-5p. (E) qRT-PCR results showed that the expression levels of miR-324-5p and miR-136-5p under 5 ng/mL TGF-β1 treatment gradually decreased from 0 h to 72 h compared with those under the…

Figure 4. Rescue experiments confirmed that the circANKRD42 profibrotic function depended on miR-324-5p and miR-136-5p (A) miR-324-5p and miR-136-5p inhibitors rescued the decreased trends of collagen, vimentin, and α-SMA levels caused by interfering with circANKRD42. (B) miR-324-5p and miR-136-5p mimics rescued the increased trends of collagen, vimentin, and α-SMA levels caused by the circANKRD42 overexpression. (C) miR-324-5p and miR-136-5p inhibitors rescued the decreased trends of myofibroblast proliferation and migration caused by interfering with circANKRD42. (D) miR-324-5p and miR-136-5p mimics rescued the increased trends of myofibroblast proliferation and migration caused by the circANKRD42 overexpression. Each bar represents mean ± SD (n = 6); ∗p < 0.05; ∗∗p < 0.01.

Figure 5. circANKRD42 sponged miR-324-5p and miR-136-5p to target AJUBA and YAP1, respectively (A) The dual-luciferase reporter gene array manifested that AJUBA was the target gene of miR-324-5p and that YAP1 was the target gene of miR-136-5p. (B) The expression levels of AJUBA and YAP1 increased at different time points under the 5 ng/mL TGF-β1 treatment. (C) Western blot detected that miR-324-5p and miR-136-5p mimics reduced the expression of AJUBA and YAP1, respectively. miR-324-5p and miR-136-5p inhibitors increased the expression levels of AJUBA and YAP1, respectively. (D) Western blot results showed that the circANKRD42 overexpression increased the expression levels of AJUBA and YAP1. The knockdown of circANKRD42 reduced the expression levels of AJUBA and YAP1. (E) Immunofluorescence results depicted that p-YAP (red) was located in the cytoplasm in the normal group and that YAP (green) was located in the nucleus in the TGF-β1 group. circANKRD42 knockdown enhanced the p-YAP expression in the cytoplasm, and the circANKRD42 overexpression promoted the YAP expression in the nucleus. DNA (blue) was stained with DAPI. (F) Rescue experiments showed that circANKRD42 knockdown repressed AJUBA…

Figure 6. circANKRD42 enhancement of mechanical stiffness depended on miR-324-5p, miR-136-5p, and YAP1 (A) Young’s modulus value increased in TGF-β1 group compared with the normal group. Knockdown circANKRD42 induced its decrease, and overexpressed circANKRD42 caused its increase. Young’s modulus values of the selected area were 225.8 kPa in normal group, 610.5 kPa in TGF-β1 group, 503.5 kPa in TGF-β1 + si-circRNA NC group, 296.9 kPa in TGF-β1 + si-circRNA group, 204.3 kPa in BP group, and 327.4 kPa in RP group. (B) F-actin was decreased by circANKRD42 knockdown and increased by circANKRD42 overexpression. (C) The F-actin expression was inhibited by miR-324-5p and miR-136-5p mimics and enhanced by miR-324-5p and miR-136-5p inhibitors. (D) The cellular F-actin stained with FITC-phalloidin was observed using fluorescence microscopy. The cytoskeleton tension was alleviated and aggravated by circANKRD42 knockdown and overexpression, respectively. The images of rescue experiments showed that the alleviation of si-circANKRD42 on the cytoskeleton tension was partly reversed by miR-324-5p and miR-136-5p inhibitors and that the aggravation of circANKRD42 overexpression on the cytoskeleton tension…

Figure 7. circANKRD42 was evaluated as a potential therapeutic target in mice (A) Mice inhaled the adenovirus-packed circANKRD42 by spraying. (B) qRT-PCR results demonstrated that overexpressed circANKRD42 was highly expressed in the mouse model. Expression of miR-136-5p and miR-324-5p decreased in overexpressed circANKRD42 group. There was no significant change in the circANKRD42 mutation group compared with the control group. (C) The results of MicroCT imaging system for small animal demonstrated that fibrotic changes occurred in the left lower lung of mice in overexpressed circANKRD42 group. The degree of fibrosis in overexpressed circANKRD42 + BLM group was significantly higher than that in BLM + circANKRD42 NC group. The degree of fibrosis in circANKRD42 mutation + BLM group was similar to that in the BLM + circANKRD42 NC model group and had no change compared with the control group. (D) H&E manifested that the alveolar walls were thickened and filled with pink collagen in the overexpressed circANKRD42 group compared with the BLM group. The degree of fibrosis in overexpressed circANKRD42 + BLM group had become more deteriorated than that in BLM + circANKRD42 NC group. The degree of…

Figure 8. Assessment of circANKRD42 value in IPF patients and its regulatory mechanism (A) circANKRD42 expression was tested from blood of healthy people at different ages by using qRT-PCR method. The results manifested that circANKRD42 expression increased with age. n = 8. (B) ROC curve between circANKRD42 and FVC showed that the sensitivity and specificity values are 60.0% and 93.3% in IPF patients, respectively. The area under the ROC curve was 0.769. Confidence interval (CI): 0.593–0.945. (C) ROC curve between circANKRD42 and DLco showed that the sensitivity and specificity values are 80.0% and 53.3% in IPF patients, respectively. The area under the ROC curve was 0.649. CI: 0.441–0.857. (D) RNA scope detection of circANKRD42 location in IPF patient’s lung samples. Tissues were counterstained with circANKRD42 probes (red) and hematoxylin (blue) and imaged using 20× and 40× objective. Left panel: positive control is shown; right panel: representative images of circANKRD42 are shown. (E) hnRNP L activated the reverse splicing biogenesis of circANKRD42 and the circANKRD42-mediated crosstalk between mechanical force and biochemical signals to drive pulmonary fibrogenesis.