Radiation-sensitive circRNA promotes intestinal regeneration

Abstract

Background: The intestine is one of the most sensitive organs to ionizing radiation (IR), and radiation-induced intestinal injury (RIII) impacts the quality of life of patients undergoing radiotherapy. There are limited early diagnostic biomarkers and specific medicines clinically approved for RIII. Therefore, we sought to identify new theranostic targets to prevent RIII and to facilitate the reestablishment of the intestinal epithelium. Circular RNAs (circRNAs) are widely appreciated as pervasive regulators of many diseases and multiple biological processes, while whether and how specific circRNAs are involved in radiation-induced intestinal injury remains largely unknown.

Methods: Differentially expressed circRNAs were analyzed and verified via RNA sequencing. The function of an intestine-specific circRNA (termed circDmbt1(3,4,5,6)) on cell proliferation, apoptosis, and DNA damage level after radiation was explored in vitro, and the underlying mechanism was further investigated. Ultimately, intestinal organoids and mice model were used to verify the role of circDmbt1(3,4,5,6) on radiation-induced intestinal injury.

Results: Primarily expressed in intestinal stem cells, CircDmbt1(3,4,5,6) was downregulated in mice intestines after 14 Gy abdominal radiation and showed timely relationship with intestinal injury level. CircDmbt1(3,4,5,6) promoted the proliferation and alleviated cell apoptosis and DNA damage level of intestinal epithelial cells and promoted organoids survival after radiation compared with control groups. In vivo experiments showed that compared with control groups, overexpression of circDmbt1(3,4,5,6) could increase intestinal length; enhance epithelial integrity and the percentage of proliferative cells, stem cells, paneth cells, and goblet cells; and promote intestinal adaption after radiation. Mechanistically, circDmbt1(3,4,5,6) protects intestines from IR via circDmbt1(3,4,5,6)/miR-125a-5p/STAT3.

Conclusions: CircDmbt1(3,4,5,6), a novel promising RIII bio-marker, responses rapidly at the early stage after 14 Gy abdominal irradiation, and exogenous expression of circDmbt1(3,4,5,6) could promote intestinal fitness in RIII. We reveal that the circDmbt1(3,4,5,6)/miR-125a-5p/STAT3 axis is important to the regeneration of the intestinal epithelium after radiation-induced damage, providing a potential diagnostic and therapeutic target for RIII.

Keywords: circDmbt1(3,4,5,6); Intestinal stem cell; Radiation-induced intestinal injury; STAT3; miR-125a-5p.

Fig. 1

The identification and characteristics of the novel circRNA named circDmbt1(3,4,5,6). A schematic diagram of the mouse model of RIII. B Hematoxylin and eosin staining of sections from the jejuna of control group mice (NR) and from ABI-treated mice (IR 2 days and IR 3.5 days). C Volcano plots of differentially expressed circRNAs between control (NR) mice and mice 2 days after ABI (IR 2d). D RT–qPCR analysis of circDmbt1(3,4,5,6) in jejuna and crypts of control mice (NR) and of ABI-treated mice (IR 2d), N = 6. E Schematic illustration indicating the formation of circDmbt1(3,4,5,6) via circularization of exons 3, 4, 5, and 6 from its host gene Dmbt1 on chromosome 7 and Sanger sequencing of the RT–PCR products of circDmbt1(3,4,5,6) were used to verify the site of back-splicing. The black arrow indicates the back-splice junction (BSJ). F cDNA and gDNA were amplified with convergent (ponit-to-point triangles) and divergent (back-to-back triangles) primers, respectively. β-actin served as negative control. G RT–PCR was used to measure the abundance of circDmbt1(3,4,5,6) and its host gene mRNA in intestinal tissues treated with RNase R. H RT–qPCR assay of circDmbt1(3,4,5,6) and Dmbt1 expression in circDmbt1(3,4,5,6)-overexpressing CT26 cells following treatment with the transcriptional inhibitor actinomycin D for 0 h, 2 h, 4 h, and 10 h, N = 3. I RT–qPCR was used to test the expression of circDmbt1(3,4,5,6) in intestinal villi, crypts, and remnants, N = 3. J RNA FISH analysis of the sub-tissue localization of circDmbt1(3,4,5,6) in mouse intestinal paraffin sections. K RT–qPCR analysis of circDmbt1(3,4,5,6) in Lgr5⁻ and Lgr5⁺ intestinal epithelial cells, N = 3. Bars represent mean ± standard error of the mean (SEM). d days; *P < 0.05 and ***P < 0.001

Fig. 2

circDmbt1(3,4,5,6) alleviates radiation-induced damage in intestinal epithelial cells. A EdU staining was used to analyze cell proliferation of circDmbt1(3,4,5,6)-overexpressing and control MODE-K cells after IR, N = 2. B EdU staining was used to analyze cell proliferation of circDmbt1(3,4,5,6)-knockdown (si-circDmbt1(3,4,5,6)) and control (siNC) MODE-K cells after IR, N = 3. C Flow cytometry was used to detect apoptosis rates of circDmbt1(3,4,5,6)-overexpressing and control MODE-K cells after IR, N = 3. D Flow cytometry assays were used to detect apoptosis rates of circDmbt1(3,4,5,6)-knockdown and control MODE-K cells after IR, N = 3. E Representative images of comet assays (left) and quantitative determination of Olive tail moment (right) of circDmbt1(3,4,5,6)-overexpressing and control MODE-K cells after IR, n = 100. F Representative images of comet assays (left) and quantitative determination of Olive tail moment (right) of circDmbt1(3,4,5,6)-knockdown and control MODE-K cells after IR, n = 100. G Representative images of γ-H2AX immunofluorescence (left) and determination of the numbers of γ-H2AX foci (right) in circDmbt1(3,4,5,6)-overexpressing and control MODE-K cells after IR, n = 50. H Representative images of γ-H2AX immunofluorescence (left) and determination of the numbers of γ-H2AX foci (right) in circDmbt1(3,4,5,6)-knockdown and control MODE-K cells after IR, n = 100. Bars represent mean ± SEM. d days; *P < 0.05, **P < 0.01 and ***P < 0.001

Fig. 3

Ability of circDmbt1(3,4,5,6) to function as a sponge of miR-125a-5p. A RNA FISH analysis of the sub-cellular location of circDmbt1(3,4,5,6) in circDmbt1(3,4,5,6)-overexpressing CT26 cells. B RT–qPCR analysis for RNA immunoprecipitation products of circDmbt1(3,4,5,6) in an anti-AGO2 immunoprecipitated fromcircDmbt1(3,4,5,6)-overexpressing 293 T cells. C Lysates prepared from mouse jejunal crypt cells were incubated with RNA enriched from circDmbt1(3,4,5,6)-overexpressing 293 T cells by incubating with biotinylated probes against circDmbt1(3,4,5,6). An RNA pulldown assay was then performed, and immunoblotting analyses were performed with anti-AGO2 antibody. Pulldown+ refers to the sample being pulled down with the biotin-labeled circDmbt1(3,4,5,6) probe and pulldown R refers to the sample being pulled down with the negative control probe. D The sequences of circDmbt1(3,4,5,6) containing the miR-125a-5p binding sites or mutant binding sites are shown. E Lysates prepared from mouse jejunal crypt cells were incubated with RNA enriched from circDmbt1(3,4,5,6)-overexpressing CT26 cells by incubating with biotinylated probes against circDmbt1(3,4,5,6). RT–qPCR was used to determine miR-125a-5p expression. F Dual-luciferase reporter assay was used to detect the luciferase activities of pmirGLO-NC, circDmbt1(3,4,5,6)-WT, and circDmbt1(3,4,5,6)-MUT luciferase reporter plasmids to evaluate the interaction between miR-125a-5p and circDmbt1(3,4,5,6) in 293 T cells. G CCK-8 assay was used to detect the cell proliferation after IR between miR-125a-5p-overexpressing and control MODE-K cells. H Quantitative assessment of percentage of EdU-positive cells in MODE-K cells overexpressing miR-125a-5p. I Flow cytometry was used to detect the percentage of apoptotic cells in MODE-K cells transfected with NC mimics or miR-125a-5p mimics after 8 Gy IR. J Comet assay was performed and quantitative results of Olive tail moment of MODE-K cells transfected with miR-125a-5p mimics after 8 Gy IR. K Representative plots and quantification of γ-H2AX foci in MODE-K cells transfected with NC mimics or miR-125a-5p mimics after 4 Gy IR. Bars represent mean ± SEM. *P < 0.05, **P < 0.01 and ***P < 0.001

Fig. 4

circDmbt1(3,4,5,6)/miR-125a-5p/STAT3 axis is partially responsible for intestinal epithelium protection against IR of circDmbt1(3,4,5,6). A Venn diagram analysis of candidate miR-125a-5p-targeted mRNA as predicted by Pic Target, Target Scan, StarBase, and miRDB databases. B The sequences of circDmbt1(3,4,5,6) containing the miR-125a-5p binding sites or mutant binding sites are shown. C Dual-luciferase reporter assay was used to detect the luciferase activities of pmirGLO-NC, Stat3-WT and Stat3-MUT luciferase reporter plasmids to evaluate the interaction between miR-125a-5p and STAT3 in MODE-K cells. D Representative blots (left) and quantitative assessment of the percentage of EdU-positive cells (right) in MODE-K overexpressing circDmbt1(3,4,5,6) and miR-125a-5p under irradiated and non-irradiated conditions. E Flow cytometric detection of the rate of apoptosis in MODE-K cells co-transfected with circDmbt1(3,4,5,6) plasmids and miR-125a-5p mimics after 8 Gy IR. F Comet assay was performed, and quantitative results of Olive tail moment of MODE-K cells co-transfected with circDmbt1(3,4,5,6) plasmids and miR-125a-5p mimics after 8 Gy IR were displayed. G Representative plots (left) and quantification (right) of γ-H2AX foci in MODE-K cells co-transfected with circDmbt1(3,4,5,6) plasmids and miR-125a-5p mimics 6 h and 24 h after 4 Gy IR. Bars represent mean ± SEM. *P < 0.05 and ***P < 0.001

Fig. 5

Overexpression of circDmbt1(3,4,5,6) promotes the proliferation of small intestinal organoids after IR. A Green fluorescence of small intestinal organoids of circDmbt1(3,4,5,6) mice overexpressing circDmbt1(3,4,5,6) labeled with GFP was observed with fluorescence microscopy, indicating successful lentivirus infection. B Lentivirus-mediated overexpression of circDmbt1(3,4,5,6) in small intestinal organoids was verified by RT–qPCR. C Small intestine organoids overexpressing circDmbt1(3,4,5,6) and the organoids were observed microscopically after IR. D A Cell Titer Glo assay was performed to quantify the proliferation of small intestinal organoids overexpressing circDmbt1(3,4,5,6) and that of control organoids after 8 Gy IR. E, F Control organoids and circDmbt1(3,4,5,6)-overexpressing organoids received 8 Gy IR, and EdU staining assays were performed 1 day after IR. The results showed higher proliferation in small intestinal organoids overexpressing circDmbt1(3,4,5,6) after IR than that of control organoids. G, H Control organoids and circDmbt1(3,4,5,6)-overexpressing organoids received 4 Gy IR, and KI67 immunofluorescence assays were performed 1 day after IR. The results showed higher proliferation in small intestinal organoids overexpressing circDmbt1(3,4,5,6) after IR than that of control organoids. Bars represent mean ± SEM. *P < 0.05 and ***P < 0.001

Fig. 6

Overexpression of circDmbt1(3,4,5,6) alleviates ABI-induced acute intestinal injury in vivo. A Schematic diagram of the experimental process for in vivo overexpression of circDmbt1(3,4,5,6) by tail vein injection of AAV-circDmbt1(3,4,5,6), AAV-circNC or PBS, followed by 14 Gy ABI. B Verification of circDmbt1(3,4,5,6) overexpression was carried out by RT–qPCR, n = 3. C Verification of miR-125a-5p down-expression in AAV-circDmbt1(3,4,5,6) group compared with PBS and AAV-circNC groups was verified by RT–qPCR, n = 3. D Early survival analysis of different treated groups after 18 Gy ABI. E Small intestine lengths of virus-injected mice at different time points after IR, n = 6. F Mice were administered FITC-DexTran 4000 intragastrically after IR and sacrificed 50 min later. Serum FITC fluorescence intensity was detected to characterize integrity of the jejunum, n ≥ 3. G, H Small intestinal pictures and histological score are displayed. I, L Representative images of H&E staining (I) and mice jejunum villi lengths (L) after IR, n ≥ 12. J, M Representative images (J) and quantification (M) of immunohistochemistry of KI67, a marker of small intestinal proliferation, n ≥ 8. K, N Representative images (K) and quantification (N) of immunohistochemistry of OLFM4, a marker of small intestinal stem cells, n ≥ 43. O, P Representative images of TUNEL staining for cell apoptosis of jejunal paraffin section in mice treated with PBS, AAV-NC, or AAV-circDmbt1(3,4,5,6) (O) and bar graph showing fluorescence intensity in the indicated cells (P), n ≥ 4. Scale bar: 100 μm for (I), (J), and (K), scale bar: 400 μm for (O). Bars represent mean ± SEM. d days; *P < 0.05 and **P < 0.01