MicroRNA-146a-mediated downregulation of IRAK1 protects mouse and human small intestine against ischemia/reperfusion injury

Abstract

Intestinal ischemia/reperfusion (I/R) injury causes inflammation and tissue damage and is associated with high morbidity and mortality. Uncontrolled activation of the innate immune system through toll-like receptors (Tlr) plays a key role in I/R-mediated tissue damage but the underlying mechanisms have not been fully resolved. Here, we identify post-transcriptional upregulation of the essential Tlr signalling molecule interleukin 1 receptor-associated kinase (Irak) 1 as the causative mechanism for post-ischemic immune hyper-responsiveness of intestinal epithelial cells. Increased Irak1 protein levels enhanced epithelial ligand responsiveness, chemokine secretion, apoptosis and mucosal barrier disruption in an experimental intestinal I/R model using wild-type, Irak1(-/-) and Tlr4(-/-) mice and ischemic human intestinal tissue. Irak1 accumulation under hypoxic conditions was associated with reduced K48 ubiquitination and enhanced Senp1-mediated deSUMOylation of Irak1. Importantly, administration of microRNA (miR)-146a or induction of miR-146a by the phytochemical diindolylmethane controlled Irak1 upregulation and prevented immune hyper-responsiveness in mouse and human tissue. These findings indicate that Irak1 accumulation triggers I/R-induced epithelial immune hyper-responsiveness and suggest that the induction of miR-146a offers a promising strategy to prevent I/R tissue injury.

Figure 1. Hypoxia increases epithelial Irak1 protein and causes innate immune hyper-responsiveness

*Student's t-test p < 0.05, **p < 0.01, ***p < 0.001 compared to controls (B,C) or hypoxia LPS-treated (D). Values are means ± SEM from 3 to 5 independent experiments, n = 4/group.

1. Time kinetic of Irak1 and Hif-1α protein expression in mIC_cl2 cells after incubation in hypoxic chambers.

2. mIC_cl2 cells were incubated in hypoxic chambers for indicated time, and subsequently stimulated under normoxic conditions with 1 ng/ml LPS or 1 µM PMA for 6 h, and the secretion of Cxcl2 was determined. For each data point, n = 4. normoxia: LPS 0.536 ± 0.048 and PMA 0.239 ± 0.091 versus control 0.063 ± 0.008, p = 2 × 10⁻⁶ and p = 0.09, respectively; 0.5 h: LPS 0.728 ± 0.117 versus control 0.073 ± 0.033, p = 3 × 10⁻³; 1 h: LPS 1.156 ± 0.325 versus control 0.094 ± 0.011, p = 6 × 10⁻⁴; 1.5 h: LPS 1.336 ± 0.144 versus control 0.065 ± 0.015, p = 2 × 10⁻⁶; 2 h: LPS 1.580 ± 0.291 versus control 0.037 ± 0.034, p = 4 × 10⁻⁵; 4 h: LPS 1.821 ± 0.425 versus control 0.117 ± 0.082, p = 2 × 10⁻⁴).

3. mIC_cl2 cells were incubated in hypoxic chambers for 2 h and subsequently stimulated with various concentrations of LPS under normoxic conditions for 6 h, and the secretion of Cxcl2 was determined. For each data point, n = 4.0.05: hypoxia 0.188 ± 0.040 versus control 0.032 ± 0.020, p = 4 × 10⁻⁴; 0.1: hypoxia 0.316 ± 0.098 versus control 0.059 ± 0.025, p = 0.002; 0.5: hypoxia 0.931 ± 0.080 versus control 0.215 ± 0.034, p = 3 × 10⁻⁶; 1: hypoxia 1.455 ± 0.201 versus control 0.525 ± 0.075, p = 10⁻⁴.

4. mIC_cl2 cells were left untreated or incubated in hypoxic chambers for 2 h followed for one fraction of cells by overnight incubation in fresh medium in normoxic conditions (recovery). Subsequently, the levels of Irak1 and Cxcl2 secretion were determined after 6 h stimulation with 1 ng/ml LPS. For each data point, n = 4. Hypoxia/recovery + LPS (0.690 ± 0.173) versus Hypoxia + LPS (1.456 ± 0.317), p = 0.005.

Figure 2. Hypoxia alters the ubiquitination pattern and induces Irak1 SUMOylation

*Student's t-test p < 0.05, ***p < 0.001. Values are means ± SEM, from four independent experiments, n = 4/group.

1. A. mIC_cl2 cells were transfected with an ubiquitin-encoding plasmid and incubated in hypoxic chambers or treated with 10 ng/ml LPS for 2 h in the presence of the proteasome inhibitor MG132 (100 nM). Irak1 was immunoprecipitated, and K48 or K63 ubiquitination was individually detected by immunoblot.

2. B,C. mIC_cl2 cells were transfected with a plasmid-encoding haemagglutinin (HA)-tagged SUMO, left untreated (co.) or subjected to hypoxia in hypoxic chambers (Hyp.) for 2 h in the presence of 10 mM NEM and 100 nM MG132. Irak1 protein was detected by immunoblot (B), or (C) IRAK1 was immunoprecipitated, and SUMO-HA and Irak1 were detected by immunoblot.

3. D. mIC_cl2 cells were left untreated (control), treated with the hypoxia mimetic DMOG (1 mM), or incubated in hypoxic chambers for 2 h and Irak1 levels were determined by immunoblot.

4. E. mIC_cl2 cells were left untreated (co.) or incubated in hypoxic chambers (Hyp.) for 2 h, Senp1 was immunoprecipitated, and IRAK1 and SENP1 were detected by immunoblot.

5. F,G. mIC_cl2 cells were transfected with siRNA directed against Senp1 or Ubc9, and incubated in hypoxic chambers for 2 h. (F) The level of Irak1 was determined by immunoblot, and (G) the amount of Cxcl2 secretion after stimulation with 1 ng/ml LPS for 6 h was quantified by ELISA. Hypoxia/Senp1 + LPS (0.331 ± 0.092) and Hypoxia/Ubc9 + LPS (2.564 ± 0.751) versus Hypoxia + LPS (1.481 ± 0.122), p = 5 × 10⁻⁶ and p = 0.02, respectively.

Figure 3. Ischemia-induced Irak1 accumulation leads to innate immune hyper-responsiveness and Tlr4- and Irak1-dependent tissue injury

***Student's t-test p < 0.001, **p < 0.01, compared to LPS-treated (B–C), WT (D), or WT I/R (F). Values are means ± SEM from 4 to 5 independent experiments. Magnification ×100.

1. A. Wild-type mice (n = 12) were subjected to ischemia for 30 min (I), and IECs were isolated. Irak1 protein was determined by immunoblot.

2. B. An ischemic segment and an unaffected control segment of the intestine were removed and incubated ex vivo for 2 h in the presence of 100 ng/ml LPS at 37°C. IECs were isolated and Cxcl2 mRNA was quantified by real-time PCR. I/R + LPS (71.66 ± 1.56) versus I/R (44.20 ± 2.12), p = 0.004. n = 4 for each data point.

3. C. After 30 min ischemia, 200 µl of a solution of 100 ng/ml LPS was injected intraluminally during a 1 h reperfusion period. IECs were then isolated, and Cxcl2 mRNA was quantified by real-time RT-PCR. I/R + LPS (90.87 ± 20.70) versus I/R (15.32 ± 3.67), p = 3 × 10⁻⁵. n = 4 for each data point.

4. D–F. Wild-type (WT), Tlr4^−/−, and Irak1^−/− mice (n = 10 for each group) were subjected to I/R. (D) IECs were isolated, and Cxcl2 mRNA was quantified by real-time RT-PCR. Tlr4^−/− (1.72 ± 1.91) and Irak1^−/− (2.67 ± 2.70) versus WT (21.23 ± 5.08), p = 5 × 10⁻⁶and p = 10⁻⁵, respectively. (E) H&E staining was performed using formalin-fixed tissue sections from I/R-treated or untreated (Co) intestinal segments. (F) Permeability of the intestinal barrier was measured by injecting 200 µl of a 25 mg/ml FITC-dextran solution into the intestinal lumen during ischemia (I/R) or into untreated control sections (Co). Subsequently, blood samples were collected and the fluorescence intensity was measured. Tlr4^−/− I/R (44.03 ± 7.06) and Irak1^−/− I/R (41.05 ± 13.64) versus WT I/R (119.95 ± 7.00), p = 5 × 10⁻⁶. n = 5 for each data point.

5. G. TUNEL staining was performed using formalin-fixed tissue sections from I/R-treated or untreated (Co) intestinal segments.

6. H. Expression of phospho-jnk (P-Jnk), Jnk, phosphor-Bax and Bax in IECs were assessed by Western blotting.

7. I. The translocation of Aif was measured in nuclear extract (NE) and cytosolic extract (CE) of IECs (TATA binding protein Tbp and Gapdh expression were used as nuclear or cytosolic loading control respectively).

Figure 4. DIM-mediated induction of miR-146a diminishes the hypoxia-induced hyper-sensitivity towards LPS and protects against I/R injury in mice

***Student's t-test p < 0.001, **p < 0.01 compared to control (A), LPS-treated (C), I/R-treated (E, G, H), or I/R and LPS-treated (F). Values are means ± SEM from 4 to 5 separated experiments. Magnification ×100.

1. A,B. mIC_cl2 cells were incubated in hypoxic chambers for 2 h in the absence or presence of 25 µM diindolylmethane (DIM), and the levels of miR-146a (for each point, n = 4; Control: DIM 3.65 ± 0.58 versus Co 1.03 ± 0.64, p = 9 × 10⁻⁴; Hyp: DIM 3.99 ± 0.34 versus Co 1.21 ± 0.37, p = 3 × 10⁻⁵) (A) and Irak1 (B) were determined by real-time PCR and immunoblot, respectively.

2. C. mIC_cl2 cells were transfected with microRNA (miR)-146a mimic, an anti-miR-146a (anti-miR) or miR control (miRco), and/or treated with DIM and subjected to hypoxia for 2h. Cxcl2 mRNA was quantified by real-time RT-PCR after stimulating the cells with 1 ng/ml LPS for 6h (for each point, n = 4; miR-146a/LPS 0.440 ± 0.184 and DIM/LPS 0.475 ± 0.194 versus LPS 1.1516 ± 0.363, p = 10⁻³ and p = 0.002, respectively).

3. D. 200 µl of 25 µM DIM or the solvent control DMSO (Mock) was intraluminally injected into the small intestine of WT mice (n = 10 for each group). Subsequently, the intestinal tissue was subjected to ischemia for 30 min followed by reperfusion for 1 h. IECs were isolated and Irak1 protein was determined by immunoblotting.

4. E. 100 nM miR-146a mimic, 100 nM miR control (miRco), 25 µM DIM or the solvent control (Mock) was intraluminally injected into the small intestine of WT mice (n = 10 for each group). Subsequently, the intestinal tissue was subjected to ischemia for 30 min, followed by reperfusion for 1 h. IECs were isolated and Cxcl2 mRNA was quantified by real-time PCR. miR-146a/IR 3.17 ± 1.17 and DIM/IR 3.59 ± 0.85 versus IR 21.24 ± 2.20, p = 6 × 10⁻⁶ and p = 5 × 10⁻⁶, respectively.

5. F. Epithelial Cxcl2 mRNA in response to intraluminal injection of 100 ng/ml LPS during the reperfusion period after pretreatment with miRco, miR-146a, DIM or Mock was determined by quantitative PCR. miR-146a/IR/LPS 34.90 ± 5.98 and DIM/IR/LPS 42.19 ± 8.88 versus IR/LPS 80.51 ± 6.09, p = 4 × 10⁻⁵ and p = 4 × 10⁻⁴, respectively. For each data point, n = 5.

6. G. Lipid peroxidation, used as an indirect index of the oxidative injury induced by the reactive oxygen species, was determined by measuring the malonedialdehyde (MDA) concentration with the thiobarbiturate reaction. miR-146a/IR 0.2633 ± 0.0198 and DIM/IR 0.2467 ± 0.715 versus IR 0.5304 ± 0.0446, p = 3 × 10⁻⁵ and p = 5 × 10⁻⁴, respectively. For each data point, n = 5.

7. H. The permeability of the intestinal barrier was measured by intraluminal injection of 200 µl 25 mg/ml FITC-dextran during the ischemia period after pretreatment with miRco, miR-146a, DIM or Mock. Blood samples were collected after reperfusion, and the fluorescence intensity was determined. miR-146a/IR 35.66 ± 12.99 and DIM/IR 45.16 ± 7.99 versus IR 122.61 ± 3.51, p = 9 × 10⁻⁴ and p = 2 × 10⁻⁴, respectively. For each data point, n = 4.

8. I. H&E staining of small intestinal tissue sections left untreated (co) or after I/R with or without pretreatment with miR-146a or DIM.

9. J. Levels of phospho-Jnk (P-Jnk), phospho-Bax (P-Bax) and nuclear Aif was determined by immunoblot.

10. K. TUNEL staining of small intestinal tissue sections left untreated (co) or after I/R with or without pretreatment with miR-146a or DIM.

Figure 5. DIM-mediated decrease in IRAK1 reduces hypoxia-induced innate immune hyper-responsiveness in human intestinal mucosa

1. A. Immunohistology for IRAK1 and HIF-1α in tissue sections of human small intestine obtained from normoxic controls or patients suffering from mesenteric infarction. Magnification ×100.

2. B,C. Human small intestinal biopsies (n = 5) were left untreated or subjected to hypoxia for 2 h in the absence or presence of DIM. (B) IRAK1 and HIF-1α was determined by immunoblot and (C) LPS susceptibility was analysed by quantitative RT-PCR for Cxcl8 RNA after incubating for 6 h in the absence or presence of 10 ng/ml LPS. Hypoxia/DIM 29.45 ± 3.04 versus hypoxia 83.88 ± 29.90, p = 0.02. For each data point, n = 5.*Student's t-test p < 0.05 between groups. Values are means ± SEM from five separated experiments.