Irisin reverses intestinal epithelial barrier dysfunction during intestinal injury via binding to the integrin αVβ5 receptor

Abstract Disruption of the gut barrier results in severe clinical outcomes with no specific treatment. Metabolic disorders and destruction of enterocytes play key roles in gut barrier dysfunction. Irisin is a newly identified exercise hormone that regulates energy metabolism. However, the effect of irisin on gut barrier function remains unknown. The therapeutic effect of irisin on gut barrier dysfunction was evaluated in gut ischemia reperfusion (IR). The direct effect of irisin on gut barrier function was studied in Caco‐2 cells. Here, we discovered that serum and gut irisin levels were decreased during gut IR and that treatment with exogenous irisin restored gut barrier function after gut IR in mice. Meanwhile, irisin decreased oxidative stress, calcium influx and endoplasmic reticulum (ER) stress after gut IR. Moreover, irisin protected mitochondrial function and reduced enterocyte apoptosis. The neutralizing antibody against irisin significantly aggravated gut injury, oxidative stress and enterocyte apoptosis after gut IR. Further studies revealed that irisin activated the AMPK‐UCP 2 pathway via binding to the integrin αVβ5 receptor. Inhibition of integrin αVβ5, AMPK or UCP 2 abolished the protective role of irisin in gut barrier function. In conclusion, exogenous irisin restores gut barrier function after gut IR via the integrin αVβ5‐AMPK‐UCP 2 pathway.

apoptosis. 5 On the other hand, ROS accelerates inflammatory cell infiltration and cytokine release, which further fuel ROS generation as a positive feedback during gut IR. 6 Massive enterocyte apoptosis and disrupted intercellular tight junctions are the main mechanisms of intestinal barrier dysfunction. 7 As a protective mechanism, mitochondrial uncoupling protein 2 (UCP 2) serves as a negative feedback regulator in the presence of excessive ROS. Overexpression of UCP 2 significantly decreased oxidative stress and cell apoptosis. 8 Accumulating evidence suggests that AMP-activated protein kinase (AMPK) plays a pivotal role in gut barrier function. 9 As an energy sensor, AMPK regulates mitochondrial function and the energy requirement of enterocytes. [10][11][12] Besides, AMPK can directly facilitate the aggregation of cytoskeletal proteins and formation of intercellular tight junctions via activation of Rac1. 9,13 Since first reported in 2012, irisin has been an intriguing option for solving obesity problems. 14 As a newly identified hormone, the major function of irisin is regulating glucose/lipid metabolism and mitochondrial function. 14,15 However, as research continues, it has been shown that irisin also benefits type 2 diabetes, ageing and some cardiovascular diseases. [16][17][18] In addition, many studies have shown that irisin can facilitate AMPK activation to further regulate energy metabolism. 13 A recent study indicated that irisin restrains bone loss by binding to the αv class of integrins in osteocytes. 19 However, the effect of exogenous irisin on gut barrier function has not been elucidated to date. We therefore suggested that irisin restores gut barrier function after gut IR via activation of the integrin αvβ5-AMPK-UCP 2 pathway. The main purpose of this study was to determine the effects of exogenous irisin on gut barrier function after gut IR. In addition, our study also sought to clarify the effects of irisin on the integrin αvβ5-AMPK-UCP 2 pathway during gut IR injury.

| Experimental animals
Experiments were performed on male wild-type C57BL/6 J mice

| Mouse model of gut IR
A mouse model of gut IR was conducted as described previously. 20 The superior mesenteric artery (SMA) was occluded with an atraumatic clip for 60 minutes, and then, reperfusion was allowed under anaesthesia with isoflurane. Sham group mice were given 0.5 mL saline after sham operation without ischemia treatment; mice were intravenously administered 250 μg/kg irisin (067-29A; Phoenix Pharmaceuticals, Inc), 20 mg/kg cilengitide trifluoroacetate (S707; Selleck) or 20 mg/kg genipin (S2412; Selleck, China) immediately after reperfusion. Four hours later, the mice were euthanized, and the following experiments were performed. In additional groups of animals, anti-irisin (4 mg/kg; Phoenix Pharmaceuticals) blocking antibody was administered at 24 hours before gut IR.

| Histological analysis and gut IR score
Haematoxylin and eosin staining of fixed intestinal tissues was performed as described previously. 22 Images were collected by a light microscope, and a representative field was chosen for assessment.
The gut IR score was graded as follows: 0, normal mucosal villi; 1, minor subepithelial space and capillary congestion; 2, extensive subepithelial space with little epithelial layer lifting from the lamina propria; 3, massive epithelial layer lifting from the lamina propria; and 4, villi detachment and haemorrhage. 23

| Water content
Gut tissues were weighed immediately (wet weight) and at 48 hours after drying in a 60°C oven (dry weight). Gut water content was calculated as H 2 O % = (1 − dry weight/wet weight) × 100%.

| FITC-dextran permeability assay
Mice were administered 200 μL FITC-dextran (25 mg/mL) immediately after reperfusion by gavage. Four hours later, mice were sacrificed, and blood FITC-dextran concentrations were assessed with a Varioskan™ LUX multimode microplate reader (Thermo Scientific™) at an excitation wavelength of 485 nm and an emission wavelength of 515 nm.

| Bacterial content
The mesenteric lymph node complex and lung tissues were harvested after euthanasia. The tissues were homogenized and centrifuged to obtain a supernatant. After serial log dilutions, 500 μL each dilution was evenly coated onto chocolate agar plates. The plates were incubated at 37°C for 24 hours, and colony-forming units (CFUs) were counted.

| Western blot analysis
Western blot analysis was performed as described previously. 22 PVDF membranes were incubated with primary rabbit anti-irisin was incubated for 1 hour at room temperature. Protein expression was detected by a chemiluminescence system (Bio-Rad) and quantified by ImageJ2x software.

| Activation of Rac1
Activated and total Rac1 levels were determined with a Rac1 Activation Assay Combo Kit (STA-405; Cell Biolabs) following the manufacturer's instructions.

| Immunofluorescence
Immunofluorescence staining was performed as described previously. 22 Samples were incubated with rabbit anti-irisin antibody

| Measurement of TER
The TER of Caco-2 cells was determined by an electrical cell-substrate impedance sensing system (Applied Biophysics) as described previously. 25

| Transwell permeability assays
Transwell permeability assays were performed using 6.5-mm transwell dishes with 0.4 µm pore polycarbonate membrane inserts (3413; Corning). FITC-albumin concentrations were assessed with a Varioskan™ LUX multimode microplate reader (Thermo Scientific™) at an excitation wavelength of 485 nm and an emission wavelength of 515 nm.

| Flow cytometry analysis
An annexin V-FITC/PI Apoptosis Detection Kit (AD10; Dojindo Laboratories) was used to detect Caco-2 cell apoptosis with flow cytometry (ACEA Biosciences, Inc) according to the manufacturer's instructions. The percentage of apoptotic cells was calculated from the sum of early apoptosis and late apoptosis (n = 3 per group).

| Analysis of mitochondrial DNA content
Mitochondrial DNA (mtDNA) content was detected as mtDNA encoded NADH dehydrogenase-1 and normalized against the nuclear encoded POU class 5 homeobox 1 gene as described previously. 22

| Determination of LDH levels
Serum LDH was quantified by using an assay kit (A020-2; Nanjing Jiancheng Bioengineering Institute) according to the manufacturer's instructions.

| Statistical analysis
Results were expressed as the means ± standard error of the mean (SEM). t Test or one-way ANOVA was applied to analyse the differences between groups by SPSS 18.0. P < .05 represents a significant difference.

| Exogenous irisin restores gut barrier function after gut IR
A significant reduction in serum irisin was observed after gut IR treatment, while mice received recombinant irisin treatment (250 μg/kg, iv) showed higher irisin levels at 4 hours after gut IR ( Figure 1A). Irisin levels in the intestine were detected by Western blot as shown in Meanwhile, irisin-treated mice exhibited lower water content than the control-treated mice after gut IR ( Figure 1H). Consistent with the histological changes, a significant increase in serum FITC-dextran was detected after gut IR, while irisin treatment significantly reversed this change ( Figure 1I). Mesenteric lymph node (MLN) and lung bacterial loads were determined, and the results showed that irisin treatment significantly reduced the increase of bacterial translocation to the MLN and lung that occurred after gut IR ( Figure 1J,K).
Additionally, the neutralizing antibody against irisin significantly aggravated gut injury and increased the levels of water content, serum FITC-dextran and bacterial loads in gut IR mice ( Figure 1F-K) In addition, irisin treatment markedly decreased the levels of serum LDH and lactate ( Figure 1L,M). Moreover, the irisin-treated group showed lower levels of serum tumour necrosis factor α (TNF-α) and cold-inducible RNA binding protein (CIRP) than the control-treated group ( Figure 1N,O).

| Irisin increases the intercellular tight junctions between enterocytes after gut IR
Western blot revealed a conspicuous decrease in tight junctionrelated claudin-1 and occludin expression during gut IR injury, but these changes were reversed by exogenous irisin treatment ( Figure 2A

| Irisin protects mitochondrial function to increase intercellular tight junctions and reduce enterocyte apoptosis
Mitochondrial dysfunction is the key cause of cell apoptosis. MitoTracker Red staining was conducted to determine the number of mitochondria.
MitoTracker distribution and fluorescence intensity were markedly decreased after H/R treatment. Irisin treatment significantly reversed the changes in mitochondrial visualization ( Figure 4A,B). Meanwhile, we found that irisin administration markedly increased the ATP concentration after H/R treatment in Caco-2 cells ( Figure 4C). mtDNA copy number in intestine tissues was detected to assess the number of mitochondria. Consistent with the in vitro results, irisin-treated mice had higher mtDNA copy number and ATP concentration compared with the saline-treated mice after gut IR ( Figure 4D,E).
TUNEL staining was conducted to assess apoptosis of enterocytes. Mice that underwent gut IR exhibited a large number of apoptotic cells. Treatment with exogenous irisin significantly reduced the F I G U R E 2 Irisin increases intercellular tight junctions between enterocytes after gut IR. Irisin (250 μg/kg in 0.5 mL saline, a single dose, iv) was administered immediately after reperfusion. Anti-irisin (4 mg/kg; Abcam) blocking antibodies were administered at 24 h before gut IR. Four hours after reperfusion, mice were sacrificed, and tissue samples were collected. A,B, Western blot analysis of claudin-1 and occludin expression; (C) immunofluorescence staining of JAM-1 and ZO-1 (red) and the corresponding nuclear counterstaining (blue) in gut tissues. n = 6 per group, mean ± SEM, *P < .05 vs the sham group, # P < .05 vs the gut IR group. Caco-2 cells were exposed to hypoxia for 90 min, and 10 nmol/L irisin was added at the beginning of reoxygenation.

| Irisin protects against gut IR injury via binding to integrin αvβ5 receptor in enterocyte
Immunofluorescent staining revealed an observable co-localization of irisin and integrin αvβ5 proteins after irisin administration in H/R treated caco-2 cells ( Figure 5A). Cilengitide trifluoroacetate, a cyclic RGD-containing pentapeptide, is an inhibitor of integrin αvβ5. F I G U R E 3 Irisin decreases the oxidative stress, calcium influx and ER stress after gut IR. Irisin (250 μg/kg in 0.5 mL saline, a single dose, iv) was administered immediately after reperfusion. Anti-irisin (4 mg/kg, Abcam, USA) blocking antibody was administered at 24 h before gut IR. Four hours after reperfusion, mice were sacrificed, and tissue samples were collected. A,B, DHE fluorescence staining of gut tissues; (C-G) levels of gut malonaldehyde (MDA), xanthine oxidase (XO), 4-hydroxynonenal (4-HNT), superoxide dismutase (SOD) and glutathione peroxidase activity (GSH-PX), respectively; n = 6 per group, mean ± SEM, *P < .05 vs the sham group, # P < .05 vs the gut IR group. Caco-2 cells were exposed to hypoxia for 90 min, and 10 nmol/L irisin was added at the beginning of reoxygenation. (H,I) DHE fluorescence staining; (J,K) Fluo-4 AM staining of Ca 2+ ; (L,M) Western blot analysis of IRE1 and CHOP expression at 4 h after reoxygenation in Caco-2 cells; n = 3 per group, mean ± SEM, *P < .05 vs the sham group, # P < .05 vs the H/R group

| Irisin restores gut barrier function after gut IR by the integrin αVβ5-AMPK-UCP 2 pathway
In the process of clarifying the potential mechanisms for irisin increasing intercellular tight junctions, we found that irisin dra-  F I G U R E 4 Irisin protects mitochondrial function to increase intercellular tight junctions and reduce enterocyte apoptosis. Caco-2 cells were exposed to hypoxia for 90 min, and 10 nmol/L irisin was added at the beginning of reoxygenation. A,B, MitoTracker Red fluorescence staining of mitochondria; (C) ATP concentration at 4 h after reoxygenation in Caco-2 cells; n = 3 per group, mean ± SEM, *P < .05 vs the sham group, # P < .05 vs the H/R group. Irisin (250 μg/kg in 0.5 mL saline, a single dose, iv) was administered immediately after reperfusion. Anti-irisin (4 mg/kg; Abcam) blocking antibodies were administered at 24 h before gut IR. Four hours after reperfusion, mice were sacrificed, and tissue samples were collected. (D) Gut mtDNA copy numbers; (E) gut ATP concentration; (F,G) TUNEL fluorescence staining (green) and corresponding nuclear counterstaining (blue); n = 6 per group, mean ± SEM, *P < .05 vs the sham group, # P < .05 vs the gut IR group.  Figure 7J-M).

| D ISCUSS I ON
In this study, we found that irisin restores gut barrier function after gut IR via relieving oxidative stress, calcium influx, ER stress and mitochondrial dysfunction. The potential mechanism is that irisin activates the AMPK-UCP 2 pathway via binding to integrin αVβ5 receptor in enterocyte (Figure 8). Irisin therefore exhibits promising practical application prospects to solve gut barrier dysfunction-related diseases in the future.
It has been thoroughly proven that physical exercise, mainly skeletal muscle activity, benefits the whole body, especially organs such as the heart, lung, brain and gut. 28 Secreted irisin is derived from fibronectin type III domain containing 5 (FNDC5) protein, mainly in skeletal muscle during exercise. 14,29 The discovery of F I G U R E 6 Irisin restores gut barrier function after gut IR by the integrin αVβ5-AMPK-UCP 2 pathway. Irisin (250 μg/kg, iv), cilengitide trifluoroacetate (20 mg/kg, iv) or genipin (20 mg/kg, iv) were administered immediately after reperfusion. Four hours after reperfusion, mice were sacrificed, and tissue samples were collected. A-D, Western blot analysis of the activation of AMPK, UCP 2 and UCP 1 in gut tissues; n = 6 per group, mean ± SEM, *P < .05 vs the sham group, # P < .05 vs the gut IR group. Caco-2 cells were exposed to hypoxia for 90 min, and that irisin is involved in obesity, cardiovascular diseases, telomere length and ageing, and hippocampal neurogenesis. 14 Moreover, a recent study verified that irisin can directly bind to the αv class of integrin receptors in osteocytes. 19 Previous studies have demonstrated that gut barrier dysfunction after gut IR is mainly caused by metabolic disorders and destruction of enterocytes, which can benefit from exercise. 30,31 However, the effects of irisin on the intestinal barrier have not been elucidated to date. In this study, F I G U R E 7 Genipin abolished the protective role of irisin in gut IR. Irisin (250 μg/kg in 0.5 mL saline, a single dose, iv) was administered immediately after reperfusion. Four hours after reperfusion, mice were sacrificed, and tissue samples were collected. A-C, levels of gut malonaldehyde (MDA), xanthine oxidase (XO) and 4-hydroxynonenal (4-HNT), respectively; n = 6 per group, mean ± SEM, *P < .05 vs the gut IR group, # P < .05 vs the gut IR + irisin group. Caco-2 cells were exposed to hypoxia for 90 min, and 10 nmol/L irisin was added at the beginning of reoxygenation. Moreover, previous studies have demonstrated that macrophages play a crucial role in ischemia reperfusion injury. 32 A recent report has confirmed the macrophages, which is a target of FNDC4, a homologue of irisin, is associated with intestinal inflammation. 33 Therefore, it is possible that macrophages may be another target of irisin during gut IR. Mitochondrial dysfunction is one of the main mechanisms of ischaemia-reperfusion injury. A decrease in mitochondrial and ATP contents results in cellular energy stress and apoptosis. 36 As an energy sensor, AMPK regulates energy metabolism via its phosphorylation and maintains cellular mitochondrial and ATP homeostasis. 13 Moreover, previous studies proved that AMPK can directly strengthen the aggregation of cytoskeletal proteins and intercellular tight junctions via activating Rac1. 37 Rac1 is a member of the GTPase family. Activated Rac1 (GTP-bound state) maintains the integrity of intercellular tight junctions in the epithelial monolayer by the formation of cortical actin. 38 As described above, irisin plays a pivotal role in energy metabolism and mitochondrial function, but whether irisin facilitates AMPK-dependent mitochondrial protection and intercellular tight junctions has not been previously elucidated.
In this study, we showed that irisin increased AMPK and Rac1 activation and relieved mitochondrial dysfunction and enterocyte apoptosis. AMPK siRNA abolished the protective effects of irisin on gut barrier function. Irisin therefore might restore gut barrier function via activation of AMPK during gut IR injury.
Enterocyte apoptosis is a key mechanism of gut barrier dysfunction. 39 Gut IR facilitates ROS generation and eventually exceeds the antioxidant capacity of enterocytes. 40 Excessive ROS accumulation-induced oxidative stress reactions finally result in calcium influx, ER stress and mitochondrial dysfunction and eventual cell death. 41 A previous study indicated that ROS scavenging is an effective method to improve cell survival. 42 As a protective mechanism, mitochondrial UCP 1 and UCP 2 serve as a negative feedback regulator in the presence of excessive ROS.
Overexpression of UCP 2 significantly decreased oxidative stress and cell apoptosis. 8 Interestingly, irisin was proven to have an antioxidant capacity in multiple diseases. 14 In this study, we found that irisin decreased ROS accumulation as well as increased UCP This study has some limitations. First of all, this study mainly clarified the therapeutic implications with the administration of irisin F I G U R E 8 The exercise hormone irisin protects against gut barrier function after gut IR via relieving oxidative stress, calcium influx, ER stress and mitochondrial dysfunction. The potential mechanism is that irisin activates the AMPK-UCP 2 pathway via binding to integrin αVβ5 receptor in enterocytes 24 hours prior to gut ischemia reperfusion injury and the additional pre-clinical studies with post-treatment of irisin are needed in the future. Furthermore, our study focused on the effects of irisin on the enterocyte barrier. The roles of irisin in the permeability of intestinal lymphatic vessels and blood vessels and other mechanisms of gut barrier dysfunction need further exploration. What's more, the impressive therapeutic effects of exogenous irisin on the gut IR-induced gut barrier dysfunction were only based on basic experiments, and prospective clinical studies are needed.
In conclusion, exogenous irisin restores gut barrier function after gut IR via integrin αVβ5-AMPK-UCP 2 pathway. Irisin therefore exhibits promising practical application prospects to solve gut barrier dysfunction-related diseases in the future.

ACK N OWLED G EM ENTS
We thank Dr Ying Hao at the Instrument Analysis Center of Xi'an Jiaotong University for her assistance with confocal analysis. This work was supported by grants from the National Nature Science

CO N FLI C T S O F I NTE R E S T
We declare that there are no competing interests.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data are available from the corresponding author on reasonable request.