Lactobacillus murinus alleviate intestinal ischemia/reperfusion injury through promoting the release of interleukin-10 from M2 macrophages via Toll-like receptor 2 signaling

Background Intestinal ischemia/reperfusion (I/R) injury has high morbidity and mortality rates. Gut microbiota is a potential key factor affecting intestinal I/R injury. Populations exhibit different sensitivities to intestinal I/R injury; however, whether this interpopulation difference is related to variation in gut microbiota is unclear. Here, to elucidate the interaction between the gut microbiome and intestinal I/R injury, we performed 16S DNA sequencing on the preoperative feces of C57BL/6 mice and fecal microbiota transplantation (FMT) experiments in germ-free mice. The transwell co-culture system of small intestinal organoids extracted from control mice and macrophages extracted from control mice or Toll-like receptor 2 (TLR2)-deficient mice or interleukin-10 (IL-10)-deficient mice were established separately to explore the potential mechanism of reducing intestinal I/R injury. Results Intestinal I/R-sensitive (Sen) and intestinal I/R-resistant (Res) mice were first defined according to different survival outcomes of mice suffering from intestinal I/R. Fecal microbiota composition and diversity prior to intestinal ischemia differed between Sen and Res mice. The relative abundance of Lactobacillus murinus (L. murinus) at the species level was drastically higher in Res than that in Sen mice. Clinically, the abundance of L. murinus in preoperative feces of patients undergoing cardiopulmonary bypass surgery was closely related to the degree of intestinal I/R injury after surgery. Treatment with L. murinus significantly prevented intestinal I/R-induced intestinal injury and improved mouse survival, which depended on macrophages involvement. Further, in vitro experiments indicated that promoting the release of IL-10 from macrophages through TLR2 may be a potential mechanism for L. murinus to reduce intestinal I/R injury. Conclusion The gut microbiome is involved in the postoperative outcome of intestinal I/R. Lactobacillus murinus alleviates mice intestinal I/R injury through macrophages, and promoting the release of IL-10 from macrophages through TLR2 may be a potential mechanism for L. murinus to reduce intestinal I/R injury. This study revealed a novel mechanism of intestinal I/R injury and a new therapeutic strategy for clinical practice. Video Abstract. Supplementary Information The online version contains supplementary material available at 10.1186/s40168-022-01227-w.


Background
Intestinal ischemia reperfusion (I/R) is a common but grave condition in some critical clinical settings such as acute mesenteric ischemia; hemorrhagic, or septic shock; severe burns; or some surgical procedures, including cardiopulmonary bypass (CPB), small bowel transplantation, and abdominal aortic surgery [1]. It not only causes local intestinal injury, but can also disrupt intestinal mucosal barrier [2], which allows enteric bacterial endotoxins to penetrate the blood and cause extraintestinal multiple organ dysfunction or even failure with high morbidity and mortality [3,4]. Currently, the mechanisms of intestinal I/R injury are not understood fully, and effective approaches for its clinical application are still lacking.
The human gastrointestinal tract houses a vast and diverse intestinal micro ora that provides nutrients and intrinsic immunity, regulates epithelial cell growth, and fundamentally affects human health and disease [5]. Few studies have reported the key role of the intestinal microbiome in intestinal I/R injury [6].
Our work stems from the previous observation that some SPF-free C57BL/6 mice survived 5 days after intestinal I/R, but some mice died quickly within 1 h after intestinal I/R. To better explore the above phenomenon, mice that died within 1 hour after reperfusion were de ned as intestinal I/R-sensitive (Sen) mice, while those that survived up to 5 days were de ned as intestinal I/R-resistant (Res) mice. These mice had the same genetic background, but their sensitivity to intestinal I/R injury was signi cantly different. We speculated that the different outcomes of mice suffering from intestinal I/R could be related to the difference in the gut microbiota. Furtherly, we conducted 16S DNA sequencing and fecal microbiota transplantation (FMT) experiment to study the intestinal bacterial community and speci c bacterial strains in the intestinal I/R model of mouse. The results showed that the relative abundance of Lactobacillus murinus (L.murinus) at the species level was higher in the Res mice than in the Sen mice, suggesting that L.murinus could prevent from intestinal I/R injury. It has been demonstrated that L. murinus, as a potential probiotic, plays an important role in maintaining intestinal immune homeostasis in the mouse model of colitis through regulating T lymphocyte activity [7]. However, the role of L. murinus in intestinal I/R injury remains unclear.
Toll-like receptor 2 (TLR2) and myeloid differentiation factor 88 (Myd88) are involved in the recognition of, and signaling in response to, a variety of pathogen-associated molecular patterns from gram-positive bacterium by innate immune cells including macrophages [8,9]. L.murinus is a gram-positive bacterium that can be used as a ligand for TLR2. It is unclear whether L.murinus plays a role in intestinal I/R injury through mediating TLR2 signaling on macrophages.
Interleukin-10 (IL-10) is an in ammatory and immunosuppressive factor and plays a vital role in controlling in ammation and preventing enteritis [10,11]. Its main sources are monocytes, macrophages and T helper cells. Murine Kupffer cells are protective in total hepatic I/R injury with bowel congestion through the release of IL-10 [12]. However, the role of IL-10 secreted by macrophages during intestinal I/R injury is yet to be revealed.
Based on the above analysis, we hypothesized that gut microbiome is associated with the postoperative outcome of intestinal I/R, and L.murinus alleviates intestinal I/R injury through promoting the release of IL-10 from macrophages via activating TLR2 signaling. In the present study, we investigated the relationship between gut microbiome and postoperative outcome of intestinal I/R as well as the effects of L.murinus on intestinal I/R injury, and elucidated the potential mechanisms. This study will shed light on a novel mechanism of intestinal I/R injury related on gut microbiota and provide a potential therapeutic strategy in the clinical settings in the foreseeable future.

Animal experiments
All animal experimental procedures were carried out in accordance with the National Institutes of Health guidelines and were approved by the local Animal Care and Use Committee of the Nanfang hospital of Southern Medical University (Guangzhou, China). Six-to eight-week-old speci c pathogen-free male C57BL/6 mice were purchased from the animal center of Nanfang Hospital of Southern Medical University (Guangzhou, China). TLR2 -/and IL-10 -/mice were purchased from Shanghai Model Organisms Center, Inc. All mice were housed under controlled temperature and humidity conditions with a 12-hour light-dark cycle, had free access to food and water and were fasted overnight before the experiment.
The model of intestinal I/R was established as described in our previous study [13]. Brie y, mice were anesthetized with 4% iso urane, and a non-invasive microvascular artery clip was placed on the superior mesenteric artery (SMA) for 60 minutes (min) followed by clip removal for reperfusion of 120 min or 5 days. During the study period, body temperature was maintained at 37 °C with a heating pad and liquid resuscitation was performed by injecting 0.5 ml of physiological saline subcutaneously, just after reperfusion.
Extraction and culture of organoids and the establishment of hypoxia-reoxygenation (H/R) models in vitro The extraction and culture of small intestinal organoids was performed as previously described [14,15]. The separated intestinal crypts were xed onto the bottom of the dish with Matrigel (STEMCELL Technologies Inc., Shanghai, China) drops and covered with IntestiCult medium (STEMCELL Technologies Inc.). For the establishment of the organoid H/R model, the organoids were placed in a humid, anaerobic environment at 37 °C for 12 hours and then placed in an aerobic environment containing 5% CO2 in a 37 °C incubator for 4 hours.

Patient Samples
From 2019 September to 2020 January, we recruited consecutive patients who underwent elective cardiac valve replacement or coronary artery bypass graft under cardiopulmonary bypass (CPB) at the Department of Cardiac Surgery, and healthy volunteers who underwent physical examination at the Department of Health Management, in Southern Medical University Nanfang Hospital, Guangzhou, China.
Participants were not included if they (1) aged <18 or >75 years, (2) had chronic kidney disease, (3) had chronic digestive system diseases, previous gastrointestinal surgery, or con rmed or suspected intestinal ischemia/necrosis, (4) used antidiarrheals, laxatives or prebiotics within one week, or used antibiotics within 3 months. A total of 26 participants were enrolled, including 20 patients who underwent CPB and 6 healthy volunteers. There was no signi cant difference between the patient group and the healthy group in terms of demographic characteristics. The study protocol was approved by the Ethical Committee of Nanfang hospital, Southern Medical University (approval number NFEC-202009-k2-01).
Fecal and blood samples: Blood samples were collected at preoperatively (T0) and at 0 h (T1), 2 h (T2), 6 h (T3), 12 h (T4) and 24 h (T5) after surgery in EDTA plasma tubes as well as in serum separator tubes for analyses of intestinal fatty acid-binding protein (IFABP) and citrulline respectively. Fecal samples were collected at preoperatively, and the relative abundance of L. murinus was quanti ed by real-time PCR. IFABP and citrulline in the plasma samples were measured at multiple time points, to allow for of (T3-T0) concentration differences, respectively by means of a human IFABP ELISA Kit (Bio-Swamp, Wuhan, China) and Citrulline ELISA Kit (USCN, Wuhan, China), following the manufacturer's instructions. The gastrointestinal complication score of the patient on the seventh day after surgery was performed according to the acute gastrointestinal injury (AGI) standard described previously [16]. The detection of L. murinus, IFABP, citrulline and AGI scores were performed by researchers blinded to the group allocation.

Microbe analysis
Feces were collected with sterilized 1.5 ml tubes before intestinal I/R and frozen at − 80 °C until DNA extraction. All extracted DNA was stored in − 20 °C until further test. The extracted fecal DNA concentration was diluted to 10 ng/μl and quantitative real-time polymerase chain reaction (PCR) was processed according to 16S rRNA primers, Firmicutes primers and Bacteroidetes primers, the primers were listed in Table S1. Moreover, 16S rRNA abundance from blood was normalized to host 18S.
All samples were paired-end sequenced on the Illumina Hiseq PE2500 sequencing platform. Low-quality reads were ltered after quality control, and high-quality reads were assigned to operational taxonomic units (OTUs) of ≥ 97 % similarity using UPARSE pipeline [19]. QIIME was applied to analyze the alpha and beta diversities, based onweighted and unweighted UniFrac distances successively Metastasis (version 20090414) and Linear discriminant effect size (LEfSe) software [20] (version 1.0) were used to explore biomarker features in each group. The KEGG pathway analysis of the OTUs was performed using Tax4Fun [21] (version 1.0) and was performed using the OmicShare tools, a free online platform for data analysis (www.omicshare.com/tools). The calculated p-value was gone through FDR Correction, taking FDR ≤ 0.05 as a threshold.

Bacterial Strains and Growth Conditions
L. murinus freeze-dried powder (BN, Beijing, China) was dissolved with 0.5 ml MRS medium (HKM, Guangdong, China), then the bacteria liquid was coated on the blood plate, and bacterial colonies appeared after about 24 h. Single colonies were picked into MRS medium, were incubated at 37 °C under anaerobic conditions, and OD600=0.6-0.7 of cultures was measured until mid-log phase after 12-16 h of growth, at which time the colony count was 6.8×10 8 CFU/mL by plate count. Frozen stocks of L. murinus (in MRS medium with 25% glycerol) were prepared, stored at -80 °C for further experiments. 50 µl frozen stocks of L. murinus were added to 5 ml MRS medium and incubated at 37 °C under aerobic conditions for 12 h, and then used for gavage of mice. In order to evaluate the total amount of L. murinus DNA in cecum and stool samples, was quanti ed by quantitative real-time PCR using the following primers: L. murinus [22], LactoM-F (5'-TCGAACGAAACTTCTTTATCACC-3') and LactoM-R (5'-CGTTCGCCACTCAACTCTTT-3').
Experimental design L. murinus pretreatment experiment: As shown in Additional le 1a, mice were randomly divided into 3 groups: (1) Sham group; (2) I/R group; (3) I/R+L. murinus group. The sham group of mice were gavaged daily for 7 days with 200 µl MRS medium and then intestinal I/R was performed without SMA occlusion. The I/R group of mice were gavaged daily for 7 days with 200 µl MRS medium, and intestinal SMA was occluded for 60 min followed by 120 min reperfusion. The I/R+ L. murinus group of mice were gavaged daily for 7 days with 200 µl 6.8×10 8 CFU/ml L. murinus prior to establishing intestinal I/R. In addition, to explore a single strain of L. murinus alleviates intestinal I/R-induced intestinal injury, mice were allocated randomly to 2 groups: Group ABX+I/R, in which antibiotics (ABX) was administered 1w before intestinal ischemia; Group ABX+I/R+L. murinus, treatment of mice with ABX for a week, and then mice were gavaged daily for 7 days with 200 µl 6.8×10 8 CFU/ml L. murinus prior to establishing intestinal I/R. Macrophage depletion experiment: Mice were allocated randomly to 4 groups: (1) I/R group; (2) PBS+Lipo+I/R group; (3) Clodronate-Lipo+I/R group; (4) Clodronate-Lipo+L. murinus+I/R group. I/R group of mice, in which intestinal SMA was occluded for 60 min followed by 120 min reperfusion was established as described previously; PBS+Lipo+I/R group of mice, which were intraperitoneally injected with 200 µl empty liposomes (Yeasen, Shanghai, China) 24 h before establishing I/R; Clodronate-Lipo+I/R group of mice, which were intraperitoneally injected with 200 µl clodronate liposomal (Yeasen, Shanghai, China) 24 h before establishing I/R model; Clodronate-Lipo+L. murinus+I/R group of mice, which were given treatment of L. murinus and clodronate liposomal injection before establishing I/R (Additional le 1b).
To explore L. murinus promotes the release of IL-10 by macrophages through TLR2 to reduce organoid H/R injury, we established the transwell co-culture system of macrophages separately extracted from WT, or TLR2 -/or IL-10 -/mice and small intestinal organoids extracted from WT mice. In our experiment, the transwell chamber (0.4 μm pore size polyester membrane; Corning, Inc.) were placed in a 6-well culture plate, small intestinal organoids were planted in the upper chamber, and macrophages were planted in the lower chamber. The 200 µl supernatant of L. murinus stimulates the upper chamber to further con rm the L. murinus may require the participation of macrophages to improve the intestinal injury in vitro.

Fecal microbiota transplantation
Fecal microbiota transplantation (FMT) was performed according to the modi ed method described previously [18]. Brie y, 6-8-week-old male C57BL/6 mice were given antibiotics (vancomycin, 100 mg/kg; neomycin sulfate 200 mg/kg; metronidazole 200 mg/kg; and ampicillin 200 mg/kg) intragastrically once each day for 1 week to deplete the gut microbiota (receptor mice). Preoperatively collected feces of sensitive and resistant mice or low L. murinus patient feces and high L. murinus patient feces (donor mice) were resuspended in PBS at 0.125 g /ml. An amount of 0.1 ml of the solution was administered to mice in the corresponding groups orally via gastric gavage tube 1 week. All the mice had free access to food and water, and mice were performed intestinal ischemia 1 h and reperfusion 2 h surgery after1 weeks of transplantation, and then blood, ileum, kidney, lung and liver samples were harvested in a sterile manner for further examination.

Gene expression analysis
A reverse transcript enzyme (TOYOBO, Tokyo, Japan) was applied to prepare cDNA according to the manufacturer's protocol. The real-time PCR reaction was performed using the ABI Q5 real-Time PCR System with SYBR Green detection protocol (TOYOBO, Tokyo Japan). The expression of target genes in mice were normalized to house-keeping gene 18S using the 2 -^^C T method. The target genes primers were listed in Additional le 2

Protein expression and biochemical analysis
The endotoxin level was measured via commercial kit (GenScript, Nanjing, China). Plasma AST was determined manually with a commercial kit (KeyGene, Nanjing, China). The protein level of IL-10 in the ileum was measured by IL-10 ELISA kit (ABdonal, Wuhan,China).

Histological staining
Ileum, liver, lung and kidney samples tissue were collected and xed in 4% paraformaldehyde. Then, the samples were embedded in para n, 5-μm-thick sections were sliced and were stained with hematoxylineosin (HE) according to the experimental protocol. The degree of ileum injury after reperfusion was evaluated using a modi ed Chiu method [23] according to changes of the intestinal mucosa villus and glands. Liver, lung and kidney tissue histological damage were assessed according to previously described scoring system [24,25,26]. Images were captured at 200 X with Olympus uorescence microscope. (Olympus, Japan).

Immuno uorescence
Para n section samples were generated as above, blocked for1 h, and incubated overnight at 4℃ with anti-ZO-1 antibody (Abcam, Cambridge, MA, USA) and anti-Occludin antibody (Abcam, Cambridge, MA, USA). Tissues were then washed, stained with DAPI for 10 min, and Images were captured at 200 X with uorescent microscopy (Olympus, Japan). Quanti cation of the uorescence intensity of ZO-1 and Occludin staining were performed by automated image analysis in ve randomly chosen 200 X elds of each sample.

Peritoneal Macrophage Collection
Injection of 4 ml of normal saline solution into the peritoneum was used for peritoneal macrophage collection, and the mice's abdomen was gently rubbed for 2 min to make the liquid ow in the abdominal cavity. The peritoneal uid was sucked out and transferred into a centrifugal tube with a glue-head dropper. The amount of each suction was about 4 ~ 5 ml. The collected peritoneal lavage uid was centrifuged at 1000 r / min for 10 min and the supernatant was removed, then further ow experiment was carried out.

Detection of organoid injury by lactate dehydrogenase (LDH) assays
The LDH kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, China) was used to detect the level of LDH in the culture medium to assess organoid damage. The detection of LDH was carried out based on the manufacturers' protocols.

Statistical analysis
All analyses were performed using GraphPad Prism software (version 7.0). Data are presented as means ± standard error of mean (SEM). Statistical analyses were performed using two-sided Student's t tests or by one-way analysis of variance (ANOVA) as indicated in the gure legends. p values <0.05 were considered statistically signi cant.

Page 10/20
Characteristics of tissue injury and the gut microbiota in intestinal Sen and Res mice.
Feces from C57BL/6 mice were collected before intestinal I/R following which the mice underwent intestinal ischemia and the survival rate was observed after reperfusion. Mice that died within 1 h after reperfusion were de ned as Sen mice, while those that survived up to 5 days were de ned as Res mice (Fig. 1a). Compared with the Res mice, the Sen mice showed signi cantly aggravated HE injury and scores in the ileum, liver, lung and kidney as well as up-regulated mRNA levels of IL-1β, IL-6 and TNF-α. (Fig. 1b, c and Additional le 3a). Moreover, the mRNA levels of intestinal tight junction markers in the ileum of the Res mice were markedly higher, whereas, the plasma endotoxin level was reduced compared with those in Sen mice (Fig. 1d, e).
Real-time PCR and 16S DNA sequencing were used to explore whether the susceptibility or resistance of the Sen or Res mice, respectively, to intestinal I/R was related to differences in their gut micro ora. There was a signi cant increase in the relative total bacterial load; the relative abundance of Firmicutes and Bacteroidetes as well as the Firmicutes/Bacteroidetes ratio in the feces of Sen mice were lower than those in the feces of Res mice (Additional le 3b, c). 16S DNA sequencing results showed that the bacterial composition in fecal samples from Sen and Res mice in terms of both bacterial phyla and class were signi cantly different ( Fig. 1f and Additional le 3d). Alpha diversity analysis and principal coordinates analysis (PCoA) indicated that the overall structure of the gut microbiota was signi cantly different between the Sen and Res mice (Fig. 1g, h). To further explore the differences of gut microbial metabolic function in Sen and Res group, the 16S rRNA data were annotated with metabolic pathways from the KEGG data using Tax4Fun prediction analysis. The relative abundances of 20 metabolism related KEGG pathways were shown in Pathway abundance heat map (Additional le 3e). In particularly, the relative KEGG pathways abundances of "Metabolism; Energy Metabolism; Oxidative phosphorylation(ko00190)" and "Environmental Information Processing; Membrane Transport; Bacterial secretion system(ko03070) were statistically upregulated in the Sen group compared with those of the Res group. LEfSe analysis indicated that the higher relative abundance of Coprostanolignes-group and Ruminococcaceae-UCG-010 at the genus level in the Res mice than in the Sen mice was statistically signi cant (linear discriminant analysis [LDA] score > 2, Fig. 1i). Moreover, at the species level, the relative abundance of L. murinus in the Res mice was signi cantly higher than in the Sen mice, and the fold change was the largest (Fig. 1j). The relative abundance of L. murinus was negatively correlated with in ammatory factors of intestinal I/R injury (Fig. 1k).

Gut microbiota from Res mice could independently alleviate I/R-induced tissue damage
To further demonstrate that the gut microbiota plays an important role in intestinal I/R injury, FMT experiment was performed (Fig. 2a). Recipient mouse feces were collected on the third and seventh days after FMT. The mice that received feces from Sen mice (Sen feces group) showed marked decrease in the levels of Firmicutes and increase in the relative total bacterial load compared with the mice that received feces from Res mice (Res feces group) (Fig. 2b, c). The above results demonstrated that the FMT experiment was successful.
Intestinal I/R (60 min/120 min) were performed, after a week of gavage. The Res feces group had signi cantly reduced intestinal I/R-induced HE injury and decreased mRNA levels of IL-1β, TNF-α, Cxcl2, Ccl2 and Ccl5 in the ileum compared with the Sen feces group (Fig. 2d, e). As presented in Fig. 2f-h, the Res feces group had markedly up-regulated mRNA and protein levels of ZO-1 and Occludin in the ileum and reduced plasma endotoxin level compared with the Sen feces group. Further, the Res feces group had signi cantly reduced HE injury and down-regulated mRNA levels of in ammatory factors in the liver, lung and the kidney when compared with the Sen feces group (Fig. 2i-l). The FMT experiment con rmed that the gut microbiome prior to intestinal ischemia is closely related to the postoperative outcome of intestinal I/R.

A single strain of L. murinus alleviates intestinal I/R-induced intestinal injury
From the 16S rRNA gene sequencing analysis, it was found that the relative abundance of L. murinus was four times higher in the Res mice than in the Sen mice (Fig. 1j). Moreover, the relative abundance of this species decreased signi cantly after intestinal I/R, as evidenced by from real-time PCR (Fig. 3a). This prompted us to investigate whether oral administration of L. murinus could ameliorate intestinal I/Rinduced intestinal injury. Treatment with L.murinus signi cantly increased the relative abundance of L. murinus in the cecum, improved survival rate, and reduced HE injury compared with those in I/R mice (Fig. 3a-c). The I/R + L.murinus and sham mice had markedly increased ZO-1 and Occludin mRNA and protein levels compared with I/R mice (Fig. 3d-g). To further demonstrate the role of a single strain of L. murinus in intestinal I/R, mice were treated with ABX and then were gavaged for 7 days with L. murinus before I/R (Supplementary Fig. 1a). Real time PCR revealed that the ABX + L. murinus + I/R group had signi cantly up-regulated relative abundance of L. murinus compared with the ABX + I/R group (Fig. 3h). As presented in Fig. 3i-l, L. murinus reduced HE injury, while increasing ZO-1 and Occludin mRNA and protein levels in the ileum following I/R. The above results demonstrated that L. murinus alleviates intestinal I/R-induced intestinal injury in mice.
Relationship between L. murinus and postoperative intestinal injury in patients undergoing cardiopulmonary bypass surgery.
In patients undergoing cardiac surgery, CPB is potentially responsible for the reduced intestinal ischemia (reduced blood supply and oxygen delivery) and injury; thus, it was used as a clinical model of intestinal I/R in this study [27,28]. Firstly, the relative abundance of L. murinus in preoperative feces of patients undergoing CPB surgery was less than 0.1 and more than 0.1, which were categorized into low and high L. murinus abundance, respectively (Fig. 4a, b). There were no differences in baseline characteristics between the two groups (Additional le 4). Plasma concentrations of citrulline and intestinal fatty acidbinding protein IFABP in both groups were measured as markers of absorptive enterocyte mass and intestinal failure in humans [29,30]. There were no signi cant differences in the levels of citrulline and IFABP between the two groups before surgery (Fig. 4c). Surprisingly, as shown in Fig. 4d-f, the abundance of L. murinus in preoperative feces of patients undergoing CPB surgery was negatively correlated with changes in IFABP concentration and positively correlated with changes in citrulline concentration at 6 hours after operation. The gastrointestinal function of the patients was further evaluated after 7 days post operation, and the occurrence rate of gastrointestinal injury in the low L. murinus abundance group patients was 60%, while the high L. murinus abundance group patients remained clinically devoid of gastrointestinal complications (Fig. 4g).
To further prove the role of L. murinus in ameliorating intestinal injury, FMT experiment was carried out where fecal bacteria from the high and low L. murinus abundance groups of patients collected preoperatively were transplanted into mice separately and were categorized correspondingly into high L. murinus feces group and low L. murinus feces group, respectively. Real-time PCR revealed that the relative abundance of L. murinus in feces of the high L. murinus feces group was markedly higher than that in the low L. murinus feces group after FMT (Fig. 4h). HE staining revealed that the high L. murinus feces group had markedly reduced I/R-induced intestinal damage, but increased mRNA and protein levels of ZO-1 and Occludin in the ileum compared with the low L. murinus feces group (Fig. 4i-k).

Improvement in intestinal I/R injury by L. murinus depends on macrophage participation
The above animal and clinical experimental results have veri ed that L. murinus has a protective effect on in intestinal I/R injury, but the mechanism has not yet been elucidated. L. murinus affects immune response, but whether it regulates macrophages in the intestinal I/R is unclear. Flow cytometry results showed that intestinal I/R induced a signi cant increase in the number of macrophages (F4/80 + CD45 + ), which was reversed by L. murinus treatment (Fig. 5a). To further verify whether the protective effect of L. murinus is related to macrophages, clodronate liposomal was administered intraperitoneally to mice two days before I/R to deplete macrophages ( Supplementary Fig. 1b). The Clodronate-Lipo + I/R group exhibited decreased the number of macrophages (Fig. 5b) and reduced I/R-induced HE injury (Fig. 5c, d), as well as up-regulated mRNA and protein levels of ZO-1 and Occludin (Fig. 5c, e-f) compared with the I/R mice and the PBS + Lipo + I/R group, but showed no statistical difference when compared with the Clodronate-Lipo + L. murinus + I/R group. Hypoxia-reoxygenation (H/R) model of small intestinal organoids was established to further con rm the results of in vivo experiments. As shown in the Fig. 5g, h, L. murinus reduced morphological damage of intestinal organoids injury and LDH levels, but increased the mRNA levels of ZO-1 and Occludin (Fig. 5i) in the co-culture system of macrophages and organoids following H/R, but had no obvious effect on these data in the organoids cultured alone. Consistent with the results of in vivo experiments, the protective effect of L. murinus on intestinal I/R injury may depend on the participation of macrophages.
L. murinus promotes the release of IL-10 from M2 macrophages through activating TLR2 signaling to alleviate intestinal I/R injury As shown in the Fig. 6a, ow cytometry results showed that intestinal I/R induced a signi cantly decreased the proportion of M2 macrophages (CD11C − CD206 + %) in peritoneal, which was reversed by L. murinus treatment in WT mice. In addition, intestinal I/R induced a signi cant increase in the expression of TLR2 on the surface of M2 macrophages, which was enhanced by L. murinus treatment. TLR2 −/− mice was used to explore the effect of TLR2 on the surface of M2 macrophages in the protection of L. murinus on intestinal I/R injury. L. murinus signi cantly increased the proportion of M2 macrophages and TLR2 + M2 macrophages/total M2 macrophages in WT mice following intestinal I/R, but not in TLR2 −/− mice (Fig. 6a). Compared with in the I/R group, L. murinus signi cantly reduced HE injury and scores in WT mice, but in TLR2 −/− mice, the protective effect of L. murinus on intestinal I/R injury disappeared (Fig. 6b). Consistent with the changes in TLR2 expression on the surface of macrophages, the mRNA levels ZO-1 and Occludin (Fig. 6c, d) in the ileum of mice in the I/R + L. murinus group were higher than those in the ileum of mice in the I/R group in WT mice, but not in TLR2 −/− mice. Furthermore, the proportion of IL-10 + M2 macrophages/total M2 macrophages in peritoneal, the mRNA and protein levels of IL-10 in the ileum of mice in the I/R + L. murinus group were higher than those of mice in the I/R group in WT mice, but not in TLR2 −/− mice (Fig. 6e-g). These indicated that L. murinus reduced intestinal I/R injury through TLR2, accompanied by an increase in the proportion of IL-10 + M2 macrophages.
Then IL-10 −/− mice were used to explore the role of IL-10 in the protection offered by L. murinus in intestinal I/R injury. We found that the mRNA and protein levels of IL-10 in the ileum of the I/R + L. murinus group were higher than those in the I/R group in WT mice, while the mRNA and protein levels of IL-10 in IL-10 −/− mice were drastically reduced (Fig. 6h, i). Compared to the I/R group, L. murinus treatment and rmIL-10 treatment signi cantly reduced HE injury and scores, as well as increased the mRNA levels of ZO-1 and Occludin in WT mice. However, L. murinus treatment had no obvious effect on HE injury and scores, the mRNA levels of ZO-1 and Occludin in IL-10 −/− mice following intestinal I/R (Fig. 6j, k). The above results indicated that L. murinus promotes the release of IL-10 from macrophages through TLR2 to alleviate intestinal I/R injury in mice.
L. murinus alleviates organoid H/R injury through promoting the release of IL-10 from macrophages via TLR2 signaling.
As we showed that the protective effect of L. murinus on organoid injury depends on the participation of macrophages (Fig. 5g-i), we decided to explore the speci c mechanism of macrophages in small intestinal organoids. A transwell co-culture system of macrophages extracted from TLR2 −/− mice (TLR2 −/ − -MΦ) or IL-10 −/− mice (IL-10 −/− -MΦ) or WT mice (WT-MΦ) and organoid extracted from WT mice (WT-Org) was established. It was uncovered that L. murinus treatment increased the proportion of M2 macrophage, the expression of TLR2 on the surface of M2 macrophages and the mRNA level of Myd88 compared with H/R group in the co-culture system of WT-MΦ and WT-Org, but not in the co-culture system of TLR2 −/− -MΦ and WT-Org (Fig. 7a-c). L. murinus reduced organoids HE injury and LDH levels, while increased the mRNA and protein levels of ZO-1 and Occludin in the co-culture system of WT-MΦ and WT-Org, but not in the co-culture system of TLR2 −/− -MΦ and WT-Org ( Fig. 7d-g).In addition, ow cytometry results showed that L. murinus suppressed the decrease in the proportion of IL-10 + M2 macrophages/total M2 macrophages induced by H/R in the transwell co-culture system of WT-MΦ and WT-Org, but not in the transwell co-culture system of TLR2 −/− -MΦ and WT-Org, IL-10 −/− -MΦ and WT-Org (Fig. 7h). The co-culture system of IL10 −/− -MΦ and WT-Org was established to explore the role of IL-10 released by macrophages in the protective effect of L. murinus on organoid H/R injury. L. murinus reduced organoids HE injury and LDH levels, while increased the mRNA and protein levels of ZO-1 and Occludin in the co-culture system of WT-MΦ and WT-Org during H/R, but not in that of IL-10 −/− -MΦ and WT-Org (Fig. 7i-l). The above results indicated that L. murinus promotes the release of IL-10 from M2 macrophages through TLR2 to alleviate organoid H/R injury in the co-culture system of macrophages and organoids.

Discussion
In the present study, we for the rst time con rmed that mice with the same genetic background had huge differences in sensitivity to intestinal I/R injury, which, at least partly, results from the difference in the gut microbiome in the feces of mice before intestinal ischemia. Furtherly, our results revealed that the relative abundance of L. murinus at the species level in the feces from the Res mice before intestinal ischemia was signi cantly higher than that in the Sen mice, and was negatively correlated with the levels of in ammatory factors following intestinal I/R. Clinical evidence also showed that the abundance of L. murinus in preoperative feces of patients undergoing CPB surgery was closely related to the degree of intestinal I/R injury after surgery. FMT indicated that the abundance of L. murinus in preoperative feces of patients is related to different sensitivities of mice to intestinal I/R injury. More importantly, treatment with L. murinus signi cantly prevented from intestinal I/R-induced intestinal injury and improve the survival of mice, which depends on the participation of macrophages. Another important nding of the current study was that L. murinus promotes the release of IL-10 from M2 macrophages through activating TLR2 signaling to alleviate intestinal I/R injury, which shed light on a novel mechanism of intestinal I/R injury and new therapeutic strategy for clinical practice.
It is well known that patients suffering from shock or undergoing some surgical procedures including CPB surgery have different outcome after receiving medical treatment. The reason for the difference has remained being unclear. In the present study, we employed the mouse model of intestinal I/R and identi ed the difference that mice with the same genetic background have signi cant differences in sensitivity to intestinal I/R injury. Especially, 16S DNA sequencing and FMT results showed that gut microbiome is involved in postoperative outcome of intestinal I/R. This is a novel nding.
Gut microbiota is a complex ecosystem susceptible to the surrounding environment and diet. At present, only a few studies have reported the changes of gut microbiota in intestinal I/R by the traditional DGGE method [31]. Previous antibiotic experiments have shown that depletion of gut commensal bacteria can attenuate intestinal I/R injury [32]. However, the extensive diversity of the gut microbiome makes it di cult to precisely determine whether speci c microbes are associated with I/R development or progression. Here, we for the rst time found that the abundance of L. murinus was signi cantly increased in the Res mice compared with the Sen mice, and the fold change was the largest. L. murinus strains have previously been isolated and identi ed from rat, mice, porcine and canine species, and humans [33,34]. Few reports have indicated the application of L. murinus in host health and disease. It has been shown that L. murinus may be used as a potential probiotic to reduce the incidence of delayed sepsis in neonates [35]. In addition, another study showed that L. murinus HU-1 improved the abnormal neural behavior of offspring mice caused by maternal dysregulation [36]. Further, treatment of mice with L. murinus prevented salt-sensitive hypertension by modulating Th17 cells [22]. Furthermore, L. murinus improved colitis by inducing Treg cell differentiation in colitis model [7]. Various L. murinus strains have been further characterized as potential probiotics in the food formulation industry. With increased public interest in L. murinus containing probiotics, the impacts of L. murinus on intestinal I/R injury are beginning to be unraveled.
At present, macrophages are considered to be a potential therapeutic intervention target for in ammatory diseases and cancer [37]. The switch from M1 to M2 phenotype may reduce intestinal I/R injury [23], and M1 polarization of liver macrophages may be one of the mechanisms of intestinal I/R-induced hepatic injury in mice [38]. However, the mechanisms underlying the speci c role of macrophages in intestinal I/R are not fully elucidated. Clodronate-liposome was able to eliminate macrophages and had a certain protective effect on the intestinal morphological damage in the intestinal I/R mouse model, consistent with the results reported in the literature [39]. In addition, L. murinus reduced morphological damage of intestinal organoids injury and LDH levels in the co-culture system of macrophages and small intestinal organoids following H/R, but had no obvious effect on these data in the organoids cultured alone, which showed that the protective effect of L. murinus on organoid injury depends on the participation of macrophages. 3D intestinal organoid culture was rst established in 2009 [40]. Compared with simple intestinal epithelial cell lines, organoids have the physiology of natural intestinal epithelium and functional diversity [15,41,42]. Compared with in vivo experiments, organoid systems in vitro avoided the interference of multiple complex factors in vivo and thus are more convenient and accurate. Here, we established an intestinal organoid H/R model in vitro to simulate intestinal I/R injury and con rmed that it is reliable for studying the mechanism and treatment of intestinal I/R injury.
In our study, although L. murinus treatment signi cant decreased in the number of macrophages, compared with I/R in WT mice, L. murinus reduced the proportion of M1 macrophages and increased the M2 polarization of macrophages. In addition, L. murinus signi cantly increased the proportion of M2 macrophages and TLR2 + and IL-10 + M2 macrophages/total M2 macrophages in WT mice following intestinal I/R, but not in TLR2 −/− mice. The reported role of TLR2/Myd88 signaling in intestinal I/R injury are controversial. Compared with WT mice, TLR2 −/− mice have a dysregulated mucosal innate immune response and fail to produce a protective response after intestinal I/R [43]. However, some groups found that TLR2 −/− mice have less intestinal damage and in ammation compared with WT mice [44,45]. In this study, compared with WT mice, TLR2 −/− mice fail to produce a protective response after I/R. In addition, Liang et al. revealed that bi dobacterial and lactobacilli induced macrophages to secrete IL-10 by activating TLR2 and MyD88 pathways, conferring a protective effect in hosts suffering from in ammation diseases [46]. Another group found that TLR2-mediated secretion of IL-10 and immune suppression in response to phagosome-con ned Listeria monocytogenes [47]. Clostridium butyricum directly triggered IL-10 production by intestinal macrophages in in amed mucosa via the TLR2/MyD88 pathway, thereby preventing experimental colitis in mice [48]. In this study, we found that L. murinus alleviates intestinal I/R injury through promoting the release of IL-10 from M2 macrophages via activating TLR2 signaling, which shed light on a novel mechanism of intestinal I/R injury.
Although, differences in the gut microbiota at the individual level may lead to differences in the severity of intestinal I/R injury in patients, the gut microbiome is a complex system where different strains may act together to affect the intestine, and thus there is no clear evidence that one individual strain is the most bene cial. Our pilot study on patients was conducted on a limited number of individuals and needs to be validated in a larger cohort. Moreover, metagenomic and metabolomic analyses and clinical studies need to be performed to con rm the involvement of L. murinus in intestinal I/R injury and develop effective probiotics in the future.

Conclusions
Taken together, our results suggest that gut microbiome is involved in postoperative outcome of intestinal I/R. Furthermore, the present results reveal that L. murinus alleviates intestinal I/R injury through promoting the release of IL-10 from M2 macrophages via activating TLR2 signaling, which shed light on a novel mechanism of intestinal I/R injury and suggest that the therapy via targeting microbiome, TLR2 and IL-10 is a promising strategy to prevent intestinal I/R injury.

Declarations
Ethics approval and consent to participate: The study protocol was approved by the Ethical Committee of Nanfang hospital, Southern Medical University (approval number NFEC-202009-k2-01), and informed consent was obtained from participants. Consent for publication: Not applicable.
Availability of data and material: The raw sequencing data generated in this study have been deposited in NCBI Sequence Read Archive (http://www.ncbi.nim.nih.gov/sra) under the accession number PRJNA661144. All other data associated with this study are present in the paper or Supplementary Materials.