Association of peripheral blood-derived mesenchymal stem cells exosomal circular RNAs with diagnosis and inflammatory response in acute mesenteric ischemia

Abstract

BACKGROUND

Acute mesenteric ischemia (AMI) is associated with high mortality owing to delayed diagnosis and the absence of biomarkers capable of distinguishing disease subtypes or inflammatory burden. Mesenchymal stem cells (MSCs) exosomal circular RNAs (circRNAs) show diagnostic potential in other diseases.

AIM

To evaluate the utility of MSCs exosomal circRNAs in AMI, particularly for early detection and subtyping.

METHODS

Peripheral blood-derived MSCs (PBMSCs) from 80 patients with AMI (stratified by etiology: 48 arterial AMI, 32 venous AMI, 36 early reversible, 44 late necrotic) and 125 controls were obtained at initial hospital admission. Exosomal circRNAs were isolated via ultracentrifugation, validated by transmission electron microscopy and nanoparticle tracking analysis, and quantified by quantitative real-time polymerase chain reaction. Plasma intestinal fatty acid binding protein (IFABP), d-lactate, interleukin-6 (IL-6), tumor necrosis factor-α, IFABP, D-lactate, IL-6, and neutrophil-to-lymphocyte ratio were measured. Pearson’s test was used to assess the diagnostic performance of receiver operating characteristic analysis.

RESULTS

Patients with AMI exhibited significantly elevated PBMSCs exosomal circ-Eya3 (P < 0.001) and reduced circEZH2_005 levels (P < 0.001). Circ-Eya3 correlated positively with IFABP (r = 0.606), D-lactate (r = 0.384), IL-6 (r = 0.551), and neutrophil-to-lymphocyte ratio (r = 0.601) (all P < 0.001), whereas circEZH2_005 showed inverse correlations (r = -0.580 to -0.403; P < 0.001). Critically, circEZH2_005 expression was lower in arterial AMI than in venous AMI (P = 0.003). The combination of circ-Eya3, circEZH2_005, and D-lactate achieved excellent diagnostic accuracy (area under the curve = 0.891). Post-reperfusion, circ-Eya3 increased (P < 0.05), while circEZH2_005 decreased (P < 0.05).

CONCLUSION

PBMSCs exosomal circ-Eya3 and circEZH2_005 serve as novel biomarkers for AMI, reflecting intestinal injury severity, systemic inflammation, and disease subtype. Their integration with D-lactate enables high-accuracy diagnosis, whereas differential expression across arterial/venous AMI and ischemia stages offers clinical utility for guiding intervention strategies.

Key Words: Peripheral blood-derived mesenchymal stem cells; Exosome; Circular RNAs; Acute mesenteric ischemia; Diagnosis; Inflammatory factors

Core Tip: Peripheral blood-derived mesenchymal stem cells exosomal circ-Eya3 and circEZH2_005 serve as novel biomarkers for acute mesenteric ischemia (AMI), reflecting intestinal injury severity, systemic inflammation, and disease subtype. Their integration with D-lactate enables high-accuracy diagnosis, whereas differential expression across arterial/venous AMI and ischemia stages offers clinical utility for guiding intervention strategies. Peripheral blood-derived mesenchymal stem cells exosomal circ-Eya3 and circEZH2_005 are promising biomarkers for the early diagnosis, subtyping, staging, and post-reperfusion monitoring of AMI.

INTRODUCTION

Early treatment of acute mesenteric ischemia (AMI) contributes to the prevention of linked short bowel syndrome and death[1,2]. According to recent findings, patients may have better results if they receive a diagnosis and standardized multidisciplinary expert care early on[3,4]. Tools to obtain an instant diagnosis are urgently needed in order to do this. Despite advancements in imaging, endovascular techniques, and critical care medicine, the rates of mortality and intestinal resection have not changed over the past decades. In fact, unlike advanced AMI with permanent transmural necrosis, early AMI is completely reversible[5-7]. However, the lack of specificity in acute abdominal pain in AMI patients makes it difficult to identify and raise clinical suspicions[8,9], which frequently results in missed or delayed diagnosis and treatment. Furthermore, if suspicion is not aroused, contrast-enhanced computed tomography (CT) of the acute abdomen may underdiagnose AMI[10]. Therefore, it is of great significance to identify precise biomarkers for the early diagnosis of AMI to address the current challenges.

Mesenchymal stem cells (MSCs) are a type of adult stem cells that possess the potential for multi-directional differentiation, immune regulatory functions, and tissue repair capabilities[11]. MSCs possess anti-inflammatory and immunomodulatory functions and have been utilized in the diagnosis and treatment of various inflammatory diseases[12]. The latest evidence indicates that MSCs transplantation can enhance the intestinal repair ability and promote the regeneration of the small intestinal mucosa. These studies have made it a new direction for the diagnosis and treatment of intestinal diseases such as AMI[13]. The paracrine effects of MSCs and the secretion of important extracellular vesicle are even more the focus of attention in the medical field.

Exosomes, minute vesicles secreted by cells, display a diameter typically ranging between 35 nm and 100 nm[14]. These exosomes are replete with proteins, nucleic acids, and a diverse array of other biomolecules[14-16]. Much evidence demonstrated that molecule carried by exosomes indicates the occurrence and progression of many diseases, such as cancers[17] and neurodegenerative diseases[18]. Circular RNAs (circRNAs) are a class of non-coding RNA molecules with a covalent closed-loop structure and do not contain 5’ to 3’ polar or polyadenylate tails[19], which were also carried by exosomes. CirRNA molecules are highly abundant and show specific expression patterns in cell types, tissues, and developmental stages[20]. Additionally, circRNAs are closely associated with a variety of physiological and pathological processes, such as cell survival, growth, and differentiation, and have important regulatory roles in these biological processes[21,22]. A recent study on intestinal ischemia/reperfusion injury has reported that circARHGAP12 inhibits DNA damage and cell apoptosis of intestinal epithelial cells by regulating MDC1[23]. Another research reported the results of intestinal exosomal circRNA sequencing of intestinal ischemia/reperfusion injury, and the top five differentially expressed circRNAs circ-Eya3, circ-Lbr, circEZH2_005, circ-Tmem267, and circ-Herc3 were detected in mice with intestinal ischemia/reperfusion injury[24]. However, whether these five circRNAs were differentially expressed in peripheral blood-derived MSCs (PBMSCs) exosomes of AMI patients was unknown and worthy of further exploration.

Therefore, this study aimed to evaluate the potential diagnostic utility of PBMSCs exosomal circ-Eya3 and circEZH2_005 levels in distinguishing AMI patients from controls, to assess the correlations of these circRNAs with traditional biomarkers or inflammatory factors, to investigate the potential differential expression patterns of circ-Eya3 and circEZH2_005 among clinically relevant AMI subgroups, and to profile their dynamic expression changes at various time points before and after surgery.

MATERIALS AND METHODS

Study design

This observational study analyzed archived samples and clinical data from patients treated between December 30, 2024, and September 1, 2025, at the First Hospital of Jiaxing. The cohort included 80 surgically managed patients with AMI and 125 non-AMI patients with abdominal pain. The study was approved by the Ethics Committee of First Hospital of Jiaxing (Approval No. 2024-LP-159). All participants provided written informed consent.

Patients and controls

The patients included in the study were those who presented with symptoms such as abdominal pain and came to our center for enhanced CT examination. According to the previous guidelines and standards, the definition of AMI is as follows: (1) The manifestations of intestinal injury in patients were evaluated through clinical diagnosis, biological tests and enhanced CT scans; (2) The display of functional disorders of the peritoneal trunk artery, superior mesenteric artery/vein; and (3) No other diseases with similar symptoms have been diagnosed[25]. Multimodal and multidisciplinary treatment approaches are employed for AMI patients as their treatment methods. In simple terms, the intervention for patients involves the use of antibiotics and antithrombotic drugs. If an emergency occurs, an AMI reconstruction surgery will be performed on the patient[26]. If AMI cannot be repaired, an open surgical procedure will be used for treatment[27]. As the control group for AMI patients, other patients who experienced abdominal pain but whose diagnosis was not AMI were also included in the study. AMI is classified into the following three types based on the nature of the lesion: Arteriogenic AMI, venous AMI and late necrotic or early AMI subgroups. In this study, irreversible AMI is defined as follows: (1) The pathological diagnosis shows necrosis with features of hypoxia, being extensive and involving the entire layer. At the same time, there are hemorrhagic and gangrenous infarctions; (2) The CT scan results indicated that the patient had a bowel perforation; and (3) During the surgery, extensive necrosis was discovered[27].

Data collection

Upon admission, clinical records including sex, age, body mass index, cardiovascular history, and serological indicators were collected. Peripheral venous blood was drawn from all enrolled patients at initial hospital admission using standardized EDTA-containing vacutainers. The samples were immediately processed via centrifugation at 1500 × g for 15 minutes at ambient temperature. The resultant plasma supernatant was aliquoted into cryovials and preserved at -80 °C in ultralow-temperature freezers until subsequent biomarker assays. Preoperative hematological and inflammatory biomarkers were systematically documented for analytical purposes, encompassing: (1) Absolute neutrophil; (2) Absolute lymphocyte; and (3) Platelet counts. The neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio were calculated to investigate their potential correlation with intestinal ischemia. The most recent preoperative blood sample obtained 12 hours prior to surgical intervention was utilized for each patient.

Enzyme-linked immunosorbent assay

Diluent plasma collected within 12 hours prior to surgical intervention was added to the sample wells. Following the manufacturer’s instructions for human ELISA kits (Solarbio, Beijing, China), the levels of intestinal fatty acid binding protein (IFABP), D-lactate, interleukin-6 (IL-6), and tumor necrosis factor-α (TNF-α) in plasma were determined at OD 450 nm using a spectrophotometer (Thermo Fisher Scientific, Waltham, MA, United States).

Isolation and culture of PBMSCs

Stem cells were isolated from peripheral blood using the Rvy kit through an automatic blood cell separator (FRESENIUS KABI, Baden-Hoewerk, Germany). MSCs were further sorted by flow cytometry after cells were labeled with CD73 and CD105 (Abcam, Cambridge, United Kingdom). The obtained cells were cultured in DMEM medium (Thermo Fisher Scientific, Waltham, MA, United States) containing 10% foetal bovine serum and maintained at 37 °C, 5% CO₂ and 95% humidity.

Exosome isolation

The exosome isolation process was performed in accordance with the instructions of the exosome isolation kit (QIAGEN, Duesseldorf, Germany). First, the PBMSCs were centrifuged at 2000 × g at 4 °C for 20 minutes to remove large cell fragments or debris, the resulting supernatants were collected and centrifuged at 12000 × g at 4 °C for 45 minutes to remove smaller cellular debris, and the supernatants were transferred to a new tube and ultracentrifuged at 120000 × g at 4 °C for 120 minutes to pellet the small vesicles. The pellets were then resuspended in a sufficient amount of phosphate buffered saline (PBS, Thermo Fisher Scientific, Waltham, MA, United States) and filtered through a filter with a particle size of roughly 0.22 μm to remove any possible impurities. The pellets produced by this procedure were resuspended in PBS following a second cycle of ultracentrifugation under the same conditions as previously described (120000 × g and maintained at a temperature of approximately 4 °C). Finally, a bicinchoninic acid test kit (MeilunBio, Dalian, Liaoning Province, China) was used to measure the protein quantities found in these separated exosomes.

Transmission electron microscope

After separation, exosomes were placed on a formvar-coated copper grid and left in 2% paraformaldehyde (Sigma, Darmstadt, Germany) for ten minutes. The exosomes were then embedded in a solution containing 0.4% uranyl acetate (Sigma-Aldrich, Darmstadt, Germany) and 0.13% methylcellulose (Sigma-Aldrich, Darmstadt, Germany). The grid was then examined using electron microscopy (HITACHI, Tokyo, Japan) running at 80 V, and pictures were taken while doing so.

Nanoparticle tracking analysis

A Nanoparticle Tracking Analyzer (Zeta View, Particle Metrix, Dusseldorf, Germany) was used to measure the size distribution of the exosomes. Between 50 and 400 exosome samples were used and diluted with 1 × PBS to the appropriate concentration. The size of the exosomes was measured in accordance with the instrument parameters.

Western blot

The characterization of exosomes was confirmed by the presence of exosomal protein markers by western blotting using the following antibodies: Anti-CD63 (1:2000, Abcam, Cambridge, United Kingdom) and anti-tumor susceptibility gene-101 (1:1000, Abcam, Cambridge, United Kingdom). RIPA lysis solution (Thermo Fisher Scientific, Waltham, MA, United States) containing proteinase inhibitors was used to lyse exosomes. The lysates were centrifuged at 16100 × g at 4 °C for 15 minutes after incubation on ice for 30 minutes. A total of 40 μg of total protein was successively run on 10% sodium-dodecyl sulfate gel electrophoresis gels after determining the protein concentration using the bicinchoninic acid technique. The proteins were then transferred onto polyvinylidene fluoride membranes and blocked with 5% skim milk for one hour at room temperature. After overnight incubation at 4 °C with the primary antibodies, the membranes were cleaned using TBST solution containing 1% Tween-20 and then incubated for two hours at room temperature with secondary antibodies. The SuperSignal West Pico Chemiluminescent Substrate Kit (Thermo Fisher Scientific, Waltham, MA, United States) was used to identify the protein bands.

Quantitative real-time polymerase chain reaction analysis

TRIzol reagent (Thermo Fisher Scientific, Waltham, MA, United States) was used to extract total RNA from the exosomes (50 μL). Subsequently, 1 μg of the total RNA was subjected to reverse transcription of cDNA using a HiCapacity RT kit (Thermo Fisher Scientific, Waltham, MA, United States). The SYBR Green detection protocol (TOYOBO, Osaka, Japan) was used for quantitative real-time polymerase chain reaction (qRT-PCR) analysis using the ABI 7500 Real-Time PCR System (Thermo Fisher Scientific, Waltham, MA, United States). Relative RNA expression levels were calculated using the 2^-ΔΔCt method and normalized to GAPDH. The primer sequences used were as follows: For circEZH2_005, the forward primer was 5’-TTTAGTTCCAATCGTCAG-3’, and the reverse primer was 5’-CTCAGATTTCTTCCCAGT-3’; for circ-Eya3, the forward primer was 5’-GACAGACTCAATACCAGACACTACAGC-3’, and the reverse primer was 5’-TCTTGCTC TTCTTCCATGAGGACC-3’; for circ-Lbr2, the forward primer was 5’-CTGCTCCACTTCCCTCCA-3’, and the reverse primer was 5’-CACGCTGACGCTGTTTCC-3’; for circ-Herc3, the forward primer was 5’-ACAACAGGTGGCAGTCAA-3’, and the reverse primer was 5’-ACAAGGTGGTTCCTACGG-3’; for circ-Tmem267, the forward primer was 5’-ACACACA CAAGCATCTACTTTGA-3’, and the reverse primer was 5’-AGAGGACTGCACATTCACTGT-3’; for GAPDH, the forward primer was 5’-GCTCTCTGCTCCTCCTGTTC-3’, and the reverse primer was 5’-GACTCCGACCTTCACCTTCC-3’.

Statistical analysis

Data analysis was performed using GraphPad Prism 8.4 (GraphPad, Inc., La Jolla, CA, United States) and SPSS Statistics 27.0 software (IBM Corp., Armonk, NY, United States). Pearson’s χ² test was used to assess the associations between discrete variables. Pearson’s correlation coefficient was used to analyze the correlation between the two continuous variables. Continuous variables with normal distribution were subjected to Student’s t-test. Non-normally distributed variables were analyzed using the Mann-Whitney U test. Subgroup comparisons were performed using the Kruskal-Wallis test for non-parametric distributions. Upon achieving statistical significance (P < 0.05) in global testing, pairwise post hoc analyses were conducted with Bonferroni adjustment. One-way analysis of variance was used to analyze differences among multiple groups. A receiver operating characteristic (ROC) curve was used to evaluate diagnostic value (specificity and sensitivity). All statistical tests were two-tailed, and a P value less than 0.05 was considered statistically significant.

RESULTS

The characteristics of subjects

We enrolled 80 patients with AMI and 125 healthy controls in this study. Compared AMI group to control group, there were differences in age, body mass index, and history of atrial fibrillation (Table 1, all P < 0.001). Among AMI patients, 48 were ranked as having arterial AMI, while 32 were diagnosed with venous AMI (Table 2). Additionally, there were 36 patients diagnosed with early AMI, and another 44 patients with necrotic AMI (Table 2). NLR and plasma levels of IFABP, D-lactate, and IL-6 were different between patients with AMI and controls (Table 3, all P < 0.001).

Table 1 Baseline data.

Variables	AMI group (n = 80)	Control group (n = 125)	P value
Gender (male/female)	51/29	88/37	0.320
Age	62.850 ± 6.963	52.590 ± 10.798	< 0.001
BMI	28.530 ± 4.927	22.280 ± 1.604	< 0.001
Cardiovascular history
Arterial hypertension	37	44	0.114
Atrial fibrillation	26	8	< 0.001
Coronary heart disease	14	19	0.662
Cerebral infarction	20	20	0.113
Clinical symptoms
Vomiting	36	53	0.714
Diarrhea	16	26	0.890
Hematochezia	28	37	0.418

AMI: Acute mesenteric ischemia; BMI: Body mass index.

Table 2 Acute mesenteric ischemia subgroups.

Variables	AMI patients (n = 80)
Cause of vascular insufficiency
Arterial stenosis	16
Arterial thrombosis	22
Arterial embolus	10
Mesenteric vein thrombosis	9
Portal vein thrombosis	7
Mesenteric vein compression	6
Portal vein compression	10
Stage
Early AMI	36
Necrotic AMI	44

AMI: Acute mesenteric ischemia.

Table 3 Preoperative biochemical and inflammatory variables.

Variables	AMI group (n = 80)	Control group (n = 125)	P value
Biochemical variables
Neutrophils count (× 109)	8.480 ± 5.346	7.692 ± 6.603	0.408
Platelets count (× 109)	264 ± 42	221 ± 23	0.395
Lymphocytes count (× 109)	1.302 ± 2.944	1.024 ± 3.750	0.502
NLR	10.124 ± 2.724	6.086 ± 1.702	< 0.001
PLR	195.537 ± 169.450	178.056 ± 153.799	0.466
IFABP, pg/mL	615.762 ± 348.743	447.806 ± 185.106	< 0.001
D-lactate, nmol/mL	446.387 ± 129.760	274.089 ± 97.689	< 0.001
Inflammatory variables
IL-6, pg/mL	24.5585 ± 18.742	4.876 ± 4.332	< 0.001
TNF-α, pg/mL	86.630 ± 82.076	70.258 ± 56.476	0.118

AMI: Acute mesenteric ischemia; NLR: Neutrophil-to-lymphocyte ratio; PLR: Platelet-to-lymphocyte ratio; IFABP: Intestinal fatty acid binding protein; IL: Interleukin; TNF: Tumor necrosis factor.

The high expression of circ-Eya3 and the low expression of circEZH2_005 in PBMSCs exosomes of AMI patients

Some researchers have reported five differentially expressed circRNAs in PBMSCs exosomes from mice with intestinal I/R.21 We extracted PBMSCs exosomes from patients and controls to verify the expression of these five circRNAs. Transmission electron microscopy and nanoparticle tracking analysis characterization results demonstrated that exosomes derived from patients with AMI and controls had similar morphologies and particle sizes (Figure 1A and B). Western blot analysis revealed the surface characteristic proteins tumor susceptibility gene-101 and CD63 in the exosomes (Figure 1C). These results suggested the success of exosome extraction. qRT-PCR results showed that circ-Tmem267 and circ-Lbr had no detection, circ-Herc3 was not significantly different between AMI group and control group in PBMSCs exosomes (Figure 2A, P > 0.05). In contrast, circ-Eya3 was highly expressed, while circEZH2_005 was lowly expressed in exosomes derived from AMI patients compared with exosomes from controls (Figure 2B and C, all P < 0.001). These results verified that patients with AMI had increased PBMSCs exosomal circ-Eya3 and decreased PBMSCs exosomal circEZH2_005.

Figure 1 The characteristics of peripheral blood-derived mesenchymal stem cells exosomes from acute mesenteric ischemia patients and controls. A: Transmission electron microscopy characterization of exosomes isolated from peripheral blood-derived mesenchymal stem cells of acute mesenteric ischemia patients and controls (scale bar: 100 nM); B: Nanoparticle tracking analysis characterization of exosomes isolated from peripheral blood-derived mesenchymal stem cells of acute mesenteric ischemia patients and controls; C: Expression levels of two exosome marker proteins (tumor susceptibility gene 101 and CD63) were measured by western blot. AMI: Acute mesenteric ischemia; TSG101: Tumor susceptibility gene 101.

Figure 2 The expression levels of circ-Herc3, circ-Eya3, and circEZH2_005 in peripheral blood-derived mesenchymal stem cells exosomes from acute mesenteric ischemia patients and controls. A-C: Quantitative real-time polymerase chain reaction was used to measure the circ-Herc3 (A), circ-Eya3 (B), and circEZH2_005 (C) levels in peripheral blood-derived mesenchymal stem cells exosomes of different groups. ^cP < 0.001. NS: No significant; AMI: Acute mesenteric ischemia.

Correlations between PBMSCs exosomal circ-Eya3 and circEZH2_005 with plasma IFABP and D-lactate

IFABP and D-lactate levels were correlated with AMI[28]. As shown, IFABP and D-lactate levels were lower in AMI patients than in controls (Figure 3A and B, P < 0.001). The Pearson correlation coefficient showed that PBMSCs exosomal circ-Eya3 was positive correlated with plasma IFABP levels (Figure 3C, r = 0.521, P < 0.001), and was also positively correlated with plasma D-lactate levels (Figure 3D, r = 0.347, P < 0.01). In addition, PBMSCs exosomal circEZH2_005 was negatively correlated with PBMSCs IFABP levels (Figure 3E, r = -0.360, P < 0.01) but not with plasma D-lactate levels (Figure 3F, r = -0.123, P > 0.05). These results show that exosomal circ-Eya3 was positively correlated with plasma IFABP and D-lactate, while exosomal circEZH2_005 was negatively correlated with plasma IFABP.

Figure 3 Correlations between peripheral blood-derived mesenchymal stem cells exosomal circ-Eya3 and circEZH2_005 with plasma intestinal fatty acid binding protein and D-lactate. A and B: The levels of plasma intestinal fatty acid binding protein (IFABP) (A) and D-lactate (B) in acute mesenteric ischemia (AMI) patients and controls were determined by enzyme-linked immunosorbent assay; C and D: Correlations between the levels of peripheral blood-derived mesenchymal stem cells exosomal circ-Eya3 with plasma IFABP (C) and D-lactate (D) in AMI patients; E and F: Correlations between the levels of peripheral blood-derived mesenchymal stem cells exosomal circEZH2_005 with plasma IFABP (E) and D-lactate (F) in AMI patients. ^cP < 0.001. IFABP: Intestinal fatty acid binding protein; AMI: Acute mesenteric ischemia.

Correlations between PBMSCs exosomal circ-Eya3 and circEZH2_005 with NLR and IL-6

Compared with the controls, the NLR in patients were significantly increased (Figure 4A, P < 0.001). The Pearson correlation coefficient showed that PBMSCs exosomal circ-Eya3 was positive correlated with NLR (Figure 4B, r = 0.601, P < 0.001), and circEZH2_005 was negatively correlated with NLR (Figure 4C, r = -0.580, P < 0.001). In addition, higher levels of IL-6 were observed in patients with AMI than in controls (Figure 4D, P < 0.001). PBMSCs exosomal circ-Eya3 positively correlated with plasma IL-6 levels (Figure 4E, r = 0.551, P < 0.001), and circEZH2_005 negatively correlated with plasma IL-6 levels (Figure 4F, r = -0.532, P < 0.001). These results showed that exosomal circ-Eya3 was positively correlated with NLR and IL-6, whereas exosomal circEZH2_005 was negatively correlated with NLR and IL-6.

Figure 4 Correlations between peripheral blood-derived mesenchymal stem cells exosomal circ-Eya3 and circEZH2_005 with neutrophil-to-lymphocyte ratio and interleukin-6. A: The levels of neutrophil-to-lymphocyte ratio in acute mesenteric ischemia (AMI) patients and controls were calculated by the ratio of neutrophils count/Lymphocytes count; B and C: Correlations between the levels of neutrophil-to-lymphocyte ratio with peripheral blood-derived mesenchymal stem cells exosomal circ-Eya3 (B) or circEZH2_005 (C) in AMI patients; D: The levels of interleukin-6 in AMI patients and controls were determined by enzyme-linked immunosorbent assay; E and F: Correlations between the levels of interleukin-6 with peripheral blood-derived mesenchymal stem cells exosomal circ-Eya3 (E) or circEZH2_005 (F) in AMI patients. ^cP < 0.001. NLR: Neutrophil-to-lymphocyte ratio; AMI: Acute mesenteric ischemia; IL: Interleukin.

Diagnose values of PBMSCs exosomal circ-Eya3 and circEZH2_005

Much evidence has improved the diagnostic values of IFABP and D-lactate in AMI[29,30]. Considering the differential expression levels in AMI patients and correlations with IFABP and D-lactate, it is important to identify the diagnostic values of PBMSCs exosomal circ-Eya3 and circEZH2_005 in AMI. For circ-Eya3, the area under the ROC curve (AUC) was 0.677 (Figure 5A). For circEZH2_005, the AUC was 0.741 (Figure 5A). In addition, the AUC of IFABP ROC was 0.645, and that of D-lactate ROC was 0.787 (Figure 5A). We then predicted the combined diagnostic value of the two or multiple biomarkers. As shown in Figure 5B, the AUC of the combined curve for circ-Eya3 and circEZH2_005 was 0.798. For circ-Eya3 and D-lactate, the AUC of the combined ROC curve was 0.854 (Figure 5B). For circEZH2_005 and D-lactate, the AUC of the combined ROC curve was 0.819 (Figure 5B). The AUC of the combined ROC curve for circ-Eya3, circEZH2_005, and D-lactate was 0.891 (Figure 5B). Therefore, these two novel biomarkers, circ-Eya3 and circEZH2_005, combined with traditional D-lactate, showed excellent diagnostic value for AMI.

Figure 5 Diagnose values of peripheral blood-derived mesenchymal stem cells exosomal circ-Eya3 and circEZH2_005 to acute mesenteric ischemia patients. A: Separate receiver operating characteristic of circ-Eya3, circEZH2_005, intestinal fatty acid binding protein, and D-lactate; B: Combined receiver operating characteristic of circ-Eya3, circEZH2_005 and D-lactate. AUC: Area under the curve.

The differential levels of circ-Eya3, circEZH2_005, IFABP and D-lactate in AMI subtypes

We included 80 AMI patients who were diagnosed with arterial (n = 48) or venous (n = 32) AMI. As exhibited, circ-Eya3, IFABP and D-lactate were higher, while circEZH2_005 was lower in arterial AMI than in controls (Figure 6, P < 0.001). In venous AMI patients, circ-Eya3 and D-lactate levels were significantly higher than in controls (Figure 6B and C, P < 0.001). The levels of IFABP and circEZH2_005 were not significantly different between controls and venous AMI patients (Figure 6A and D, P > 0.05). The levels of IFABP and D-lactate were decreased and circEZH2_005 was increased in venous AMI patients compared with arterial AMI patients (Figure 6A, B, and D, P < 0.05).

Figure 6 The differential levels of circ-Eya3, circEZH2_005, intestinal fatty acid binding protein and D-lactate in arterial vs venous acute mesenteric ischemia subtypes. A and B: Enzyme-linked immunosorbent assay was applied to determine the intestinal fatty acid binding protein (A) and D-lactate (B) levels in plasma from different groups; C and D: Quantitative real-time polymerase chain reaction was utilized to measure the expression of circ-Eya3 (C) and circEZH2_005 (D) in peripheral blood-derived mesenchymal stem cells exosomes of different groups. ^aP < 0.05, ^cP < 0.001. NS: No significant; IFABP: Intestinal fatty acid binding protein; AMI: Acute mesenteric ischemia.

Patients were divided into early AMI (n = 36) and late necrotic AMI (n = 44) subgroups. As shown in Figure 7, IFABP and D-lactate were increased in the early AMI subgroup and late necrotic AMI subgroup (Figure 7A and B, P < 0.01) when compared with the control group. Of note, circ-Eya3 expression was increased, while circEZH2_005 expression was decreased in the late necrotic AMI subgroup compared with the control group (Figure 7C and D, P < 0.001). In addition, circ-Eya3 was highly expressed, while circEZH2_005 was lowly expressed in the late necrotic AMI subgroup compared to the early AMI subgroup (Figure 7C and D, P < 0.001). These results suggest the potential value of circ-Eya3 and circEZH2_005 in the identification of AMI subgroups.

Figure 7 The differential levels of circ-Eya3, circEZH2_005, intestinal fatty acid binding protein and D-lactate in early vs late necrotic acute mesenteric ischemia subtypes. A and B: Enzyme-linked immunosorbent assay was applied to determine the intestinal fatty acid binding protein (A) and D-lactate (B) levels in plasma from different groups; C and D: Quantitative real-time polymerase chain reaction was utilized to measure the expression of circ-Eya3 (C) and circEZH2_005 (D) in peripheral blood-derived mesenchymal stem cells exosomes of different groups. ^bP < 0.01, ^cP < 0.001. IFABP: Intestinal fatty acid binding protein; AMI: Acute mesenteric ischemia.

The differential levels of circ-Eya3, circEZH2_005, IFABP and D-lactate before and after operation in AMI patients

In 80 patients included in this study, 15 patients were carried out arterial embolectomy or mesenteric artery revascularization. We monitored the levels of PBMSCs exosomal circ-Eya3 and circEZH2_005, and plasma IFABP and D-lactate before and after the operation. It was observed that the levels of IFABP, D-lactate, and circ-Eya3 were increased after surgery compared with before operation, while circEZH2_005 expression was decreased after operation compared with before operation (Figure 8, P < 0.05). However, there were no significant differences in circ-Eya3, circEZH2_005, IFABP, and D-lactate at 6 hours, 12 hours, and 24 hours after the operation (Figure 8).

Figure 8 The dynamic levels of circ-Eya3, circEZH2_005, intestinal fatty acid binding protein and D-lactate before and after operation in acute mesenteric ischemia patients. A and B: Plasma intestinal fatty acid binding protein (A) and D-lactate (B) levels in acute mesenteric ischemia patients before and after operation were detected by enzyme-linked immunosorbent assay; C and D: Circ-Eya3 (C) and circEZH2_005 (D) expression levels in acute mesenteric ischemia patients before and after operation were analyzed through quantitative real-time polymerase chain reaction. ^aP < 0.05. IFABP: Intestinal fatty acid binding protein.

DISCUSSION

Early diagnosis and targeted therapeutic intervention are critical to reduce mortality in patients with AMI. Consequently, the ongoing discovery of novel early diagnostic biomarkers is a primary focus of AMI research. In this study, we identified significantly elevated levels of circ-Eya3 and reduced levels of circEZH2_005 in PBMSCs exosomes derived from patients with AMI. Correlation analyses revealed that exosomal circ-Eya3 was positively correlated with plasma levels of IFABP and D-lactate. Conversely, exosomal circEZH2_005 was significantly negatively correlated with both IFABP and D-lactate. Furthermore, circ-Eya3 was positively associated with the inflammatory markers NLR and IL-6, whereas circEZH2_005 was inversely related to these inflammatory indices. ROC curve analysis indicated that the combined utilization of circ-Eya3, circEZH2_005, and D-lactate has substantial diagnostic value for AMI. Notably, distinct expression patterns of PBMSCs exosomal circ-Eya3 and circEZH2_005 were observed between patients with early stage AMI and those with later necrotic presentations. Additionally, circEZH2_005 expression differed significantly between arterial occlusion and venous thromboembolism subtypes, highlighting its potential utility in subtyping AMI. Post-reperfusion therapy [e.g., percutaneous coronary intervention (PCI)], plasma levels of IFABP, D-lactate, and exosomal circ-Eya3 significantly increased, while exosomal circEZH2_005 levels markedly decreased.

Exosomes can be detected in various biological fluids, including urine, blood, saliva, and breast milk, making them potential biomarkers for disease diagnosis and monitoring, especially plasma exosomes[31]. Our study identified two PBMSCs exosomal circRNAs-circ-Eya3 (upregulated) and circEZH2_005 (downregulated)-as potential diagnostic biomarkers for AMI. These findings align with prior reports linking circEZH2_005 to intestinal ischemia/reperfusion injury, where its upregulation protected intestinal crypt cells via GPRC5A stabilization[24]. Similarly, another study implied that circ-Eya3 plays an oncogenic role in PDAC by the miR-1294/c-Myc axis to affect the formation of ATP[32]. In the condition of intestinal ischemia, intestinal epithelial cells are in an acute hypoxia and glucose deficiency environment, which may affect the production of ATP, which also explains the increased expression of circ-Eya3 in the PBMSCs exosomes of AMI patients. Meanwhile, previous researchers have shown that various types of MSCs exosomes can carry genetic material, including non-coding RNAs, to participate in the regulation of intestinal injury. Bone marrow MSCs (BMMSCs) exosome has been reported to carry miR-143-3p to alleviate intestinal ischemia-reperfusion injury by regulating pyroptosis[33]. Similarly, exosomes derived from BMMSCs have been shown to carry genetic material into intestinal epithelial cells under hypoxia. Exosomes entering intestinal epithelial cells can alleviate ulcerative colitis injury by regulating oxidative stress of the cells[34]. Therefore, the present study identified two circRNAs that may also be involved in the regulation of AMI. More studies are needed to clarify the specific mechanisms of these two circRNAs in AMI regulation.

Critically, circ-Eya3 exhibited a positive correlation with established intestinal injury markers (IFABP and D-lactate) and inflammatory indices (NLR and IL-6), while circEZH2_005 showed inverse correlations with these parameters. D-lactate indicates intestinal barrier malfunction and microbial translocation[35,36]; and IFABP indicates enterocyte damage[36,37]. A systematic review and meta-analysis revealed the moderate predictive value of IFABP and D-lactate for AMI[38]. Our study assessed the correlations of circ-Eya3 and circEZH2_005 with IFABP and D-lactate. These findings further confirm that circ-Eya3 and circEZH2_005 are related to intestinal injury in AMI. IL-6 promotes neutrophil infiltration and endothelial dysfunction, exacerbating microvascular thrombosis in AMI. Sutherland et al[39] demonstrated that the plasma IL-6 level in patients diagnosed with AMI was significantly elevated compared with other diagnoses. Elevated NLR reflects systemic inflammation and is a potential predictor of AMI[40]. These relationships position circ-Eya3 and circEZH2_005 not only as intestinal injury markers, but also as potential indicators of the inflammatory burden in AMI, offering a more comprehensive risk stratification tool. On the other hand, although no researchers have yet demonstrated that the exosomal circRNAs of PBMSCs have a regulatory effect on the inflammatory response of AMI. However, in other diseases, it has been proven that MSCs exosomes can carry circRNAs into inflammatory cells such as macrophages, thereby regulating the inflammatory responses of various diseases[41,42]. Other researchers have shown that BMMSCs exosome could carry miR-200b into intestinal epithelial cells. MiR-200b could target Hmgb3 to alleviate inflammatory injury of intestinal epithelial cells[43]. Therefore, we believe that circ-Eya3 and circEZH2_005 may also be involved in the inflammatory response of AMI. More in vivo and in vitro studies need to be conducted to elaborate on this hypothesis.

The observed association of circ-Eya3 and circEZH2_005 with potential diagnostic markers underscores their potential clinical diagnostic value. Here, we found that PBMSCs exosomal circ-Eya3 exhibited a certain predictive value for AMI and circEZH2_005 showed a moderate predictive value for AMI. More importantly, the combination of circ-Eya3, circEZH2_005, and D-lactate showed great diagnostic value for AMI. This multi-marker approach addresses the limitations of single biomarkers (e.g., IFABP heterogeneity in advanced necrosis). In recent years, advances in microfluidics (e.g., exosome-on-chip) and point-of-care PCR devices have enabled the rapid profiling of circRNAs. The inherent stability of circRNAs, which are resistant to RNase and have a longer half-life than linear RNAs, makes them promising targets for liquid biopsy. Collectively, our study provides AMI diagnostic markers with the advantages of clinical stability and accuracy.

Two important hospital-based factors that influence postoperative mortality and short bowel syndrome in AMI patients are delayed vascular consultation and vascular surgery[1]. However, the identification of early and advanced AMI still depends on imaging and pathology, which is the cause of delayed diagnosis and treatment. In addition, arterial occlusion AMI (e.g., embolism/thrombosis) requires urgent revascularization (e.g., PCI or thrombectomy), whereas AMI (e.g., mesenteric vein thrombosis) may prioritize anticoagulation[44]. CircEZH2_005 expression differed significantly between arterial occlusion and venous thromboembolism AMI subtypes. The ability of circEZH2_005 to subtype AMI could guide targeted interventions, addressing a critical diagnostic gap in which conventional biomarkers are specific. Additionally, the divergent expression patterns of both circRNAs between early reversible ischemia and advanced necrotic AMI highlight their utility in identifying surgical urgency, potentially reducing delays in vascular consultation.

Post-reperfusion (e.g., PCI), the sharp increase in circ-Eya3/IFABP/D-lactate and decline in circEZH2_005 suggest that these markers reflect real-time intestinal reperfusion injury, aiding in post-interventional monitoring. While dynamic changes post-surgery (6-24 hours) did not predict prognosis, their peak alterations immediately post-PCI may indicate acute intestinal damage severity.

CONCLUSION

In summary, our study demonstrated that PBMSCs exosomal circ-Eya3 was significantly upregulated, while circEZH2_005 was downregulated in AMI patients, establishing these circRNAs as novel biomarkers for AMI-associated intestinal injury. Critically, circ-Eya3 exhibited positive correlations with both intestinal damage markers (IFABP, D-lactate) and systemic inflammatory indices (NLR, IL-6), whereas circEZH2_005 showed inverse relationships with these parameters, underscoring their dual role in reflecting intestinal barrier disruption and inflammatory burden. The combined use of circ-Eya3, circEZH2_005, and D-lactate achieved superior diagnostic accuracy for AMI, providing a clinically actionable multi-marker strategy. Importantly, circEZH2_005 expression distinguishes arterial occlusion from venous thromboembolism AMI subtypes, a finding with direct therapeutic implications for guiding intervention selection (e.g., anticoagulation vs revascularization). Furthermore, the differential expression patterns of both circRNAs between early reversible ischemia and late necrotic AMI highlight their utility in triaging surgical urgency, potentially reducing critical delays in vascular consultation. These findings collectively suggest that PBMSCs exosomal circ-Eya3 and circEZH2_005 are promising biomarkers for the early diagnosis, subtyping, staging, and post-reperfusion monitoring of AMI. Nevertheless, several limitations of this study warrant consideration. While robust clinical correlations have been established, validation in larger multicenter cohorts is essential to define standardized reference ranges and address pre-analytical variability in exosome isolation. The molecular mechanisms by which circ-Eya3 and circEZH2_005 participate in regulating the inflammatory response and disease progression of AMI require further investigation using in vitro and in vivo models. Additionally, technical challenges in PBMSCs exosomal circRNA detection workflows must be overcome to enable clinical translation. Therefore, we will validate the findings of this study via more clinically feasible isolation methods (e.g., polymer-based precipitation, immunoaffinity capture) and detection platforms, such as digital PCR or streamlined microfluidic devices in the future.