Ferroptosis in acinar cells of traumatic pancreatitis: implications for predictive, preventive, and personalized approaches in intra-abdominal infection management

Abstract

Background

Traumatic pancreatitis (TP) is a distinct subtype of pancreatitis. Although ferroptosis of pancreatic acinar cells is well documented in acute pancreatitis (AP), studies on ferroptosis in TP and its relationship with subsequent intra-abdominal infections (IAIs) remain limited and unclear. Incorporating predictive, preventive, and personalized medicine (3PM) strategies could significantly enhance TP management, particularly through early detection and targeted interventions.

Methods

A total of 60 male rats were divided into four groups: TP model, ferroptosis activation, and inhibition groups. Physiological parameters, mortality rates, and serum cytokine, pancreatic enzyme expression, and oxidative stress factor levels were observed. Pathological evaluation of pancreatic tissue was performed. Subsequently, iron staining, ACSL4 immunofluorescence detection, and mRNA and protein expression of ferroptosis-related molecules were assessed in pancreatic tissues. Furthermore, 16S rDNA sequencing of peritoneal lavage fluid was performed to evaluate the impact of TP and ferroptosis modulation on the intra-abdominal microbiota and infection. Finally, clinical data from TP patients with IAIs were analyzed to identify commonalities with animal findings, offering predictive insights for human treatment.

Results

TP caused severe pancreatic tissue damage, and activation of ferroptosis further exacerbated tissue damage, systemic inflammation, and animal mortality. Inhibition of ferroptosis improved these indicators in TP rats. Furthermore, 16S rDNA sequencing results showed that TP rats had enhanced intra-abdominal microbiota dysbiosis and an increased proportion of pathogenic bacteria. The use of ferroptosis activators further aggravated the IAIs in TP rats. Clinical analysis showed elevated serum ferroptosis biomarkers in TP patients, with higher proportions of antibiotic-resistant Acinetobacter and Pseudomonas aeruginosa found in abdominal cultures, highlighting the need for predictive biomarkers and personalized therapeutic strategies.

Conclusions

Inhibiting ferroptosis alleviates pancreatic damage and reduces mortality in TP rats. Ferroptosis exacerbates IAIs by increasing oxidative stress and inflammatory responses. From a 3PM perspective, targeting ferroptosis offers predictive and preventive potential, enabling earlier interventions and personalized therapies to improve patient outcomes and reduce complications.

Supplementary Information

The online version contains supplementary material available at 10.1007/s13167-025-00418-3.

Background

Traumatic pancreatitis (TP) is a rare but clinically significant form of pancreatitis resulting from external mechanical trauma to the pancreas, which often leads to severe complications [1]. Despite being a subtype of acute pancreatitis (AP), TP exhibits distinct pathophysiological and clinical outcomes due to the nature of the injury [2, 3]. The extensive tissue damage caused by TP increases the risk of secondary complications, particularly intra-abdominal infections (IAIs), which contribute significantly to high mortality rates [4, 5]. This manuscript emphasizes how predictive, preventive, and personalized medicine (3PM) strategies can be integrated into TP management. By focusing on early identification, intervention, and individualized therapeutic approaches, we aim to address the critical challenges posed by TP and its complications, with particular attention to the role of ferroptosis in driving these outcomes.

Prevalence of IAIs following TP

IAIs are a frequent and serious complication following traumatic pancreatitis [6, 7]. The pancreas is located deep within the abdominal cavity, making it particularly vulnerable to damage from external trauma, such as blunt or penetrating abdominal injuries [8]. Though TP itself is relatively rare, accounting for approximately 2% of all closed abdominal injuries, it presents unique challenges due to the potential for severe pancreatic tissue injury and subsequent complications. In TP, significant damage to pancreatic acinar cells often leads to the release of digestive enzymes into the surrounding tissues, initiating autodigestion [9, 10]. This process not only contributes to local inflammation but also facilitates bacterial translocation and infection. These infections are a critical determinant in the prognosis of TP, making early detection and individualized prevention strategies essential for improving patient outcomes [11]. The integration of predictive biomarkers for IAIs within a 3PM framework offers the potential for earlier intervention and more effective personalized treatments for TP patients [12–14].

The role of ferroptosis in oxidative stress imbalance and iron deposition as a key contributor to IAIs and high mortality in TP

Ferroptosis, a novel form of regulated cell death distinct from necrosis, apoptosis, and autophagy, has emerged as a critical player in the pathophysiology of acute pancreatitis, including TP [15, 16]. Ferroptosis is driven by an imbalance in cellular iron homeostasis and oxidative stress. Iron accumulation in pancreatic cells leads to lipid peroxidation, which triggers cellular damage and inflammation [17–19]. This oxidative stress exacerbates tissue injury and promotes the activation of inflammatory cascades. In TP, ferroptosis is thought to contribute to the early destruction of acinar cells, leading to the release of pro-inflammatory cytokines, further amplifying the systemic inflammatory response. This cascade not only increases the risk of local infection but also facilitates bacterial overgrowth and translocation within the abdominal cavity, thereby heightening the risk of IAIs. The elevated levels of reactive oxygen species (ROS) and iron deposition in pancreatic tissues significantly contribute to the high mortality rates observed in TP, underscoring the importance of targeting ferroptosis pathways to mitigate these outcomes [20]. Studies in mouse models have demonstrated that regulating the expression of ferroptosis-related proteins offers protective effects in severe acute pancreatitis (SAP) [21, 22]. Furthermore, in an AP model constructed using caerulein-induced AR42J cells, knockdown of GPX4 effectively suppressed oxidative stress and ferroptosis, presenting a potential therapeutic strategy for AP [23]. Predictive models based on ferroptosis biomarkers could identify at-risk patients early, enabling targeted preventive interventions. Preventive strategies could modulate ferroptosis through antioxidant pathways or by inhibiting iron accumulation, thus reducing oxidative damage and improving patient prognosis. Personalized treatments targeting ferroptosis pathways could further enhance therapeutic efficacy while minimizing side effects, aligning with 3PM principles by tailoring interventions to the individual patient’s cellular profile [24–26].

Hypothesis within the framework of predictive, preventive, and personalized medicine (PPPM)

Integrating predictive, preventive, and personalized medicine (3PM) into the management of TP offers a novel approach to improving patient outcomes. Our hypothesis posits that early identification of ferroptosis-related biomarkers can predict the likelihood of IAIs and high mortality in TP patients. These biomarkers would serve as predictive tools, enabling early risk stratification and personalized intervention strategies. Preventive approaches could target ferroptosis mechanisms to mitigate the risk of secondary infections, while personalized therapies would be tailored to the individual’s ferroptotic profile to optimize treatment outcomes. By aligning our approach with the 3PM framework, we aim to provide more precise, targeted, and effective strategies for managing TP and preventing its most lethal complications.

Study design

To test this hypothesis within the 3PM framework, we used a traumatic pancreatitis rat model, with sixty male rats divided into four groups: TP model, ferroptosis activation, ferroptosis inhibition, and control. We assessed predictive markers, including mortality rates, inflammation markers, serum cytokines, and oxidative stress indicators. Pancreatic damage was evaluated histologically, focusing on iron deposition and oxidative stress markers using techniques like iron staining and ACSL4 immunofluorescence. Additionally, we measured ferroptosis-related proteins at both the mRNA and protein levels to further explore ferroptosis in TP.

The effects of ferroptosis modulation on the intra-abdominal microbiota and infection were evaluated through 16S rDNA sequencing of peritoneal lavage fluid, allowing us to examine the relationship between ferroptosis, bacterial dysbiosis, and IAIs in TP. To bridge the gap between animal models and clinical practice, we collected clinical data from TP patients with IAIs, including baseline lab results and abdominal drainage fluid cultures. This approach links experimental data with clinical outcomes, providing predictive insights for individualized treatment strategies.

By integrating experimental findings with clinical data, we aim to develop a personalized, predictive, and preventive therapeutic strategy for TP. This approach aligns with 3PM principles, enhancing patient management by targeting ferroptosis and IAIs to improve outcomes.

Materials and methods

Experimental animal grouping and TP modeling strategy

TP, as a distinct form of AP, exhibits two “peaks of mortality,” similar to SAP, in clinical presentations. The first peak occurs during the acute phase, characterized by trauma and inflammatory responses, while the second peak of mortality is associated with subsequent IAIs. Our previous study reported the survival outcomes of TP rats, with the first mortality peak typically occurring within 1 day following injury, and the second peak emerging 1 week later [27]. Therefore, in this study, animal sampling was conducted at two time points: 1 day and 7 days post-injury. The TP rat model was established using our patented multifunctional animal impact device (Patent No: ZL 2016 1 0347341.5). Anesthesia for the animals was administered via an anesthetic machine with isoflurane (purchased from Shenzhen Ruiward Life Technology Co., Ltd). SPF-grade Sprague–Dawley (SD) rats used in the experiments were purchased from Shanghai Bikai Koyi Biotechnology Co., Ltd.

We randomly assigned 60 healthy adult male rats, each weighing approximately 250 g, into four groups: a normal control group, a pancreatic injury group (TP group), a ferroptosis agonist group (TPE group), and a ferroptosis inhibitor group (TPF group). The normal control group underwent no special treatment. For the TP group, rats were anesthetized, and the pancreas was exposed via midline laparotomy. The pancreas was then impacted using a percussion device at a pressure of 400 kPa, followed by layered closure of the abdominal wall (Fig. 1a). In the TPE and TPF groups, Erastin (HY-15763, MedChemExpress, Monmouth Junction, NJ, USA) (10 mg/kg) and Ferrostatin-1 (HY-100579, MedChemExpress, Monmouth Junction, NJ, USA) (5 mg/kg) were administered intraperitoneally 1 h prior to modeling, respectively, after which they underwent the same pancreatic impact procedure as the TP group. At the conclusion of the experiment, all rats received a subcutaneous injection of 1 mL physiological saline following anesthesia. To investigate the two peaks of mortality associated with pancreatic injury, 4 rats from each group were euthanized for sampling at 1 day post-injury, while the remaining rats were euthanized at 7 days post-injury. Rectal temperature was continuously monitored throughout the experiment to track body temperature fluctuations. A detailed overview of the animal intervention strategy is shown in Fig. 1 b. This study was approved by the ethics committee, and all procedures were designed to minimize the suffering of the experimental animals. All animal experiments were approved by the Animal Ethics Committee of the Jinling Hospital (Approval No.: 2023JLHGZRDWLS-00087). All the authors complied with the ARRIVE guidelines experiments.

Fig. 1.

Experimental design and physiological parameters of traumatic pancreatitis modeling. a Schematic diagram of the traumatic pancreatitis model. b Key time points for animal modeling and sample collection. c Survival curves of experimental animals in each group. d Changes in body temperature of experimental animals in each group

Histopathological evaluation of the pancreas and serum cytokine assay

Pancreatic tissue from rats euthanized 1 day post-modeling was fixed in 4% paraformaldehyde for 48 h, followed by gradual dehydration using increasing concentrations of ethanol. The tissue was then cleared in xylene, embedded in paraffin, and sectioned. After overnight drying at 37 °C, sections were stained with hematoxylin and eosin, followed by thorough washing with distilled water and clearing in xylene before being mounted. The pancreatic tissue was examined under a microscope, and the extent of edema, hemorrhage, cell necrosis, and inflammatory cell infiltration was scored by a pathologist, according to the criteria outlined by Schmidt et al. [28].

For serum analysis, blood samples from rats euthanized 1 day post-modeling were centrifuged at 3000 rpm for 5 min, and the serum was stored at − 80 °C. The concentrations of amylase and lipase in the rat serum were measured using enzyme-linked immunosorbent assay (ELISA) kits purchased from Nanjing Jiancheng Biological Engineering Institute (Jiangsu, China). Additionally, inflammatory cytokines, including tumor necrosis factor [TNF]-α, transforming growth factor [TGF]-β, interleukin [IL]−6, and IL-10, were also assessed by ELISA. Finally, oxidative stress markers such as superoxide dismutase (SOD) and glutathione (GSH) in rat serum were measured using dry chemical methods.

Iron staining of pancreatic tissue

Pancreatic tissue previously fixed in paraformaldehyde was dewaxed to water, followed by staining with Prussian blue stain for 20 min. After washing, the tissue was stained with eosin and then dehydrated stepwise before being mounted. The pancreatic tissue was examined under a microscope, and the area of positive staining for iron was quantified using ImageJ to compare the iron content in the pancreatic tissue across different groups.

Immunofluorescence detection of ACSL4 in pancreatic tissue

Paraffin-embedded pancreatic tissue sections were subjected to antigen retrieval using citrate buffer and subsequently blocked with BSA blocking solution. Diluted primary antibody against ACSL4 (1:200) was applied and incubated overnight at 4 °C. The following day, the sections were washed three times with PBS (G4202-500 mL, Servicebio, Wuhan, China), and a fluorescent secondary antibody (dilution 1:500) was applied in the dark and incubated for 1 h at room temperature. Subsequently, the nuclei were stained with DAPI, and the sections were mounted with an anti-fade reagent. Finally, the pancreatic tissue was examined under a fluorescence microscope, and the number of positive cells was quantified using Aipathwell digital pathology image analysis software (Servicebio).

Detection of key genes and proteins involved in ferroptosis in pancreatic acinar cells

qPCR was employed to measure the mRNA expression levels of Actb, Gpx4, Slc7a11, Tfrc, Hmox1, and Nfe2l2 in the pancreatic tissue of rats from each group. The specific procedure is as follows: total RNA was extracted using a handheld tissue grinder and an RNA extraction kit on ice. A DNA removal reaction system was then employed to eliminate genomic DNA, followed by cDNA synthesis through reverse transcription. The primers for the aforementioned genes are synthesized by Shenggong Biotech (Shanghai), with the specific sequences shown in Table 1. Relative mRNA expression levels were calculated based on the CT values.

Table 1.

RT‐qPCR primer sequences

Genes	Primer sequences
Actb	F: CTGTGTGGATTGGTGGCTCT
	R: AGCTCAGTAACAGTCCGCC
Gpx4	F: CCATTCCCGAGCCTTTCAACC
	R: TCGGTTTTGCCTCATTGCGA
Slc7a11	F: ACTCGGATCCGTTTAGCACC
	R: ATTCCCAGATGAGGCCGTTAG
Tfrc	F: CCGGCCTATATGCTTGGGTAG
	R: GACAATGGCTCCCCTCCAAA
Hmox1	F: CCCCTCCTCTTCTGTCAGATTC
	R: TGCTGGGGTTAAAGTCCCG
Nfe2l2	F: CCAGACGAGGCGGTACAAG
	R: CGGAAGGTTACAACGTGGGG

In this study, the protein expression levels of β-actin (GB12001, Servicebio, Wuhan, China), GPX4 (AB125066, Abcam, Cambridge, UK), ACSL4 (DF12141, Affinity, Changzhou, China), TFR1 (AB269513, Abcam, Cambridge, UK), HO-1 (AB68477, Abcam, Cambridge, UK), and NRF2 (20,733, Abcam, Cambridge, UK) in pancreatic tissue were assessed by Western blot (WB). Briefly, pancreatic tissue was homogenized on ice using an animal protein extraction column, followed by the addition of tissue lysis buffer for further homogenization. After centrifugation at 15,000 rpm for 2 min, the supernatant containing the protein was collected. Protein concentration was measured using the BCA assay, and protein concentrations were normalized to ensure consistency across groups. Subsequently, the proteins were denatured and resolved on 10% SDS-PAGE gels, followed by electrophoresis and membrane transfer. The membranes were incubated overnight with the corresponding primary antibodies (diluted 1:2000), followed by washing and incubation with a 1:5000 dilution of secondary antibody. After additional washing, chemiluminescence was detected, and relative protein expression levels were determined by comparing the gray values of each group with the internal control.

Collection of peritoneal lavage fluid and 16S rDNA sequencing

Seven days post-TP, 5 mL of PBS was injected into the left lower abdomen of the rats. The abdomen was then gently massaged to ensure even distribution of the fluid throughout the peritoneal cavity. The abdominal cavity was subsequently opened along the midline, and peritoneal lavage fluid was aspirated from the left posterior abdominal wall using a sterile syringe. The collected lavage fluid was transferred into a centrifuge tube for subsequent 16S rDNA sequencing. The 16S rDNA sequencing of the peritoneal lavage fluid was performed on the Illumina MiSeq platform at Shanghai Meiji Biotechnology Co., Ltd. Following DNA extraction from the samples, amplification and library preparation were completed, followed by sequencing.

Selection and baseline data of patients with pancreatic trauma complicated by abdominal infection

Patients diagnosed with TP complicated by abdominal infection, who were hospitalized in the Department of General Surgery at Jinling Hospital, affiliated with Nanjing University School of Medicine, between January 2023 and December 2024, were included in this study.

Inclusion criteria: (1) Patients diagnosed with TP based on clinical symptoms, CT, MRI, and laboratory test results, who were admitted to the general surgery department; (2) patients aged over 18 years with no prior history of abdominal trauma; (3) confirmation of bacterial infection through abdominal drainage fluid or blood cultures.

Exclusion criteria: (1) Patients who did not undergo comprehensive laboratory testing (e.g., complete blood count, liver and kidney function, and electrolytes); (2) patients with multiple traumatic injuries; (3) pregnant women or those for whom the collection of clinical data was deemed inappropriate.

In total, four patients with pancreatic trauma complicated by abdominal infection were included in this study. The collection of clinical data primarily encompassed demographic information such as age, sex, height, and weight, as well as comorbidities including hypertension, diabetes, and other chronic conditions, surgical history, trauma history, and the cause of injury upon admission. This study was approved by the Ethics Committee of the Jinling Hospital (Approval number: 2024DZKY-087–02). All human participants were adults and had provided written informed consent.

Laboratory tests and analysis of ferroptosis biomarkers

The laboratory tests primarily included the measurement of serum amylase and lipase levels, complete blood count (hemoglobin, neutrophils, and platelets), liver function tests (albumin, alanine aminotransferase [ALT], aspartate aminotransferase [AST]), renal function tests (serum creatinine and blood urea nitrogen), biochemical markers (serum sodium and serum potassium), as well as C-reactive protein (CRP) and procalcitonin (PCT).

Direct analysis of ferroptosis characteristics in pancreatic tissue from patients would be inappropriate. Therefore, blood samples were collected from four patients diagnosed with TP, and serum was obtained via centrifugation. Additionally, serum samples from four healthy adult controls were included for comparison. Ferroptosis biomarkers in the serum, specifically the expression levels of TFR1, ACSL4, and GPX4, were assessed indirectly using ELISA to evaluate the ferroptosis characteristics of the patients’ pancreatic tissue.

Bacteriological culture and antibiotic resistance characterization of abdominal drainage fluid

Patients diagnosed with TP complicated by abdominal infection were required to undergo abdominal drainage. Using a sterile syringe, 4–5 mL of abdominal drainage fluid was aspirated from the drainage tube and immediately sent to the laboratory for bacteriological culture. The samples were inoculated onto blood agar plates for incubation, isolation, and purification. Antibiotic susceptibility testing was performed using the Kirby-Bauer disc diffusion method (K-B method), while bacterial identification was carried out using the MicroScan WalkAway 40S automated system.

Statistical analysis

In this study, the analysis of 16S rDNA sequencing data was performed using R software (version 3.3.1), with various packages including boot (1.3.18), stats (3.3.1), vegan (2.4.3), and pheatmap (1.0.8). LEfSe analysis was conducted using the LEfSe software (http://huttenhower.sph.harvard.edu/galaxy/root?tool_id=lefse_upload). Phylogenetic analysis was performed using FastTree (version 2.1.3). Functional prediction using PICRUSt2 was carried out via PICRUSt2 (http://huttenhower.sph.harvard.edu/galaxy). BugBase functional prediction was completed using BugBase (https://bugbase.cs.umn.edu/index.html).

Statistical analysis of experimental data was performed using SPSS 25.0 (IBM Corp., Armonk, NY, USA). Between-group comparisons were conducted using one-way analysis of variance (ANOVA), with a significance level of P < 0.05. Data visualization was completed using GraphPad Prism 9.5.0 (GraphPad Software, Boston, Massachusetts, USA).

Results

Survival rates and temperature changes in rats of each group

Prior to the intervention, each group—Normal, TP, TPE, and TPF—comprised 15 SD rats. In each group, 4 rats were euthanized at the modeling stage and 1 day post-modeling for tissue collection. The Normal group experienced no modeling or mortality during the study period, and the remaining 11 rats were euthanized on day 7 for sampling. In the TP group, 5 rats died within the first day post-modeling, and 1 additional rat died over the subsequent 7 days. In the TPE group, 6 rats died within the first day post-modeling, and 2 more rats died within the following 7 days, with 3 rats remaining for sampling on day 7. In the TPF group, 3 rats died within the first day post-modeling, and 1 additional rat died during the following 7 days, leaving 7 rats that survived for 7 days and were successfully sampled. The survival curves for each group are shown in Fig. 1 c.

Rectal temperature was measured three times daily for each rat, and the daily average temperature was calculated. The temperature changes for each group are displayed in Fig. 1 d. In the Normal group, body temperature remained stable at approximately 39 °C throughout the 7-day period with no significant variation. In contrast, the other three groups exhibited similar trends, with a rapid increase in body temperature within the first day post-modeling, reaching a peak, followed by a gradual decrease, although temperatures remained higher than those of the Normal group. Notably, the TPE group maintained a higher average body temperature, compared to the other three groups.

Pancreatic pathological injury and serum cytokine levels

Pancreatic pathological injury in each group is shown in Fig. 2 a. In the Normal group, only minimal vacuolar degeneration of acinar cells was observed. In the TP group, widening of the lobular interspace, infiltration of some granulocytes, and interstitial edema were evident. The TPE group exhibited extensive interstitial edema in the pancreatic tissue, accompanied by significant infiltration of granulocytes and lymphocytes, with widespread acinar cell necrosis at the edges of the lobules. In the TPF group, mild edema of the capsule was observed, with minimal inflammatory cell infiltration. The score of edema, hemorrhage, cell necrosis, and inflammatory cell infiltration in each group is quantified in Fig. 2 b. Overall, the TPE group showed the most severe damage, while the TPF group exhibited less tissue injury, compared to the TP group.

Fig. 2.

Histopathology and biochemical evaluation of pancreatic injury and systemic inflammation. a HE staining of pancreatic tissue (20 × and 200 × magnification). b Pancreatic injury scoring, including edema, hemorrhage, necrosis, and inflammatory cell infiltration scores. c Serum levels of amylase and lipase. d Serum levels of SOD and MPO. e Serum levels of inflammatory cytokines (TNF-α, TGF-β, IL-6, and IL-10). * P < 0.05

Further evaluation of the serum pancreatic enzyme levels in each group revealed markedly elevated amylase and lipase levels in the TP, TPE, and TPF groups, with the TPE group showing the most significant enzyme leakage (Fig. 2c). Pro-inflammatory factor TNF-α and oxidative stress marker MPO rapidly increased following pancreatic injury, indicating a strong inflammatory response and heightened oxidative stress (Fig. 2d). Additionally, serum inflammatory markers in the TPE and TPF groups displayed opposing trends, compared to the TP group, suggesting that ferroptosis activation contributes to the promotion of systemic inflammation and oxidative stress (Fig. 2e).

Pancreatic tissue iron staining and ACSL4 immunofluorescence expression

Prussian blue staining in each group indicated that normal pancreatic tissue exhibited low cellular iron content. After TP induction, pancreatic tissue showed a significant increase in iron deposition. In the TPE group, due to the application of the ferroptosis activator, iron deposition was further elevated. However, the application of ferrostatin-1 effectively blocked iron deposition (Fig. 3a). Immunofluorescence staining revealed that the expression level of ACSL4 was significantly increased in the TP group. Pre-treatment with Erastin via intraperitoneal injection further enhanced ACSL4 expression, whereas the TPF group exhibited a downregulation of ACSL4 expression (Fig. 3b).

Fig. 3.

Iron accumulation and ACSL4 expression. a Prussian blue staining of pancreatic tissue, where iron elements appear blue. b Immunofluorescence scanning of ACSL4 in pancreatic tissue, with positive cells shown in red. * P < 0.05

Expression of ferroptosis-related genes and proteins in pancreatic tissue

PCR was performed to assess the expression levels of Actb, Gpx4, Slc7a11, Tfrc, Hmox1, and Nfe2l2 (Fig. 4a). Compared to the Normal group, expression levels of Slc7a11, Tfrc, and Hmox1 were significantly elevated in the TP group, suggesting that trauma rapidly increases lipid peroxidation in pancreatic tissue. In contrast, expression levels of Gpx4 and Nfe2l2 decreased in the TP group, indicating that the pancreatic tissue becomes more susceptible to ferroptosis following injury. These ferroptosis-related genes exhibited opposite trends in the TPE and TPF groups, primarily due to the effects of the ferroptosis activator and inhibitor.

Fig. 4.

Molecular markers of ferroptosis in pancreatic tissue. a Relative mRNA expression levels of ferroptosis-related genes. b Expression levels of ferroptosis-related proteins. * P < 0.05

Western blot analysis revealed stable expression of the loading control (β-actin) across all groups, indicating consistent protein concentration (Fig. 4b). ACSL4 and GPX4 are positive and negative regulators of ferroptosis, respectively. In the TP group, ACSL4 protein expression was significantly elevated, while GPX4 expression was downregulated, suggesting increased sensitivity of acinar cells to ferroptosis. As a marker protein for ferroptosis, TFR1 was expressed at low levels in the Normal group but rapidly increased after TP, indicating intense ferroptosis in acinar cells following pancreatic injury. Finally, as a negative regulator of ferroptosis, NRF2 expression rapidly decreased after TP, and the use of ferroptosis inhibitors partially restored NRF2 expression, inhibiting the occurrence of ferroptosis.

16S rDNA sequencing of peritoneal lavage fluid reveals bacterial species composition and differences

Alpha diversity analysis was first performed on the sequencing results. The differences between groups were assessed, with the TPE group showing lower species richness, while the TP group exhibited greater variance (Fig. 5a). In beta diversity analysis, sample-level clustering analysis (Fig. 5b) indicated substantial differences in microbial composition between groups. Principal component analysis (PCA) and inter-group beta diversity comparison (Fig. 5c) further supported this observation, highlighting significant differences in the microbial community structure between the groups. Except for the higher sample dispersion within the TP group, other groups showed lower dispersion. The neutral community model (NCM) was used to evaluate the influence of environmental stochasticity on community structure. With an R² value of 0.29, it suggests that the community assembly is minimally affected by stochastic processes.

Fig. 5.

Microbial diversity and compositional analysis. a Inter-group difference test for Alpha diversity indices. b Hierarchical clustering analysis of the samples. c PCA analysis, NCM analysis, and inter-group difference test for beta diversity indices. d Venn diagram and community bar plot of species. e Bar chart comparison of species difference analysis

Next, microbial community metrics were evaluated across groups. The Gut Microbiome Health Index (GMHI), based on the species-level classification of the gut microbiome samples, was assessed to reflect the health status of each group. As shown in Fig. S1 a, the GMHI of the TP group was significantly lower than that of the Normal group. Pairwise comparisons of the four groups revealed that pancreatic injury and ferroptosis activation resulted in the lowest microbiome health index, while inhibition of ferroptosis partially restored this index. The Microbiome Disorder Index (MDI), used to assess the degree of microbial dysbiosis in each group, showed a trend opposite to GMHI, further supporting the notion that pancreatic acinar cell ferroptosis promotes dysbiosis in the IAIs microbiome (Fig. S1b).

Finally, Venn diagrams and community bar plots were used to observe trends in bacterial species variation across the groups. Both the TP and TPE groups showed increased complexity in the relative abundance of microbial communities, while the TPF and Normal groups had more similar community structures, compared to the other groups (Fig. 5d). At the genus level, species differences were evaluated across groups. The Normal group exhibited a more balanced proportion of different genera. In contrast, Corynebacterium, Staphylococcus, and Acinetobacter were more prevalent in the TP group, while Corynebacterium, Psychrobacter, and Facklamia predominated in the TPE group. The TPF group exhibited higher proportions of Corynebacterium, Jeotgalicoccus, and Achromobacter (Fig. 5e). Furthermore, LEfSe analysis was employed to assess multiple hierarchical differences from phylum to genus (Fig. S1c).

Clinical factor associations and functional prediction analysis

To further evaluate the factors influencing the microbial community, we incorporated clinical factors such as TNF-α, TGF-β, IL-6, IL-10, amylase, and lipase from the rats’ serum into the 16S rDNA sequencing analysis. After including these clinical factors, distance-based redundancy analysis (db-RDA) was performed to assess the relationships between clinical factors, sample grouping, and microbial communities (Fig. 6a). The results indicated that pro-inflammatory factors (TNF-α and IL-6) and serum pancreatic enzymes (amylase and lipase) were negatively correlated with anti-inflammatory factors (TGF-β and IL-10). The TP group and ferroptosis inhibitors had a significant impact on bacterial community distribution. Ordinal regression analysis was performed for the six clinical factors, revealing that higher levels of pro-inflammatory factors and serum pancreatic enzymes were positively correlated with bacterial species diversity in the samples, suggesting that increased inflammation and enzyme leakage enhance the bacterial diversity in IAIs (Fig. 6b). Conversely, elevated levels of anti-inflammatory factors helped reduce the species diversity of the IAIs microbiome.

Fig. 6.

Microbiota-environment interactions and functional predictions. a db-RDA analysis showing the relationship between samples, clinical factors, and microbiota. b Ordinal regression analysis evaluating the correlation between individual clinical factors and microbiota structure. c PICRUSt2 functional prediction (KEGG functional abundance statistics). d BugBase phenotype prediction of bacterial functions

Functional characteristics of the microbiomes across the groups were evaluated using PICRUSt2 functional prediction. Figure 6 c illustrates the functional differences at KEGG Level 2 between the groups. After inhibiting pancreatic acinar cell ferroptosis, functional abundances in pathways related to lipid metabolism, nucleotide metabolism, and replication were significantly reduced in the peritoneal lavage microbiota. BugBase phenotype prediction was used to assess key functional traits of the microbiomes (Fig. 6d). Both the TP and TPE groups showed increased proportions of the “Forms Biofilms” and “Aerobic” phenotypes. Ferroptosis agonist treatment led to an increase in the “Stress Tolerant” phenotype. Finally, a phylogenetic tree is constructed, and Fig. S1 d illustrates the evolutionary relationships of species within the samples from a molecular evolution perspective.

Clinical characteristics of patients and serum ferroptosis biomarker levels

A total of four patients with TP complicated by abdominal infection are included in the study, with detailed baseline data presented in Table 2. TP 2 patients underwent pancreatic necrosis debridement, abdominal hemostasis surgery, and abdominal drainage, while the remaining patients received only abdominal drainage. Figure 7 a shows a CT scan of a TP patient, where significant edema of the pancreatic tissue is clearly visible due to the trauma. Following the onset of TP, patients exhibited markedly elevated serum amylase and lipase levels (Fig. 7b). An increase in bilirubin levels was also observed after the occurrence of TP, although serum transaminase levels did not differ significantly from those in healthy individuals (Fig. 7c and d).

Table 2.

Clinical data of TP patients

Number	Age (year)	Height (m)	Weight (kg)	Hemoglobin (g/L)	WBC (109/L)	PLT (109/L)	Surgical intervention
TP 1	53	1.70	85	66	4.8	260	Abdominal drainage
TP 2	47	1.82	91	69	22.5	97	Pancreatic necrosis debridement, abdominal hemostasis surgery, and abdominal drainage
TP 3	30	1.79	77	79	11.3	190	Abdominal drainage
TP 4	62	1.68	73	75	9.8	255	Abdominal drainage

The complete blood count data were obtained from the patient’s initial laboratory test results upon admission

TP traumatic pancreatitis, WBC white blood cell count, PLT platelet

Fig. 7.

Clinical and biochemical characterization of TP patients. a Abdominal CT scans 1 week and 2 months after the onset of TP. Blue indicates significant pancreatic edema; red highlights necrotic pancreatic tissue; yellow marks peripancreatic fluid exudates. White represents the pancreas with morphological recovery 2 months post-injury. b Expression levels of serum amylase and lipase in normal individuals and TP patients. c Stacked bar chart of serum bilirubin levels in normal individuals and TP patients. d Expression levels of serum transaminases in normal individuals and TP patients. e Serum levels of CRP and procalcitonin (PCT) in normal individuals and TP patients. f ELISA measurements of serum levels of TFR1, ACSL4, and GPX4 in four TP patients over 21 days. * P < 0.05

Serum levels of CRP and procalcitonin PCT were measured, revealing a sharp increase in inflammatory biomarkers following the onset of TP, with statistically significant differences (Fig. 7e). Finally, within 21 days after the onset of TP, the ferroptosis marker TFR1 rapidly increased in the serum, peaking during the control of TP and gradually declining thereafter. The trend in the ferroptosis-promoting factor ACSL4 in the serum mirrored that of TFR1. In contrast, the ferroptosis-inhibiting factor GPX4 quickly decreased after TP occurred and gradually recovered following the resolution of inflammation (Fig. 7f).

Bacteriological culture and antibiotic sensitivity testing of abdominal drainage fluid

Given the differing strategies for antibiotic susceptibility testing of Gram-positive and Gram-negative bacteria, we separately analyzed the antibiotic resistance profiles of these two groups in the abdominal drainage fluid of the four patients. In the analysis of Gram-positive bacterial resistance, Enterococcus faecalis was detected in the abdominal drainage fluid cultures of all four patients (Fig. 8a). Notably, all Enterococcus faecalis isolates were sensitive to linezolid, and multiple other antibiotics, to which Enterococcus faecalis was susceptible, were available for therapeutic use in most patients. In the Gram-negative resistance analysis, Acinetobacter baumannii complex was identified in the cultures of three patients, which corresponded with the results of Acinetobacter identified in the abdominal lavage fluid sequencing from TP rats (Fig. 8b). Additionally, the antibiotic resistance profiles of Acinetobacter baumannii complex indicated resistance to nearly all commonly used clinical antibiotics. Finally, Pseudomonas aeruginosa was cultured from the abdominal drainage fluid of two patients, and these strains exhibited resistance to all tested antibiotics.

Fig. 8.

Antibiotic resistance profiles of bacteria isolated from TP abdominal drainage fluids. a Heatmap of antibiotic resistance profiling for gram positive bacteria in abdominal drainage fluid cultures of TP patients. b Heatmap of antibiotic resistance profiling for gram negative bacteria in abdominal drainage fluid cultures of TP patients

Discussion

AP is characterized by two peaks of mortality: one during the early phase of SIRS and organ failure, and the other during the late phase due to localized or systemic infections and complications [29, 30]. Among these complications, IAIs is a common occurrence in AP, primarily resulting from infected pancreatic necrosis and intestinal infections [31]. In TP, the ferroptotic features of pancreatic acinar cells and their subsequent role in IAIs remain underexplored. In this study, we systematically evaluated the ferroptotic characteristics of pancreatic acinar cells following TP and intervened in ferroptosis to assess its impact on tissue damage and mortality in rats. Furthermore, we also investigated the changes in the peritoneal IAIs in TP rats.

Molecular mechanisms of ferroptosis in traumatic pancreatitis and its potential as a predictive biomarker

As an acute inflammatory response, ferroptosis in various pancreatic cells following AP has been well-documented. Compared to AP, TP has two distinct characteristics that make it more prone to ferroptosis in pancreatic acinar cells. First, the risk of pancreatic duct rupture due to trauma is higher, leading to a significant leakage of pancreatic enzymes, which exacerbates local inflammation and cell death [32]. Moreover, trauma directly increases tissue oxidative stress levels, promoting lipid peroxidation and accelerating the onset of ferroptosis [33, 34]. A critical contributor to these processes is mitochondrial dysfunction. Following trauma, mitochondrial damage enhances oxidative stress, impairs energy production, and accelerates cellular injury, thereby amplifying the ferroptotic cascade. This mitochondrial dysfunction represents an important intersection between oxidative stress and ferroptosis, which plays a central role in tissue damage and inflammation in TP. In addition to mitochondrial dysfunction, we have incorporated the role of NETosis (Neutrophil extracellular traps, NETs) as a downstream effect of ferroptosis in our model. NETosis, the formation of extracellular neutrophil traps, contributes significantly to the inflammatory environment and infection [35]. In TP, NETosis exacerbates inflammation, promoting a vicious cycle of immune dysregulation and tissue damage. This process is further amplified by systemic immune reprogramming, wherein immune responses are dysregulated, contributing to secondary infections and the onset of sepsis [36].

Our findings also support these reports. ELISA assays of serum showed a significant reduction in antioxidant enzyme SOD levels following pancreatic trauma, and ferroptosis inhibitors were able to partially restore SOD expression. Further evaluation of ferroptosis characteristics in pancreatic acinar cells revealed significant iron deposition in the pancreas after trauma, which was reduced by ferroptosis inhibition. Finally, the mRNA and protein levels of various ferroptosis-related key molecules also confirmed the activation and rapid progression of ferroptosis in pancreatic acinar cells following trauma. Blocking ferroptosis improved the survival rate and prognosis of TP.

Our findings highlight the progress in trauma-related ferroptosis research, which aligns closely with the principles of 3PM. Ferroptosis biomarkers such as TFR1 and ACSL4 offer predictive value in identifying patients at high risk for complications like multi-organ failure and IAIs, enabling early intervention strategies. In the context of preventive medicine, inhibiting ferroptosis could mitigate tissue damage and prevent complications, shifting from reactive to proactive care [37–39]. Furthermore, the integration of personalized medicine allows for the development of tailored therapies based on an individual’s genetic profile and ferroptosis-related pathways, maximizing therapeutic benefits while minimizing adverse effects. Collectively, these approaches reflect a paradigm shift in trauma care, offering novel opportunities for 3PM strategies in clinical practice.

Bacterial community profile of IAIs following TP

Another focus of this study is the bacterial community characteristics of IAIs 7 days after TP. Similar to late-stage IAIs in AP, TP-induced IAIs also exhibits a surge in inflammatory factors caused by intensified pancreatic autodigestion. The exacerbation of systemic inflammation leads to the accumulation of large amounts of protein-rich exudate in the peritoneal cavity, which serves as a growth medium for bacteria [40]. Furthermore, severe trauma activates the hypothalamic–pituitary–adrenal axis, resulting in the release of large amounts of glucocorticoids that suppress the expression of MHC-II on monocytes and impair dendritic cell maturation, thereby compromising antigen presentation. This immunosuppressive state makes it easier for bacteria in the peritoneal cavity to breach host defenses, leading to severe abdominal infection [41, 42]. In this study, when simulating trauma by striking the pancreas with impact head, we sterilized the impact head in advance to prevent bacteria on the impact head from entering the peritoneal cavity. NCM analysis also indicated that environmental factors had a minimal influence on the IAIs in this study. Beta diversity analysis revealed significant differences in bacterial communities between the groups. Both GMHI and MDI scores suggested an increase in gut microbial dysbiosis in the peritoneal lavage fluid after pancreatic trauma.

At the genus level, significant differences in the bacterial communities between groups were observed. After TP, the proportion of Gram-positive bacilli (Corynebacterium), was notably higher. In rats with TP induced by the ferroptosis agonist, the proportion of Corynebacterium in the peritoneal cavity further increased. Corynebacterium is typically considered a commensal bacterium found on the skin and mucous membranes, but it can become an opportunistic pathogen in IAIs, especially in postoperative or immunocompromised patients. Its virulence factors, such as phospholipase D, can damage host cell membranes [43]. Corynebacterium is commonly associated with delayed wound healing and the formation of intra-abdominal abscesses [44]. It is noteworthy that, following TP induced by ferroptosis, the proportion of Facklamia in the peritoneal cavity significantly increased. These bacteria often migrate due to mucosal damage, leading to infections [45]. After including clinical factors in the analysis, an increase in serum amylase and pro-inflammatory cytokine levels directly correlated with an increase in bacterial species diversity, while anti-inflammatory factors were inversely related to bacterial biodiversity. These results suggest that, after TP, the ferroptosis of pancreatic acinar cells exacerbates microbial dysbiosis in the peritoneal cavity through the accumulation of inflammatory factors and oxidative stress (Fig. 9).

Fig. 9.

Ferroptosis-driven microenvironment promotes multidrug-resistant (MDR) bacterial colonization in TP. Traumatic pancreatitis induces ferroptosis of cells, leading to the accumulation of oxidative stress factors and iron ions in the abdominal cavity, thereby providing favorable conditions for the colonization of multidrug-resistant bacteria

Ferroptosis, infection dynamics, and the shift towards predictive and personalized approaches in tp

Following the onset of TP, serum biomarkers of ferroptosis in patients were significantly elevated, indicating a strong activation of this novel form of cell death. Ferroptosis, characterized by the rupture of cell membranes and the release of intracellular iron ions, may significantly exacerbate oxidative stress and trigger inflammatory cascades within the local tissue environment. This process may provide insights into the high detection rates of Acinetobacter baumannii complex in abdominal infections, as observed in this study. Iron overload, a direct consequence of ferroptosis, could serve as a critical factor in the proliferation of this bacterium [46]. Acinetobacter has evolved various iron acquisition mechanisms—such as siderophore production, heme uptake, and TonB-dependent transport systems—allowing it to thrive in the iron-limited conditions of the host. In the context of TP, where oxidative stress is markedly elevated, the increased levels of free iron further enhance the virulence of Acinetobacter and may contribute to its antibiotic resistance profiles.

These findings highlight a critical intersection between ferroptosis, microbial dynamics, and patient outcomes, underscoring the importance of targeting these mechanisms for improving clinical management [47, 48]. As iron overload facilitates bacterial virulence and antibiotic resistance, it is clear that managing ferroptosis could help control infection-related complications in TP patients, particularly in those infected with multidrug-resistant organisms like Acinetobacter [49]. Furthermore, ferroptosis-induced immunosuppression, a result of oxidative stress and tissue damage, impairs the ability of the immune system to control infections effectively. Specifically, the inhibition of NET formation and macrophage phagocytosis in the ferroptotic microenvironment provides optimal conditions for the colonization of these resistant pathogens [50, 51]. This creates a vicious cycle, where severe infections further enhance tissue damage, leading to worse clinical outcomes.

The integration of these findings into clinical practice highlights a clear trajectory towards transitioning from a reactive to a predictive and preventive care model in TP. From a predictive perspective, biomarkers related to ferroptosis can be utilized to identify patients at high risk of severe outcomes, such as infections, multi-organ failure, and IAIs. Early detection of elevated ferroptosis markers could trigger targeted interventions aimed at reducing oxidative stress, iron overload, and modulating microbial dysbiosis, which is a key factor in the development of secondary infections. The potential for microbiome modulation, particularly through probiotics, in conjunction with ferroptosis biomarkers, could offer an additional layer of precision in predicting and preventing complications. Probiotics could restore gut microbial balance and mitigate the dysbiosis-induced risk of infection, ideally within the first 48–72 h after injury, when the gut microbiome is most vulnerable to perturbations due to trauma and antibiotic use.

From a preventive approach, modulating ferroptosis pathways could reduce its detrimental effects before irreversible tissue damage occurs, potentially enhancing immune function and limiting the proliferation of harmful bacteria. This could be combined with personalized interventions based on individual microbiome profiles and ferroptosis markers. In the future, microbiome-based therapies, such as tailored probiotic strategies, could serve as adjunctive measures to conventional treatments, optimizing gut health and reducing systemic infections. Personalized treatment regimens, informed by specific patient biomarkers, would enhance therapeutic outcomes while minimizing side effects, representing a holistic and patient-centered approach to managing TP. This model of personalized, predictive and preventive medicine would not only improve survival rates but also reduce the incidence of long-term complications in TP patients.

Strength and limitations

This study found that blocking ferroptosis can alleviate pancreatic injury and reduce mortality in experimental animals caused by TP. The enhanced ferroptosis associated with TP exacerbates IAIs through the release of pro-inflammatory factors and increased oxidative stress. Future research should focus on developing therapeutic strategies that target the blockade of pancreatic acinar cell ferroptosis to treat IAIs.

This study has several limitations. First, while the trauma model in animals is well-established, the exploration of post-trauma ferroptosis at the cellular level remains insufficient. As a next step, we are planning to construct in vitro TP models by combining bio-reactive shock tubes and organoid systems to better investigate cellular-level processes. Second, the high mortality rate in rats due to traumatic pancreatitis complicated by intra-abdominal infections (TP-IAI) hindered the observation of long-term changes in the peritoneal microbiota. In response, we are exploring alternative animal models with improved survival rates and refining microbiome sequencing methods. Third, the use of only 16S rDNA sequencing to assess infection levels is limited in its resolution. We plan to integrate multi-omics approaches, including proteomics and metabolomics, to enhance the microbial profile. Finally, the relatively low incidence of TP complicated by abdominal infection resulted in a limited patient cohort. We are expanding our collaboration with multiple centers to increase sample size and improve bacteriological evaluations, particularly of drainage fluid. Future studies will address these limitations and provide more comprehensive insights.

While our study primarily focused on isolated traumatic pancreatitis, it is important to note that isolated TP is rare in clinical settings. Most pancreatic injuries occur in the context of polytrauma involving other organs such as the liver, spleen, and bowel, which presents different challenges. As such, our model may not fully reflect the complexity of such injuries, particularly in combat or multi-organ trauma scenarios. Future studies will aim to extend this model to incorporate polytrauma settings, providing a more clinically relevant framework for understanding TP in the context of broader trauma.

Conclusion and expert recommendations in the framework of PPPM

This study demonstrates the significant role of ferroptosis in TP and provides novel strategies for its management. Ferroptosis biomarkers, validated here, offer predictive value for early risk assessment, enabling timely interventions that can improve patient outcomes. Inhibition of ferroptosis was shown to reduce tissue damage and enhance survival in animal models, suggesting its potential as a preventive strategy to mitigate complications like multi-organ failure and IAIs. The study also emphasizes the importance of personalized medicine, highlighting how variations in the gut microbiota can influence therapeutic responses. Tailoring interventions based on individual molecular and microbiological profiles could optimize treatment efficacy and recovery. Overall, these findings support the integration of predictive biomarkers, preventive strategies, and personalized treatments, aligning with the principles of 3PM to improve TP management and patient outcomes.

Expert recommendation

Predictive approaches

This study demonstrates the significant predictive value of ferroptosis biomarkers, particularly TFR1 and ACSL4, in traumatic pancreatitis (TP). These biomarkers offer a promising approach for identifying high-risk patients at an early stage, allowing for timely interventions before the progression of severe complications. The integration of ferroptosis biomarkers into clinical practice could enhance predictive accuracy, facilitating early detection of patients at risk for high mortality and intra-abdominal infections (IAIs). By incorporating these predictive markers, healthcare providers can optimize patient management, ensuring that treatment strategies are adapted to individual risk profiles, thus aligning with the Predictive Medicine aspect of the 3PM framework.

Targeted prevention

The inhibition of ferroptosis represents a promising strategy for preventing tissue damage and improving survival rates in TP patients. Studies have shown that the excessive ferroptosis observed in pancreatic acinar cells significantly contributes to the progression of TP and the development of IAIs. By targeting ferroptosis early, either through the use of iron chelators or ferroptosis inhibitors, it is possible to limit the extent of tissue injury and prevent irreversible damage. This approach embodies the Preventive Medicine component of the 3PM framework, as it proactively addresses the root causes of disease progression and aims to reduce complications such as multi-organ failure and infection. Ferroptosis inhibitors should be prioritized for clinical trials to validate their therapeutic efficacy in TP patients and refine preventive strategies.

Personalization of medical services

This research underscores the importance of combining ferroptosis biomarkers with microbiome profiling to develop personalized treatment strategies for TP patients. The relationship between ferroptosis and gut microbiota offers valuable insights into how individual microbiome characteristics can influence the response to therapy. By integrating molecular and microbiological profiles, clinicians can tailor interventions that optimize therapeutic outcomes while minimizing adverse effects. This personalized approach enhances treatment precision, improving the efficacy of TP management and aligning with the Personalized Medicine component of the 3PM framework. Such an approach promises to enhance patient-specific care, improving overall survival and reducing the incidence of infection-related complications.

Abbreviations

TP

Traumatic pancreatitis

AP

Acute pancreatitis

IAIs

Intra-abdominal infections

SIRS

Systemic inflammatory response syndrome

ROS

Reactive oxygen species

SAP

Severe acute pancreatitis

SD

Sprague–Dawley

ELISA

enzyme-linked immunosorbent assay

TNF

Tumor necrosis factor

TGF

Transforming growth factor

IL

Interleukin

SOD

Superoxide dismutase

GSH

Glutathione

PCA

Principal component analysis

NCM

Neutral community model

GMHI

Gut Microbiome Health Index

MDI

Microbiome Disorder Index

db-RDA

Distance-based redundancy analysis

Author contribution

Zhirong Zhao: Writing – original draft, Visualization, Data curation. Ruiwu Dai: Validation, Methodology. Weiting Lu: Software, Resources, Methodology. Lan Ming: Methodology, Investigation. Jiamin Ji: Visualization, Validation. He Gan: Resources, Methodology, Investigation. Yuan Chen: Validation, Data curation. Ran Sun: Data curation. Qixia Jiang: Visualization, Resources. Zhaojie Wang: Data curation. Qian Huang: Writing – review & editing, Project administration, Funding acquisition, Conceptualization.

Funding

This work was supported by the National Natural Science Foundation of China (grant number 82070579).

National Natural Science Foundation of China,82070579

Data availability

The data that support the findings of this study are available on request from the corresponding author.

Declarations

Ethics approval and consent to participate

All animal experiments were approved by the Animal Ethics Committee of the Jinling Hospital (Approval No.: 2023JLHGZRDWLS-00087). The ethics of patients was approved by the Ethics Committee of the Jinling hospital (Approval No.: 2024DZKY-087–02).

Clinical trial number

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.