Application of blood biomarker CFP in preparation of diagnostic reagent for acute pancreatitis

Abstract

The application relates to the field of medicine biotechnology, and particularly relates to application of a blood biomarker CFP in preparation of an acute pancreatitis diagnosis reagent. The application proves that the CFP in blood can be used as a sensitive, effective and predictive biomarker for evaluating the severity of acute pancreatitis, and the detection convenience and result repeatability of the CFP are more advantageous in clinical application.

Description

Technical Field

[0001] The present invention relates to the field of medical biotechnology, and in particular to an application of a blood biomarker properdin (CFP) in the preparation of a diagnostic reagent for acute pancreatitis. Background Art

[0002] Acute pancreatitis (AP) is an inflammatory disease caused by abnormal activation of pancreatic enzymes. Complications can include local or systemic infection, systemic inflammatory response syndrome (SIRS), and multiple organ failure (MOF). Severe cases have a mortality rate as high as 20%-30%. The course of the disease varies greatly among patients, and early prediction of disease severity (within 48 hours) facilitates rapid and effective intervention.

[0003] Biomarkers currently in clinical use include C-reactive protein (CRP), procalcitonin (PCT), and interleukin-6 (IL-6). Other biochemical and hematological parameters, such as lipopolysaccharide binding protein, interleukin-8, neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR), are still under investigation but do not fully meet clinical needs. Clinical scoring systems such as the APACHE II score, the Ranson score, and the BISAP score also have limitations. Their analysis requires more than 48 hours of hospitalization and involves multiple complex parameters, limiting their clinical application. An ideal biomarker should be rapid, easy to use, readily accessible, cost-effective, and have high sensitivity and specificity.

[0004] As a powerful research method, proteomics has been widely used in biomarker screening for different types of AP biological samples. However, when screening candidate biomarkers derived from test results, the inherent nonlinear dynamic characteristics of the inflammatory process must be fully considered. This characteristic mainly stems from the multi-level regulatory mechanism of the pathogen and the host immune system (pro-inflammatory and anti-inflammatory responses). In this context, the limitations of traditional methods that focus on first-order statistical information (such as differential protein expression) have gradually become apparent, as they are difficult to effectively capture dynamic network interaction information. To address this challenge, the Dynamic Network Biomarker (DNB) theory proposed by researchers can identify early warning signs of disease stage transitions and rapid deterioration of complex diseases by mining high-order statistical information or differential correlations (such as correlation coefficients / covariances of proteomes).

[0005] Based on previous studies, we used the DNB method to perform proteomic analysis on peripheral blood samples from healthy people and patients with different AP subtypes, and found that CFP, a key positive regulatory protein in the complement cascade reaction, was significantly correlated with disease progression. CFP is mainly derived from neutrophils, and its expression level is not only related to the severity of the disease, but also reflects changes in the number and functional status of peripheral neutrophils, thereby revealing the patient's immunosuppressive state. Current immunosuppressive treatment strategies face major challenges in patient stratification and determining the optimal time for intervention due to the lack of reliable biomarkers. The present invention found that CFP has the potential to be a dual biomarker for both patient stratification and immunotherapy guidance, highlighting its good clinical application prospects in the precise management of AP. Summary of the Invention

[0006] The purpose of the present invention is to overcome the deficiencies of the prior art and provide an application of a blood biomarker CFP in the preparation of a diagnostic reagent for acute pancreatitis.

[0007] The objective of the present invention is achieved through the following technical solution: The present invention provides an application of properdin in the preparation of a reagent for diagnosing the severity of acute pancreatitis.

[0008] Furthermore, the diagnosis is performed by measuring the properdin concentration in the subject's blood. When the properdin concentration in the blood decreases by -40% to 50% relative to the properdin concentration in the normal control group (a negative decrease indicates an increase in the properdin concentration in the subject's blood relative to the control group), the diagnosis is made as non-severe acute pancreatitis. When the properdin concentration in the blood decreases by 50% or more relative to the properdin concentration in the normal control group, the diagnosis is made as severe acute pancreatitis. The subject is an acute pancreatitis patient.

[0009] Furthermore, the normal group refers to a group of people who have the same ethnic background as the subject and whose demographic characteristics match those of the subject, such as age, gender, and place of residence.

[0010] Furthermore, the diagnosis is performed by measuring the properdin concentration in the subject's blood, and the extent of the decrease in properdin concentration in the blood relative to the subject's own properdin concentration in the early stages of the disease or before the onset of the disease is negatively correlated with the severity of acute pancreatitis. The subject is an acute pancreatitis patient.

[0011] Furthermore, the properdin concentration in the blood of the subject and the normal group or the subject itself at the early stage of the disease or before the disease is detected using the same detection method.

[0012] Furthermore, the detection method includes electrochemiluminescence technology, enzyme-linked immunosorbent assay (ELISA), liquid phase chip technology (xMAP), single molecule immunoassay array technology (SiMoA), fully automatic capillary digital Western Blot, proximity extension assay (PEA), immunoelectrophoresis, high performance liquid chromatography tandem mass spectrometry or microfluidic chip.

[0013] Furthermore, the types of properdin in the blood include one or more of plasma properdin, serum properdin, and whole blood properdin. Existing research has demonstrated a clear correlation between the expression levels of properdin in whole blood, serum, and plasma. After hematocrit correction and standard centrifugation procedures, the detection values ​​exhibit equivalent diagnostic efficacy across the three sample types. Given that whole blood samples allow for point-of-care testing, serum samples avoid anticoagulant interference, and plasma samples maintain the continuity of routine testing, the choice of properdin can be tailored to the specific situation.

[0014] Furthermore, the uses of the reagent for diagnosing the severity of acute pancreatitis include judging the severity of acute pancreatitis patients, judging the progression or recovery of acute pancreatitis patients, using it alone or in combination to assist in guiding the formulation of acute pancreatitis medication regimens, judging the rationality of acute pancreatitis medication regimens, and judging the effectiveness of acute pancreatitis treatment drugs.

[0015] The present invention also provides a reagent for detecting blood properdin concentration for use in preparing a product for diagnosing the severity of acute pancreatitis. The reagent includes a kit.

[0016] The present invention also provides a detection system for judging the severity of acute pancreatitis, wherein the detection system comprises a biochemical index detection part and a calculation and analysis part;

[0017] The biochemical index detection part includes measuring the concentration of blood properdin;

[0018] The computational analysis part includes the classification of biochemical indicators and / or the evaluation of treatment effects and / or the formulation of treatment plans;

[0019] The classification of the biochemical indicators includes: classifying the acute pancreatitis patients' conditions into several categories, ranging from severe to non-severe acute pancreatitis, according to the range of the properdin concentration value in the blood of the acute pancreatitis patients;

[0020] The evaluation of the treatment effect includes: classifying the treatment effect of acute pancreatitis patients into several gradients from effective to ineffective according to the change of the concentration of properdin in the blood of acute pancreatitis patients over a certain period of time;

[0021] The formulation of the treatment plan includes: formulating a treatment plan based on the severity of the acute pancreatitis patient's condition; or formulating a treatment plan based on the treatment effect of the acute pancreatitis patient. Formulating a treatment plan, especially for immunotherapy, includes the administration time, duration or dosage, etc.

[0022] The beneficial effects of the present invention are:

[0023] Current clinical markers for assessing the severity of acute pancreatitis (AP) and those under development suffer from limitations such as insufficient sensitivity and specificity, and limited response times. Related hematological parameters also struggle to meet the needs for early prediction of disease progression and guidance for precise treatment. Existing clinical predictive scoring systems are limited in their application due to their complex and time-consuming analysis. The present invention provides a novel, efficient, and sensitive biomarker for early prediction of the severity of acute pancreatitis, full-course monitoring of the disease, and guidance and evaluation of immunotherapy. It has the following characteristics:

[0024] (1) Early prediction, high sensitivity and specificity:

[0025] Experimental results showed that plasma CFP, as an independent risk factor for acute pancreatitis, outperformed conventional blood markers (such as NLR, PLR, and MLR) in discriminating between severe and non-severe groups (AUC: 0.923), with a sensitivity of 92.6% and a specificity of 85.4%. The positive predictive value (PPV) was 0.926, and the negative predictive value (NPV) was 0.854. Plasma CFP levels show significant changes in the early stages of the disease, enabling early prediction of the severity of acute pancreatitis.

[0026] (2) Guiding immunotherapy:

[0027] Plasma CFP concentration not only reflects disease severity but is also closely associated with neutrophil function, characterizing the body's immunosuppressive state. In the treatment of a cerulean-zymosan (CAE-Zymosan) mouse model of SAP with the immunopotentiator GM-CSF, CFP concentration guided the selection of an appropriate intervention window, maximizing immune reconstitution while avoiding the risk of cytokine overstimulation. This significantly improved mouse survival and restored intestinal barrier integrity by reducing intestinal bacterial translocation, demonstrating the prominent role of CFP in guiding immunotherapy.

[0028] (3) Clinical translation potential:

[0029] The methods and markers provided by this invention offer advantages such as ease of use, high cost-effectiveness, and high accessibility, making them readily applicable in clinical practice. Dynamic monitoring of plasma CFP levels is expected to provide a new strategy for early assessment, full-cycle monitoring, and precise treatment of acute pancreatitis. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 Identifying critical transition states and severity biomarkers in AP progression based on DNB analysis; (A) Plasma EV proteomic analysis pipeline for three clinical cohorts. HC: healthy controls; NSAP: non-severe acute pancreatitis; SAP: severe acute pancreatitis; (B) Application of the DNB algorithm to reconstruct proteomic trajectories for detecting critical disease states; (C) l-DNB computational pipeline for identifying DNB-associated proteins in single samples; (D) Disease state-specific network perturbation profile. Nodes: proteins (red = state-specific DNB molecules); edges: Pearson correlation coefficient (PCC) fluctuations (thickness proportional to absolute value); (E) DNB-scored trajectory identification threshold for critical transitions; (F) State-stratified DNB subnetwork. Node size: change in edge connectivity (larger indicates stronger perturbation); edges: red (PCC gain), blue (PCC loss); (G) DNB subnetwork constructed based on public proteomic data (SAP vs. HC cohorts).

[0031] Figure 2Differential analysis of the DNB algorithm and private proteomics data. (A) Proteomic intensity distribution of each study group: HC (healthy control), NSAP (non-severe acute pancreatitis), and SAP (severe acute pancreatitis); (B) Dynamic evolution of the DNB network with disease state. Node size corresponds to the DNB score (quantifying contribution to network destabilization), with larger nodes representing key transitional biomarkers; (C) Disease-state-specific DNB subnetwork. Node size reflects edge perturbation strength (connectivity change magnitude), while edge thickness represents the absolute shift Pearson correlation coefficient (|sPCC|). Core DNB hubs are highlighted in dark brown; (D) Gene Ontology (GO) functional enrichment analysis of core DNB proteins that distinguish SAP from NSAP; (E) Venn diagram of the overlap of differentially expressed proteins (DEPs) between groups; (F) Volcano plot of differential protein expression: (i) NSAP vs. HC, (ii) SAP vs. HC. The dashed line marks the significance threshold (p value < 0.05 and |logFC| > 1.5); (G) Hierarchical clustering heat map of the expression profiles of differentially expressed proteins across groups.

[0032] Figure 3 Correlation between CFP and SAP severity; (A) Flowchart of patient enrollment and stratification (NT: healthy control; NSAP: non-severe acute pancreatitis; SAP: severe acute pancreatitis); (B) Comparison of plasma sC5b-9 concentrations among the NT, NSAP, and SAP groups; (C) Comparative analysis of plasma CFP levels among the three cohorts; (D) ROC curve evaluating the predictive efficacy of plasma CFP in distinguishing NSAP from SAP; (E) ROC curve evaluating the accuracy of plasma CFP in distinguishing NT from pancreatitis patients (NSAP + SAP); (F) Scatter plot of the correlation between plasma CFP concentration and the time interval from symptom onset to blood collection; (G) ROC analysis comparing CFP with traditional severity prediction indicators (NLR / PLR / MLR): CFP (AUC = 0.923), NLR (AUC = 0.752), PLR (AUC = 0.741), and MLR (AUC = 0.670). Data are presented as mean ± SD. Boxplots (B, C) show the median, interquartile range (boxes), and extreme values ​​(bars). The dashed line in the ROC curves (D, E, G) indicates the null reference line (AUC = 0.5). (H) Comparison of plasma CFP levels in patients with NSAP and SAP in this cohort whose samples were collected within 72 hours of onset.

[0033] Figure 4Mechanisms of complement activation and neutrophil granule release regulating plasma CFP; (A) C4d deposition on erythrocytes by clinical group (HC / NSAP / SAP; n = 5 biological replicates), with 1, 2, and 3 labeled in the figure representing biological replicates, the same below; (B) C4d expression on the surface of neutrophils (n = 5); (C) C5a receptor (C5aR) expression on the surface of neutrophils (n = 6); (D) CD64 and FPR1 levels on neutrophils (n = 3); (E) Western blot analysis of plasma CFP levels in HC and SAP groups treated with or without Zymosan (0.05 or 1 mg / ml); (F) Western blot analysis of CFP deposition in plasma from NT and SAP groups after incubation with Zymosan particles (0.05 or 1 mg / ml); (G) Immunofluorescence analysis showing CFP localization in neutrophil subgranules (scale bar: 5 μm); (H) Neutrophil CD66b expression (n=3); (I,J) Baseline CFP levels of freshly isolated neutrophils from each cohort; (K) Temporal CFP expression of healthy neutrophils under different conditions (serum-free RPMI, 10% FBS, 10% FBS + inflammatory stimulation); (L,M) CFP regulation of neutrophils in clinical groups under TNFα (50 ng / ml) and fMLP (100 nM) stimulation.

[0034] Figure 5 Comparative flow cytometric analysis of surface protein expression and complement deposition on peripheral blood erythrocytes and neutrophils. (A) Flow cytometric analysis of peripheral blood neutrophil isolation purity (CD11b+CD66b+) and CD235+ erythrocytes; (B) Flow cytometric analysis of CD64 and FPR1 receptor expression on neutrophils from different study groups (HC, NSAP, and SAP), with replicates; (C) C5aR receptor expression profiles on neutrophils from the HC, NSAP, and SAP cohorts, with numbers 1-4 indicating results from different independent collection groups; (D) Representative flow cytometric profiles of complement deposition and receptor expression on erythrocytes and neutrophils from synchronously processed HC, NSAP, and SAP samples. The accompanying table details the corresponding patient identifiers, isolated neutrophil concentrations, and plasma CFP levels.

[0035] Figure 6Correlation analysis between plasma CFP levels and peripheral neutrophil function in patients; (A) Neutrophil isolation yield from whole blood of healthy controls (HC), patients with non-severe acute pancreatitis (NSAP), and patients with severe acute pancreatitis (SAP) (n=8); (B) Real-time chemiluminescence kinetics of neutrophil reactive oxygen species (ROS) generation (n=3); (C) Representative images of crystal violet-stained neutrophils adhering to polycarbonate membranes in a Transwell assay (n=3); scale bar: 50 μm; (D) Quantitative analysis of neutrophils migrating to the lower chamber (n=3); (E) Neutrophil extracellular trap (NET) formation induced by Escherichia coli stimulation (neutrophil:bacteria = 1:3) in serum-free medium (0% FBS) (n=3); scale bar: 50 μm; (F) Neutrophil cytotoxicity against Escherichia coli (n=3); (G) Neutrophil signaling pathway activation (phospho-AKT, p42 / 44) Western blot analysis of ERK, p47, p40, p105, and p65, and CFP expression (stimulated with 10% FBS ± 100 ng / ml PMA). Total AKT, ERK, and GAPDH were used as internal controls. (H) ELISA analysis of plasma TNFα levels after whole blood LPS (100 ng / ml) stimulation (n = 7 samples). Data are expressed as mean ± standard deviation. *p < 0.01, **p < 0.001, ***p < 0.0001.

[0036] Figure 7 Functional and quantitative analysis of human peripheral blood neutrophils and monocytes. (A) Comparison of neutrophil yield per milliliter of whole blood among the HC, NSAP, and SAP groups (n = 8 per group); (B) Representative images of crystal violet-stained neutrophils migrating to the lower chamber in a Transwell assay, reflecting chemotactic ability (n = 3 per group); Scale bar: 50 μm; (C) Assessment of neutrophil extracellular trap (NET) formation after 4 hours of culture in RPMI 1640 medium containing Escherichia coli (neutrophil-to-bacteria ratio 1:3; n = 3 per group); Scale bar: 100 μm; (D) Quantitative ELISA analysis of plasma TNFα levels after ex vivo whole blood LPS stimulation (100 ng / mL; n = 7 per group); (E) Surface CD11b expression levels of neutrophils in the HC, NSAP, and SAP groups. Data are expressed as mean ± SD; statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001, p < 0.0001.

[0037] Figure 8Construction and phenotypic characteristics of the SAP mouse model. (A) Flowchart of the SAP induction experiment; (B) Kaplan-Meier analysis of survival rate during the 12-day observation period; (C) Dynamic changes in body weight of mice in the untreated group (NT), acute pancreatitis (AP), and SAP groups (n=5 / group); (D) Serum biochemical indicators: amylase (AMY), lipase (LIP), alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatine kinase isoenzyme (CK-MB), lactate dehydrogenase (LDH), urea, and creatinine (Cr) (n=5 / group); (E) Peripheral blood leukocyte differential count: total white blood cell (WBC), neutrophils (Neu), and lymphocytes (Lym) (n=5 / group); (F) Gross morphology of the small intestine and bacterial culture results of lavage fluid; (G) Peritoneal lavage fluid flow cytometric analysis: CD45+ leukocytes, CD11b+F4 / 80+ macrophages, and CD11b+Ly6G+ neutrophils (n = 5 / group); (H) Hematoxylin-eosin (HE) staining of pancreatic and small intestinal tissues for pathological evaluation; scale bar: 100 μm; (I) Expression of complement activation markers (CFP, C5b-9) on peripheral blood leukocytes (WBC, Neu) and bone marrow cells. Data are expressed as mean ± standard deviation. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

[0038] Figure 9 Study on plasma CFP as a biomarker for survival in the CAE-zymosan model; (A) Body weight changes of mice in the survival group (Su) and non-survival group (N-Su) on days 3.5, 5.5, and 6.5 after modeling; (B) Western blot analysis of plasma CFP and CFB in the NT, Su, and N-Su groups; (C) Peripheral blood cell counts (WBC, Neu, Lym), small intestinal pathological images, and quantification of bacterial load in lavage fluid (n=3); (D) Serum biochemical parameters (n=3); (E) Total number of bone marrow cells in bilateral femurs and tibias (n=3); (F) Number of neutrophils migrating to the lower chamber in the Transwell assay (n=3); (G) Killing efficiency of neutrophils against Escherichia coli (n=3); (H) NET formation in serum-free conditions ± E. coli (1:3) stimulation (n=3), scale bar: 50 μm; (I) LPS (5 Plasma TNFα levels after whole blood stimulation with 100 μg / ml of TNFα (n=3); (J) Western blotting of neutrophil signaling pathways (phosphorylated p65, p40, LC3B, and cleaved caspase-3) in the Su / N-Su group (10% FBS ± 100 ng / ml PMA), with Actin as an internal control. Data are expressed as mean ± SD, * p < 0.05, ** p< 0.01,*** p <0.001,**** p < 0.0001.

[0039] Figure 10 Surviving and non-surviving mice were compared for pathological features and neutrophil function. (A) Flow cytometry quantification of peritoneal leukocyte subsets: total leukocytes (CD45+), macrophages (CD1 lb+F4 / 80+) and neutrophils (CD1 lb+Ly6G+); gross morphology of small intestine and colony counts in intestinal lumen washes; (B) HE-stained sections of key organs (pancreas, small intestine, kidney, liver, lung) (scale bar: 100 pm); (C) total nucleated cell counts in bone marrow of both femur and tibia (n = 3 per group); (D) E. coli killing capacity of neutrophils (n = 3 per group); (E) Transwell migration assay quantification of neutrophils in lower chamber (n = 3 per group); (F) crystal violet staining of neutrophil adhesion on Transwell membrane (scale bar: 50 pm; n = 3 per group); (G) Western blot analysis of neutrophil signaling pathways: CFP, p-ERK, Rip3, GSDMD, CitH3 and Actin protein expression in bone marrow neutrophils of surviving (Su) and non-surviving (N-Su) mice (cultured in RPMI 1640 with 10% FBS ± 100 ng / ml PMA). Data are presented as mean ± standard deviation. Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001.

[0040] Figure 11 Longitudinal analysis of disease progression-related pathological parameters in the CAE-Zymosan model (6 time points), (A) Body weight dynamics in NT, AP and SAP groups at 12 h, 36 h, 3 d, 5 d, 8 d and 12 d after model induction (n = 3-5); (B) Plasma amylase (AMY) activity at each time point (n = 3-5; units: U / L); (C) Time distribution of circulating leukocyte subsets (WBC, Neu, Lym) (n = 3-5; units: x 103cells / mL); (D) Western blot analysis of plasma calprotectin (CFP) levels in NT, AP and SAP groups; (E) Plasma TNFa levels detected by ELISA after whole blood LPS stimulation (5 pg / mL). Data are presented as mean ± standard deviation. Statistical significance: *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

[0041] Figure 12Optimization of GM-CSF treatment regimen and efficacy evaluation in the CAE-zymosan model. (A) Body weight dynamics of mice in the non-treated (NT), SAP, and GM-CSF-treated groups (started at 0, 12, and 36 hours after zymosan administration) (n=3-5 per group); (B) Kaplan-Meier survival curves of each group; (C) Representative images of small intestinal gross morphology; (D) Colony counts in intestinal lavage fluid; (E) Flow cytometry quantification of peritoneal leukocyte subsets (n=3-5 per group); (F) Escherichia coli killing efficiency of bone marrow neutrophils (n=3-5 per group); (G) Body weight dynamics of mice in the NT, SAP, and different GM-CSF dosing regimens (200 ng / day for 3 days, 100 ng twice daily for 3 days, and 100 ng twice daily for 5 days) (n=3-5 per group); (H) Flow cytometry quantification of peritoneal leukocyte subsets at different GM-CSF doses (n=3-5 per group); (I) Small intestinal morphology and colony counts in intestinal lavage fluid; (J) Neutrophil-mediated Escherichia coli cytotoxicity under different GM-CSF regimens (n ​​= 3-5 per group); (K) Total nucleated cell counts in bilateral femoral and tibial bone marrow 0.5 hours after zymosan challenge (n = 3-5 per group); (L) Plasma TNFα levels were measured by ELISA 0.5 hours after LPS-stimulated whole blood cultures (5 μg / mL) (n = 4 per group); (M) Western blot analysis of plasma CFP protein levels in the NT, AP, and SAP groups 0.5 hours after the last GM-CSF administration. Data are expressed as mean ± SD. Statistical significance: *p < 0.05, **p < 0.01.

[0042] Figure 13Plasma properdin concentration-guided GM-CSF-based immunotherapy for SAP mice intervention; (A) Schematic diagram of the GM-CSF treatment experimental design; (B) FITC-dextran permeation assay to assess intestinal barrier integrity (n=5); (C) Kaplan-Meier analysis of survival rate; (D) Body weight dynamics in the GM-CSF-treated and untreated SAP groups (n=5); (E) Gross morphology of the small intestine (left) and colony counts on lavage fluid plates (right); (F) Flow cytometric analysis of leukocyte subsets in peritoneal lavage fluid: CD45+ total leukocytes, CD11b+F4 / 80+ macrophages, and CD11b+Ly6G+ neutrophils (n=3-5); (G) Flow cytometric analysis of the percentage of peripheral neutrophils in whole blood after erythrocyte sedimentation; (H) Quantitative analysis of the number of neutrophils migrating through the Transwell chamber in a migration assay (n=3); (I) Representative images of neutrophils adhered to polycarbonate membranes in a Transwell assay using crystal violet staining (n = 3). Scale bar: 50 μm. (J) Neutrophil extracellular trap (NET) formation induced by Escherichia coli (neutrophil to bacteria ratio of 1:3) combined with platelet-activating factor (PAF, 10 μM) in serum-free medium. Scale bar: 100 μm. (K) Neutrophil killing efficiency against Escherichia coli. Data are expressed as mean ± SD. p <0.05, p < 0.01, p < 0.001, **** p < 0.0001.

[0043] Figure 14 Table 1 Basic information of patients from whom EV proteomics was derived.

[0044] Figure 15 Table 2 Basic information of patients used for CFP concentration detection.

[0045] Figure 16 Table 3 Multivariate analysis.

[0046] Figure 17 Table 4 Comparison of the effects of CFP and common indicators for predicting the severity of acute pancreatitis.

[0047] Figure 18 Table 5 Summary of the follow-up results of the patients in different groups after blood sampling. DETAILED DESCRIPTION

[0048] The technical solution of the present invention will be further described in detail below with reference to the accompanying drawings, but the protection scope of the present invention is not limited to the following.

[0049] 1. Research ideas of the present invention

[0050] (1) Marker screening:

[0051] By using DNB analysis method, integrating private and public proteomic datasets, constructing bioinformatics analysis pipeline, through mining high-order statistical information or differential correlation (such as proteomic correlation / covariance), early warning signals of disease condition transition and deterioration can be more accurately captured.

[0052] (2) Patient sample validation:

[0053] In a prospective cohort study, patients with acute pancreatitis are included, and their plasma samples are collected. The plasma CFP level is detected by ELISA. Combined with clinical data, the correlation between plasma CFP level and disease severity and neutrophil function dynamics is analyzed to verify the potential of CFP as a biomarker for predicting disease severity.

[0054] (3) Animal model construction and evaluation:

[0055] A CAE-Zymosan mouse model is constructed to simulate the pathophysiological characteristics of human SAP, including local pancreatic inflammation, multiple organ dysfunction syndrome (MODS), and biphasic disease progression pattern. By monitoring the differences in plasma CFP levels and neutrophil function between surviving and non-surviving mice, the effectiveness of CFP as a biomarker for disease severity is further verified, and its application in guiding immunotherapy programs is explored.

[0056] II. Experimental materials and methods

[0057] 1. Patient inclusion and exclusion criteria

[0058] The present application includes patients with acute pancreatitis (AP) diagnosed in West China Hospital of Sichuan University from June 1, 2021 to December 31, 2024. The diagnosis is based on the revised Atlanta criteria. The inclusion criteria include: (1) Age 18-70 years old; (2) Meet the diagnostic criteria for AP. The exclusion criteria are: (1) Pregnancy; (2) Combined with chronic underlying diseases (such as chronic pancreatitis, hepatitis, nephritis); (3) Pancreatic or other malignancies; (4) Immune deficiency (such as HIV infection) or autoimmune diseases (such as systemic lupus erythematosus, rheumatoid arthritis); (5) History of immunosuppressive therapy; (6) Combined with non-AP related sepsis; (7) History of AP within 3 months before admission; (8) Data missing or repeated sampling. The research program is approved by the West China Hospital Ethics Committee (No. 2021-675) and follows the "Helsinki Declaration". The follow-up continues until the patient is discharged or dies in hospital. Peripheral blood samples are collected at admission for subsequent analysis.

[0059] 2. Plasma separation and neutrophil purification

[0060] Peripheral blood was collected in EDTA-2K anticoagulant tubes and centrifuged at 1500 g for 10 minutes at 4°C to separate plasma. The cell pellet was resuspended in normal saline, and peripheral blood mononuclear cells (PBMCs) and neutrophils were isolated using a human neutrophil isolation kit. Residual red blood cells were removed by hypotonic lysis. Neutrophil purity was verified by flow cytometry (CD66b+CD11b+ markers). Purified neutrophils were resuspended in RPMI 1640 medium for functional experiments.

[0061] Mouse experiment part: Neutrophils were isolated from the bone marrow of C57BL / 6J mice using the Ficoll-PLUS density gradient method, and the cell purity was verified by flow cytometry (Ly6G+ labeling).

[0062] 3. ELISA test

[0063] Commercial ELISA kits (catalog number: ab222864, manufacturer: Abcam, Cambridge, UK) were used to detect the concentrations of properdin (CFP) and soluble terminal complement complex (sC5b-9) in plasma, and the procedures were strictly followed.

[0064] 4. Dynamic Network Biomarker (DNB) Analysis

[0065] (1) Plasma collection and ultracentrifugation to separate extracellular vesicles (EVs)

[0066] Whole blood samples were collected by venipuncture in EDTA-anticoagulated vacutainer tubes and gently mixed 4–5 times to ensure adequate anticoagulant action. Plasma was separated using a two-step centrifugation procedure: an initial centrifugation at 2,500 × g for 15 minutes at 4°C, followed by a second centrifugation (using the same parameters) to obtain platelet-poor plasma (PPP). PPP samples were aliquoted and frozen at -80°C until further use. All procedures adhered to the ethical standards of the Declaration of Helsinki.

[0067] EVs were enriched using an optimized ultracentrifugation protocol: 300 μL aliquots of PPP were diluted 1:4 with phosphate-buffered saline (PBS) and transferred to polycarbonate ultracentrifuge tubes. An initial centrifugation (110,000 × g, 4°C, 90 min) was performed using a TLA-55 rotor (Beckman Coulter Optima XE). The EV pellet was resuspended in 1.2 mL of PBS and subjected to a second ultracentrifugation (using the same parameters) to remove soluble protein contamination. Finally, the purified EVs were resuspended in 30 μL of PBS and stored at −80°C for subsequent analysis.

[0068] (2) Liquid chromatography-mass spectrometry (LC-MS) analysis of EVs surface proteins

[0069] Plasma EVs proteomics analysis was performed on 88 samples (68 patients, 20 healthy controls) that met the inclusion and exclusion criteria. EVs were lysed with ice-precooled RIPA lysis buffer (containing protease / phosphatase inhibitors) and sonicated (30% amplitude, 3 seconds on / 10 seconds off, 5 minutes, ice bath) before centrifugation (10,000 x g, 4 °C, 30 minutes). The supernatant was taken to determine the protein concentration by the Bradford method. 100 μg of protein samples were sequentially reduced (10 mM TCEP, 56 °C, 1 hour), alkylated (20 mM iodoacetamide, room temperature, 30 minutes in the dark) and precipitated with methanol / chloroform / water (4:1:3, v / v). Proteins were digested with sequencing-grade trypsin (enzyme / substrate ratio 1:50, 37 °C, 12 hours), and peptides were desalted with C18 ZipTip. 5 μg of peptides were subjected to LC-MS analysis.

[0070] Desalted peptides were vacuum-dried and resuspended in mobile phase A (2% acetonitrile, 0.1% formic acid). Chromatographic separation was performed on a Thermo Scientific Nano EASY-nLC 1200 system coupled to an Orbitrap Exploris 480 mass spectrometer (nano-spray ion source). 1 μL of sample was enriched on a PepMap trap column (300 nL / min) and then separated on a C18 reversed-phase analytical column (250 mm x 75 μm, 1.9 μm Reprosil-Pur packing material) at a flow rate of 300 nL / min using a gradient elution (mobile phase B: 0.1% formic acid-80% acetonitrile; 2%-35% B in 65 minutes). Mass spectrometry parameters: survey scan (m / z 350-1800, resolution 60,000); secondary HCD fragmentation (30% collision energy, resolution 15,000); dynamic exclusion of singly charged / uncertain charge ions. Data acquisition was controlled by Xcalibur software (v4.3).

[0071] Raw data were processed by MaxQuant (v1.6.17.0): trypsin specificity was set to maximum 2 missed cleavage sites, precursor and fragment ion mass tolerances were 10 ppm and 0.02 Da, respectively. Fixed modification was cysteine carbamidomethylation, and variable modifications were methionine oxidation and N-terminal acetylation. Protein identification required >1 specific peptide (false discovery rate FDR <1%). Label-free quantification (LFQ) was performed based on iBAQ algorithm, and proteins containing >2 specific peptides were retained.

[0072] Data preprocessing included: log2 transformation of LFQ intensity mean standardization, median centering correction of batch effect between samples, and missing values were filled by mean matching method predicted by R language mice package (v3.15.0).

[0073] (3) Landscape Dynamic Network Biomarker (l-DNB) Method

[0074] In this study, we constructed a reference network to identify two disease progression states of acute pancreatitis (AP). The edges in the reference network were defined based on the Pearson correlation coefficient (PCC) calculated from protein expression values:

[0075] (Formula 1)

[0076] Where x_i and y_i represent the expression levels of proteins x and y, respectively, in the i-th reference sample, and x̄ and ȳ represent the average expression levels of proteins x and y across all n reference samples. When a new sample is added, the perturbed network is constructed by recalculating the PCC value including the new sample, denoted as PCC_(n+1). This PCC difference reflects the perturbation effect of the new sample on the network, and thus the single-sample Pearson correlation coefficient (sPCC) of the new sample relative to the reference sample is defined as:

[0077] (Equation 2)

[0078] When the sample size is large enough, the single-sample sPCC_n(x,y) follows an approximately normal distribution. This study uses statistical hypothesis tests (Z-test or U-test) to assess the significance of sPCC to determine whether proteins x and y in a specific sample are significantly correlated.

[0079] Within the framework of dynamic network biomarker (DNB) theory, when a physiological system transitions from a mild state to a critical state (severe state in this study), the protein expression of a specific molecular module usually meets the following three statistical conditions: (i) the protein expression deviation (ED_in) within the module increases significantly, indicating that protein expression fluctuations intensify; (ii) the correlation coefficient between proteins within the module (PCC_in) increases rapidly, reflecting the enhanced correlation between proteins within the module; and (iii) the correlation coefficient between proteins within and outside the module (PCC_out) decreases sharply, indicating that the correlation between modules weakens.

[0080] During the calculation process, the target protein and its main neighboring proteins in the sample-specific network (SSN) are regarded as local modules. According to the DNB theory, the local DNB score of protein x in sample E is defined as formula (3):

[0081] (Formula 3)

[0082] where sED_in represents the single-sample expression deviation of DNB module, which is used to evaluate the expression change of protein x in the sample. The DNB score of a sample is obtained by calculating the mean of local DNB scores of top K proteins (DNB module) in the SSN. According to the number of proteins in the SSN and the distribution characteristics of local DNB scores, K = 30 is set (the DNB score remains stable under different K values). The "state-level DNB score" of the disease is the mean of the DNB scores of all samples in the state, and the critical point is determined by the maximum DNB score, and the proteins in the critical point SSN are defined as DNB proteins (DNB module). To ensure the robustness of DNB proteins, only proteins ranked top K in at least 50% of patient samples in the same disease state are retained as state-specific DNB proteins, and the DNB score and ΔPCC in each state are calculated.

[0083] The state-specific DNB score is calculated using a frequency-weighted aggregation method:

[0084] The landscape DNB score is represented as , where f_s(x) is the frequency of protein x in the disease state:

[0085] (Formula 4)

[0086] N_present(x) represents the number of samples containing protein x, and N is the total number of samples in the disease state. The state-specific ΔPCC is calculated as formula (5):

[0087] (Formula 5)

[0088] where the frequency of protein pair (x, y) is defined as formula (6):

[0089] (Formula 6)

[0090] N_present(x, y) is the number of samples in which the interaction is detected, and N is the total number of samples in the disease state. This method captures early signals of critical transitions in biological systems through dynamic network features.

[0091] (4) Data analysis

[0092] Differential expression analysis of proteomics: The limma package (version 3.58.1) in R language was used to perform differential expression analysis of proteins in the healthy control group (HC), non-severe acute pancreatitis group (NSAP), and severe acute pancreatitis group (SAP) to identify differentially expressed proteins (DEPs). The identification of differentially expressed proteins is based on the criteria of adjusted p-value < 0.05 and |logFC| > 1.5.

[0093] Data Visualization: Violin plots were created to display expression distributions, volcano plots were constructed to highlight significantly differentially expressed proteins (adjusted p-value < 0.001, |logFC| > 1.5), and Venn diagrams were used to depict the overlap between differentially expressed proteins. Hierarchical clustering heat maps were generated using the pheatmap package (version 1.0.12) to reveal expression patterns among groups.

[0094] Functional enrichment analysis: The clusterProfiler package (version 4.10.1) was used to perform functional enrichment analysis, and GO and KEGG pathway analysis were performed to determine the enriched biological processes and pathways in the differentially expressed proteins (p < 0.05, q < 0.2).

[0095] Analysis environment: All analyses and visualizations were performed in R (version 4.3.3).

[0096] 5. Basic Experimental Part

[0097] (1) White blood cell (WBC) separation - erythrocyte sedimentation method

[0098] Fresh anticoagulated whole blood was mixed with erythrocyte sedimentation fluid in a 1:1 ratio and then diluted with 0.9% saline. The mixture was allowed to stand at 20 ± 2°C for 30 minutes to facilitate erythrocyte sedimentation. After phase separation, the upper leukocyte-rich fluid was carefully aspirated. Remaining erythrocytes were lysed using hypotonic ammonium chloride-potassium chloride buffer. The isolated leukocyte fraction was immunostained with fluorescein-conjugated anti-mouse antibodies (specific for CD45-APC, Ly6G-FITC, CD11b-PB, and CFP) for subsequent phenotypic analysis.

[0099] (2) Lipopolysaccharide (LPS)-induced TNFα production assay

[0100] Freshly anticoagulated whole blood was aliquoted into sterile polypropylene tubes and stimulated with ultrapure LPS (100 ng / mL for human samples and 5 μg / mL for mouse samples) for 4 hours at 37°C in a humidified incubator with 5% CO2. Following stimulation, cells were centrifuged at 3,000 g for 10 minutes at 4°C to pellet cellular components. The resulting plasma supernatant was collected, aliquoted, and stored at −80°C until cytokine quantification.

[0101] (3) Enzyme-linked immunosorbent assay (ELISA)

[0102] Human peripheral blood TNFα detection was performed using a high-adsorption 96-well plate: anti-human TNFα capture antibody (1:1000 dilution) was coated overnight at 4°C. The plate was blocked with 5% bovine serum albumin (BSA) for 1 hour at room temperature. Calibrator (recombinant human TNFα) was added and incubated with the sample for 2 hours. Subsequently, the plate was incubated with a biotinylated detection antibody (1:1000 dilution) and a streptavidin-horseradish peroxidase (HRP) conjugate (1:5000 dilution) for 1 hour each. After each step, the plate was washed four times and seven times with PBS containing 0.05% Tween-20 (PBST). The plate was developed with 3,3',5,5'-tetramethylbenzidine (TMB) and the reaction was terminated with 1 M H2SO4. The absorbance was measured at 450 nm (reference wavelength 570 nm) using a microplate reader.

[0103] (4) Flow cytometry and cell sorting

[0104] Human peripheral blood neutrophils were incubated on ice for 30 minutes and labeled with fluorescent antibodies against CD66b-FITC, CD11b-PB, C5aR-PE, FPR1-APC, and CD64-PE / Cyanine7. C4d was detected using an indirect labeling method: primary antibody (anti-C4d, 1:1500 dilution) was incubated for 1 hour, followed by secondary antibody (1:500 dilution) labeling for 30 minutes. Mouse peritoneal lavage cells were simultaneously labeled with CD45-APC, Ly6G-FITC, CD11b-PB, and F4 / 80-PE / Cyanine7. All samples were analyzed by flow cytometry using an Agilent flow cytometer and FlowJov 10.6 software.

[0105] Cell sorting was performed using human peripheral blood cells after erythrocyte depletion labeled with CD66b-FITC, CD3-APC, CD14-PE / Cyanine7, and CD11b-PB, and the cell volume was adjusted to 5×10 6 –1×10 7 Cells / mL, sorted by BD FACS Aria SORP system.

[0106] (5) Complement activation and CFP / C5b-9 deposition detection

[0107] Zymosan (1 mg / mL) was heat-activated by boiling for 30 minutes. 40% (v / v) plasma was diluted in gelatin barbiturate buffer (GVB++) containing Ca²⁺ / Mg²⁺ and incubated with zymosan at a final concentration of 0.05 or 1 mg / mL at 37°C for 2 hours. CFP deposition was quantified by Western blotting and flow cytometry, with simultaneous detection of C5b-9 complexes.

[0108] (6) Subcellular colocalization analysis

[0109] Peripheral blood neutrophils were attached to coverslips by low-speed centrifugation and fixed with 4% paraformaldehyde for 15 minutes. Blocking was performed with 10% goat serum for 1 hour at room temperature. Primary antibodies against CFP, LTF, MMP9, and MPO (1:100 dilution) were incubated overnight at 4°C, followed by incubation with AF488 / AF594-conjugated secondary antibodies (1:500 dilution) for 1 hour. After counterstaining with DAPI, the cells were visualized using an Olympus IXplore live cell imaging system.

[0110] (7) Western blot analysis

[0111] Tissue / cell proteins were extracted using pre-cooled RIPA lysis buffer containing protease / phosphatase inhibitors, and protein concentration was determined by BCA assay. After 12.5% ​​SDS-PAGE electrophoresis, the membranes were transferred to nitrocellulose membranes, blocked with 5% skim milk powder-TBST for 1 hour (room temperature), and then incubated with primary antibodies at 4°C overnight, including: anti-β-actin (1:5000), anti-Gapdh (1:5000), anti-AKT (1:5000), anti-phospho-AKT (1:2500), anti-phospho-p40phox (Thr154) (1:1000), anti-phospho-p47phox (Ser345) (1:1000), anti-ERK (1:5000), anti-phospho-p44 / 42 MAPK (Erk1 / 2) (Thr202 / Tyr204) (1:1000), anti-caspase-3 (1:1000), anti-p62 (1:1000), anti-LC3B (1:1000), anti-CFP (1:500), anti-phospho-NF-κB p105 (Ser932) (1:1000), anti-phospho-NF-κB p65 (Ser468) (1:1000), anti-DSDMD (1:1000), and anti-CitH3 (1:1000). After washing with TBST, the cells were incubated with HRP-conjugated secondary antibodies (1:5000, 1 hour at room temperature) and developed by chemiluminescence (ECL, Abbkine). Signals were acquired using a Touch Imager system. Plasma samples were diluted (human: 10x; mouse: 5x) and then denatured.

[0112] (8) Kinetics of Neutrophil CFP Release

[0113] Freshly isolated human neutrophils were cultured in RPMI 1640 supplemented with 10% fetal bovine serum (FBS), serum-free RPMI 1640, or 10% FBS-RPMI supplemented with lipopolysaccharide (LPS, 100 ng / mL), fMLP (100 nM), PMA (100 ng / mL), TNFα (50 ng / mL), C5a (20 nM), or GM-CSF (20 ng / mL). Cells were harvested at 0.5, 1, and 2 hours, and the dynamics of CFP expression were analyzed by western blotting.

[0114] 6. Neutrophil function test

[0115] (1) Reactive oxygen species (ROS) detection

[0116] Neutrophils (1×10 5 Cells (100 cells / well) were suspended in phenol red-free, serum-free RPMI 1640 medium and plated into a 96-well white microplate. After pre-equilibration at 37°C for 15 minutes, phorbol methyl paraformaldehyde (PMA, 100 ng / mL) and luminol (500 μM) were added. A control group contained luminol-only medium. Chemiluminescent signals were recorded in real time every 2 minutes for 90 minutes using a BioTek Synergy H1 microplate reader.

[0117] (2) Analysis of neutrophil extracellular trap (NET) formation

[0118] Human neutrophils (2.5×10 5 Cells (100 cells / well) were seeded in 48-well plates and stimulated with PMA (100 ng / mL) or Escherichia coli (MOI 3:1) for 3 hours at 37°C, 5% CO2. Unstimulated cells served as negative controls. Cells were fixed with 4% paraformaldehyde for 1 hour, permeabilized with 0.1% Triton X-100, and stained with SYTOX Green nucleic acid dye (100 nM) for 30 minutes in the dark. After washing three times with PBS, NET structures were observed using an Olympus IX73 fluorescence microscope.

[0119] (3) Neutrophil migration assay

[0120] Human peripheral blood neutrophils (1×10^6 cells / mL, suspended in RPMI 1640 medium containing 0.1% BSA) were seeded into the upper chamber of a 3 μm pore Transwell insert (Corning #353096). 500 μL of chemokine medium (RPMI 1640 / 0.1% BSA + 20 nM N-formylmethionyl-leucyl-phenylalanine (fMLP)) was added to the lower chamber. After incubation at 37°C for 3 hours, cells that had migrated to the lower surface of the membrane were fixed with 4% paraformaldehyde and stained with 0.1% crystal violet for 10 minutes. The slides were mounted and imaged using bright-field microscopy (Olympus IX73). Cells that had completely migrated into the lower chamber were collected by centrifugation (300 g, 5 minutes) and quantified on a hemocytometer. For experiments with mouse neutrophils, the fMLP concentration was adjusted to 50 nM, and the incubation time was extended to 4 hours.

[0121] (4) Bacterial killing experiment

[0122] Escherichia coli was cultured at 37°C in a shaker (200 rpm) until stationary phase. The cells were harvested by centrifugation (10,000 g, 3 minutes), washed once with sterile saline, and adjusted to a concentration of 1×10^9 CFU / mL. The bacterial suspension was then incubated with 20% (volume ratio) autologous serum at 37°C for 30 minutes to achieve opsonization. Neutrophils (0.8×10^6 cells / well) were co-incubated with the opsonized bacteria in RPMI-1640 at a 5:1 effector-target ratio for 90 minutes at 37°C. A control of the opsonized bacterial suspension without neutrophils was used. A 10 μL sample was lysed by adding 10 mM NaOH (pH 10.0), serially diluted, and plated on LB agar plates. The cells were incubated at 37°C for 18 hours, and the colony-forming units (CFU) were counted. The bactericidal efficiency was calculated as follows:

[0123] Bactericidal activity (%) = (CFU of experimental group / CFU of control group) × 100%

[0124] 7. Construction and Characterization of CAE-zymosan Mouse Model

[0125] (1) Optimized modeling method

[0126] Six- to eight-week-old male C57BL / 6 mice (weighing 22 ± 2 g, Beijing Huafukang Biotechnology Co., Ltd.) were housed in a specific pathogen-free (SPF) environment. All procedures adhered to the guidelines of the Sichuan University Laboratory Animal Ethics Committee (approval number #20190509023). Groups were randomly assigned to the study using ear tag numbers, and the experimental personnel and assessors were blinded.

[0127] After fasting for 12 hours, mice received nine intraperitoneal injections of 100 μg / kg cerulean (CAE) at 1-hour intervals, followed by a single intraperitoneal injection of 1 g / kg zymosan. Following model establishment, mice were allowed to resume free access to food. Pancreatic tissue and plasma samples were collected at various time points after zymosan injection for histopathological evaluation and biochemical analysis.

[0128] (2) Survival analysis experiment

[0129] Based on the optimized parameters, three models (n = 10 / group) were established: negative control (NT) treated with normal saline, acute pancreatitis (AP) treated with CAE alone, and severe acute pancreatitis (SAP) treated with CAE plus zymosan. Survival rate and clinical scores (activity, posture, and piloerection) were continuously monitored for 12 days and recorded twice daily.

[0130] (3) Hematological analysis

[0131] EDTA-K2 anticoagulated whole blood was collected by retroorbital venipuncture, and routine blood tests were performed using a Mindray BC-5120 fully automatic hematology analyzer.

[0132] (4) Plasma biochemical analysis

[0133] Plasma was obtained by centrifugation at 3,000 × g (4°C, 20 minutes) for EDTA-K2 anticoagulated blood. The following assays were performed using a Roche Cobas c311 biochemical analyzer using the following standard protocols: 1:10 diluted samples were analyzed for amylase (AMY) and lipase (LIP); stock samples were analyzed for alanine aminotransferase (ALT), aspartate aminotransferase (AST), creatine kinase isoenzyme (CK-MB), lactate dehydrogenase (LDH), blood urea nitrogen (BUN), and creatinine (Cr).

[0134] (5) Peritoneal lavage and analysis

[0135] Aseptically inject 5 mL of PBS into the peritoneal cavity, gently mix, and then recover the lavage fluid. For bacterial quantification, the lavage fluid was serially diluted and inoculated onto LB agar. Incubate at 37°C for 18 hours, and then count the colonies.

[0136] Immunophenotypic analysis: Lavage cells were labeled with CD45-APC, Ly6G-FITC, CD11b-PB, and F4 / 80-PE-Cy7 fluorescent antibodies, and leukocyte subsets were detected by flow cytometry.

[0137] (6) Histopathological evaluation

[0138] Liver, lung, kidney, pancreas, and small intestine tissues were fixed in 10% neutral buffered formalin for 24 hours, embedded in paraffin, and sectioned (5 μm). Hematoxylin and eosin (H&E)-stained sections were digitized using a 3DHISTECH Pannoramic MIDI whole-slide scanner (20× objective).

[0139] (7) Intestinal barrier function testing

[0140] Twelve hours after GM-CSF administration, FITC-dextran (40 kDa, 0.5 mg / g body weight) was administered orally. Serum was collected 4 hours later, and serum FITC-dextran concentration was measured by fluorescence (excitation wavelength 485 nm, emission wavelength 528 nm) and quantified using a standard curve.

[0141] 8. Comparative analysis of surviving and dead individuals

[0142] SAP model mice (n=20) were monitored four times daily for endpoints including hypothermia, bradykinesia, proneness, and respiratory distress. Mice meeting euthanasia criteria were euthanized via isoflurane overdose and tissues immediately processed. Surviving littermates receiving the same CAE-zymosan treatment served as controls. Measurements were consistent with those in the longitudinal study. All data were collected by a blinded investigator using precalibrated instruments, and statistical randomization was ensured by cage randomization.

[0143] 9. Optimization of GM-CSF treatment regimen guided by plasma properdin concentration

[0144] The CAE-zymosan model was longitudinally monitored for plasma CFP concentrations and whole-blood TNFα levels after LPS stimulation. Time-series analysis was performed at six time points: 12 hours, 36 hours, 3 days, 5 days, 8 days, and 12 days after model establishment (n = 30 / time point, 10 per group: NT group, AP group, and SAP group). Endpoint sample collection included systemic parameters: body weight, complete blood count (CBC), plasma amylase (AMY), and TNFα (ELISA); and semi-quantitative plasma CFP analysis (Western blot).

[0145] Based on the changes in plasma CFP concentration and TNFα production after whole blood LPS stimulation detected in the above study, the optimal dosing regimen of granulocyte-macrophage colony-stimulating factor (GM-CSF) was determined using a single-factor screening method (n=5 per group): administration timing: 0 hours, 12 hours, and 36 hours after SAP induction with zymosan; dosage regimen: 200 μg once daily for 3 days, 100 μg twice daily for 3 days, and 100 μg twice daily for 5 days.

[0146] The efficacy was evaluated by mortality, bone marrow neutrophil function testing, and intestinal barrier integrity indicators. Finally, the optimal regimen (starting 36 hours after SAP induction, 100 μg subcutaneous injection twice a day for 5 days) was selected for subsequent studies, and untreated SAP mice were used as positive controls.

[0147] 10. Data Analysis

[0148] Clinical data were analyzed using R software (v4.3.1). Categorical variables were expressed as frequencies (percentages) and analyzed using the chi-square test or Fisher's exact test. Normally distributed continuous variables were expressed as mean ± standard deviation and analyzed using the t-test. Nonparametric data were analyzed using the Mann-Whitney test. Variables with p < 0.05 in univariate analysis were included in multivariate logistic regression, and the optimal cutoff value for the receiver operating characteristic (ROC) curve was determined using the Youden index.

[0149] Experimental data were analyzed using GraphPad Prism (v10.0). Results are presented as mean ± standard deviation. Comparisons between groups were performed using unpaired t-tests, one-way analysis of variance with Bonferroni correction, or two-way analysis of variance with Tukey's post hoc test. Significance was defined as two-tailed p < 0.05.

[0150] Experimental Example 1: Identification of key states and disease severity predictors of acute pancreatitis using the landscape dynamic network biomarker (l-DNB) approach

[0151] The present invention is to study the cohort samples (demographic and clinical characteristics are detailed in Figure 14 ) were isolated and analyzed. Plasma protein mass spectrometry data were obtained from the supplementary materials of the reference (DIA-Based Proteomic Analysis of Plasma Protein Profiles in Patients with Severe Acute Pancreatitis). The raw data were pre-processed using the same workflow and incorporated into the l-DNB analysis presented in this paper. This reference contains proteomic data from three healthy volunteers (control group) and ten SAP patients (experimental group).

[0152] The present invention adopts the dynamic network biomarker (DNB) method to identify the critical state before AP worsens by analyzing the molecular correlation pattern and the interaction network topology, and construct a multi-molecular collaborative network to warn of disease progression.

[0153] Specifically, the present invention utilizes the landscape DNB (l-DNB) method to construct sample-specific molecular networks in three stages: Reference network construction: A stable reference network is established based on protein expression data from 15 healthy samples and high-confidence protein interaction information from the STRING database; Sample-specific network generation: Based on DNB theory, the Pearson correlation coefficient (PCC) of the three-stage protein expression values ​​for each patient sample is calculated, and statistical perturbation analysis is performed with control samples to construct a sample-specific network (SSN); Critical state identification: Local DNB scores are calculated step by step to form a landscape map representing the overall network state. To identify core DNB proteins and regulatory network modules, differential network analysis is further used to compare changes in SSNs before and after the estimated critical point. A voting mechanism is then used to select key nodes: only proteins that are hub nodes in more than half of the SSNs are retained as core proteins for subsequent functional validation. This method integrates single-sample expression features with a priori interaction information, significantly improving the sensitivity of network-level detection of disease critical states. The network-based key node detection (l-DNB) method is a model-free analysis framework that eliminates the large sample size required for machine learning and avoids the overfitting problem of traditional methods. This framework identifies key biomarkers in pathological progression by systematically evaluating three DNB criteria for each sample, based on the dynamic properties of critical transition states in biological systems.

[0154] Specifically, the study analyzed the proteomic profiles of plasma extracellular vesicles from 88 subjects and ultimately screened out 228 proteins that were stably expressed across samples ( Figure 1 A). The l-DNB method was used to identify the critical state of the proteome before disease progression ( Figure 1 B), Flowchart showing the three-step analysis strategy for identifying DNB core proteins from single-sample data ( Figure 1 C). There was no significant difference in protein expression intensity between the groups. The network topology constructed based on DNB theory and l-DNB method showed that ( Figure 1 D): The network of the healthy control group (HC) showed a localized fluctuation pattern, maintaining steady-state regulation; the network fluctuation of mild acute pancreatitis (NSAP) was the most significant, indicating the maximum network instability in the critical transition state; the network of severe acute pancreatitis (SAP) showed the coexistence of fluctuation amplification and topological reconstruction, and the core DNB protein formed a stable hub structure, suggesting that the pathological network architecture is strengthened through the mechanism of compensatory deregulation. DNB score analysis showed that the NSAP stage is the key window period for disease intervention ( Figure 1 E), confirmed that l-DNB can effectively identify the critical point before acute pancreatitis (AP) worsens.

[0155] We further screened novel biomarkers in DNB core proteins based on the change in the Pearson correlation coefficient (∆PCC) of the sample-specific network (SSN). Network edge types included: (i) correlation gain edge (PCC_(n+1) > PCC_n); (ii) correlation loss edge (PCC_(n+1) < PCC_n); and (iii) correlation unchanged edge (PCC_(n+1) ≈ PCC_n). Compared with the reference network, the regulatory networks at all stages of the disease showed significant changes ( Figure 1 F). APOH, APOA1, CFP, PLG, CLU and other proteins showed correlation gain or loss edge characteristics in the disease state. Among them, CFP showed correlation loss edge in both NSAP and SAP groups, indicating that its interaction partners were significantly reduced, and this trend was aggravated with disease progression. Public dataset verification showed ( Figure 1 G), the node degree of CFP in the SAP patient sample network changed most significantly, and its PCC value changed the most, further confirming the instability of the regulatory network centered on CFP during disease progression, suggesting its key role in the occurrence of SAP. In addition, CFP showed the highest local DNB score in the SAP state, indicating its transformation into a dynamic network hub, manifested by increased volatility and selective loss of connectivity, reflecting the reprogramming characteristics of the regulatory network ( Figure 2 AC).

[0156] GO enrichment analysis of DNB core proteins of SAP vs NSAP in plasma vesicle mass spectrometry data ( Figure 2 D), and it was found to be significantly enriched in pathways closely related to the severity of AP, including complement pathway (C3, C5, C9, CFP), protease inhibitory system (SERPINA1, SERPINC1, SERPIND1), oxidative stress (CAT), coagulation cascade (F2, FGA, FGG), fibrinolytic system (PLG) and inflammation-metabolism interaction pathway (TTR).

[0157] Compared with traditional differential expression analysis methods, the method of the present invention successfully detected CFP molecules that are associated with AP severity but have subtle expression fluctuations (usually <70% variation), which are easily overlooked by traditional analysis paradigms. To identify differential proteins between groups, the present invention uses conventional differential analysis methods ( Figure 2Dozens of differentially expressed molecules were identified in comparisons between the healthy control (HC) and non-severe acute pancreatitis (NSAP) groups, and between HC and severe acute pancreatitis (SAP) groups (adjusted p-value < 0.05 and |logFC| > 1.5). The most significant differences were in acute phase response proteins such as serum amyloid A (SAA), C-reactive protein (CRP), and lipopolysaccharide binding protein (LBP16). However, no significant protein expression differences were detected between the severe and mild groups (adjusted p-value < 0.05 and |logFC| > 1.5). Conventional analysis failed to identify the complement factor properdin (CFP) as a differentially expressed molecule due to two factors: limited CFP expression fluctuations: its maximum variation across study groups did not exceed 70%; and biphasic dynamics: as a complement system component, CFP exhibits a dynamic pattern of initial increase followed by a gradual decrease during disease progression. These results are consistent with literature reports and suggest that conventional differential analysis methods are insufficiently sensitive for biomarkers with complex dynamics such as CFP, making it difficult to distinguish disease severity.

[0158] Experimental Example 2 Plasma CFP concentration can be used to stratify patients with acute pancreatitis at different stages of their illness

[0159] The present invention included 142 patients with AP, and finally 122 patients were enrolled after screening by exclusion criteria, including 41 patients with SAP and 81 patients with non-severe acute pancreatitis (NSAP). Figure 3 A). There was no significant difference in baseline data between the two groups ( Figure 15 ).

[0160] Univariate analysis showed that lymphocyte count, albumin, blood glucose, CRP, CFP, IL-6, PCT, neutrophil / lymphocyte ratio (NLR) and platelet / lymphocyte ratio (PLR) were significantly different between NSAP and SAP groups. Figure 16 ).like Figure 16 As shown in the results, CFP was the only independent risk factor with statistical significance (p<0.05). Further detection of plasma sC5b-9 levels in each group showed that the expression level in the SAP group was the lowest but there was no statistical difference ( Figure 3 B).

[0161] Since CFP is not a routine clinical test indicator, the present invention determined its normal reference value range using 44 healthy controls (HC). The plasma CFP concentrations of the three groups were: HC group (23.16±5.324) μg / mL, NSAP group (19.11±6.305) μg / mL, and SAP group (9.908±3.311) μg / mL, with significant differences among the groups (p<0.0001). Figure 3 C).

[0162] When CFP was analyzed as a continuous variable, its concentration showed a significant gradient change among the three groups. Figure 3 D) shows that when the cutoff value was 11.565 μg / mL, the area under the curve (AUC) reached 0.923, sensitivity 0.926, specificity 0.854, positive predictive value (PPV) 0.926, and negative predictive value (NPV) 0.854. ROC curves of the HC group and the disease group (SAP+NSAP) ( Figure 3 E) Shows that when the cutoff value was 18.055 μg / mL, the AUC was 0.811, the sensitivity was 0.639, the specificity was 0.886, the PPV was 0.940, and the NPV was 0.470, indicating that CFP has a better predictive efficacy for distinguishing NSAP from SAP.

[0163] In view of the heterogeneity of the time window from onset to blood collection in patients, the present invention analyzed the temporal correlation between disease progression and plasma CFP levels. Figure 3 F) showed that plasma CFP concentration was significantly negatively correlated with the time since onset, and showed a gradual depletion pattern as the disease progressed. Comparative analysis with classic indicators such as NLR, PLR, and monocyte / lymphocyte ratio (MLR) further confirmed the prognostic advantage of CFP ( Figure 3 G and Figure 17 ). The plasma CFP level of the healthy control group in the present invention was 23.16±5.32 μg / mL, which was significantly higher than that reported in the literature: Stover et al. reported that the healthy population was 18.4±5.51 μg / mL (range 7.6-34.10 μg / mL), and a study of Southeast Asian populations showed 8.6 (8.0-9.3) μg / mL. These differences are due to race-related biological variations and technical differences in detection methods, suggesting that in multicenter biomarker studies and the establishment of population-specific reference ranges, it is necessary to systematically control demographic characteristics and methodological biases. Since clinical markers for early prediction of AP are needed, the patient groups in the above-mentioned NSAP and SAP groups whose blood sampling time was no more than 72 hours from the onset of the disease were analyzed and compared ( Figure 3 H), the results showed that the plasma CFP in the NSAP group was still significantly higher than that in the SAP group (p < 0.01), indicating that plasma CFP can also meet the needs and is worthy of verification in a larger cohort.

[0164] In addition, the progression of the disease in the patient cohort studied was tracked after blood sampling, and the ICU admission rate, surgery rate, length of hospital stay, and the occurrence of complications were recorded. The results showed that the prognosis of the NSAP group was significantly better than that of the SAP group. This study suggests that plasma CFP concentration not only reflects the patient's condition at the time of sampling, but also has a significant predictive effect on the patient's prognosis ( Figure 18).

[0165] In summary, plasma CFP is a dynamic biomarker associated with severity, and its predictive efficacy provides a clinically practical tool for early risk stratification of AP. The specific mechanism of action will be further elucidated in subsequent studies.

[0166] Experimental Example 3: Regulation of plasma CFP concentration mainly depends on the complement activation pathway

[0167] The mechanism of plasma CFP depletion involves three factors: (i) excessive consumption mediated by complement cascade activation; (ii) synthesis defects caused by neutrophil dysfunction; and (iii) impaired CFP mobilization due to disturbances in the granule exocytosis mechanism. This paper analyzes the mechanism based on the above hypotheses.

[0168] The degree of complement activation was assessed by flow cytometry to detect the level of C4d deposition on the surface of peripheral blood cells. Figure 4 and Figure 5 Peripheral blood neutrophils were purified by density gradient centrifugation with a purity greater than 95% ( Figure 5 A). In addition, peripheral blood red blood cells (CD235+) were also stained for analysis ( Figure 5 A). Activation of the classical / lectin pathway can cause C4 to cleave into C4d, whose covalent binding properties make it a persistent marker of complement activation. C4d on the red blood cell surface was only detected in some samples, with the SAP group expressing significantly higher levels than the other groups ( Figure 4 A). C4d on the surface of neutrophils was detected in all samples, and the expression level in the SAP group was significantly higher than that in the control group ( Figure 4 B). Neutrophil C4d deposition levels were significantly negatively correlated with plasma CFP concentrations ( Figure 5 D). However, there is no significant correlation between C5aR expression and disease severity ( Figure 4 C and Figure 5 C). In addition, the expression of CD64 (FcγRI) on the surface of neutrophils in the SAP group was upregulated, and the expression of FPR1 (fMLP receptor) was downregulated ( Figure 4 D and Figure 5 B), combined with the results of blood routine tests (increased neutrophils), it suggests that severe patients are in an infection state ( Figure 5 D) It is speculated that pathogen infection drives abnormal complement activation and accelerates CFP consumption.

[0169] In vitro complement activation experiments using zymosan, a component of the cell wall of Saccharomyces cerevisiae, showed that the CFP concentration in plasma decreased significantly after co-incubation with zymosan ( Figure 4E), confirming that CFP is consumed during complement activation. Confocal microscopy showed that the deposition of CFP on the zymosan surface was significantly reduced in SAP patient plasma compared with healthy controls ( Figure 4 F), suggesting that its complement activation capacity is weakened and pathogen clearance function is impaired.

[0170] Neutrophil quantitative analysis showed that the peripheral blood neutrophil count in the disease group was significantly higher than that in the healthy control group ( Figure 5 D), although the SAP group presented a higher level of neutrophilia ( Figure 5 D), whose plasma CFP expression is still low, excluding the possibility that insufficient neutrophil number leads to CFP deficiency. Laser confocal localization revealed that CFP in human primary neutrophils co-localized with the specific granule marker lactoferrin (LTF), but not with the gelatinase granule marker MMP-9 and the azuridine blue granule marker MPO ( Figure 4 G), confirming that it is stored in specific granules. The expression of CD66b (specific granule marker) on the surface of neutrophils in the SAP group was significantly higher than that in the NSAP and healthy groups ( Figure 4 H and Figure 5 D), suggesting that its specific granule exocytosis is enhanced. Analysis of CFP expression in neutrophils across cohorts ( Figure 4 I, J) showed heterogeneity among patients: some cases showed high CFP expression (due to inflammatory stimulation), while others showed CFP depletion (associated with sustained release). In vitro stimulation experiments confirmed that various inflammatory mediators can induce CFP release ( Figure 4 K, L, M), and no CFP secretion defect was found in severe patients.

[0171] The above results confirmed that abnormal activation of the complement system in SAP patients is the core mechanism driving the accelerated consumption of plasma CFP, providing a reasonable explanation for its decreased concentration.

[0172] Experimental Example 4 Study on the correlation between plasma CFP concentration and peripheral neutrophil function

[0173] After clarifying the association between peripheral neutrophil count and plasma CFP concentration, the present invention further explored the functional interaction mechanism between the two. Circulating neutrophil counts showed different distributions in SAP patients: 62% showed neutrophilia, 38% showed paradoxical neutropenia ( Figure 6 A and Figure 7 A). To reduce the potential interference of experimental variation on neutrophil function assessment, samples from healthy controls (HC), non-severe acute pancreatitis (NSAP) and severe acute pancreatitis (SAP) patients were collected simultaneously (completed within 40 minutes) for the experiment.

[0174] Functional characterization analysis showed that the production of reactive oxygen species (ROS) in neutrophils from SAP patients was significantly impaired ( Figure 6 B). Rapid ROS production is the core basis of neutrophil immune function, and its functional defects indicate impaired overall neutrophil function.

[0175] Transwell migration assay confirmed that SAP neutrophils had migration barriers: the cell retention rate on polycarbonate membrane increased ( Figure 6 C), chemotaxis efficiency was significantly reduced ( Figure 6 D and Figure 7 B). Migration defects are positively correlated with insufficient ROS production, while hyperactivated adhesion is associated with upregulated expression of surface markers such as selectins, integrin β2, and cell adhesion molecules. Adhesion experiments have shown that neutrophils in SAP patients have enhanced adhesion but significantly reduced migration activity. Adhesion is mediated by selectins, β2 integrins, and cell adhesion molecules (CAMs), with dynamic changes reflecting cell activation. Previous studies have shown that upregulated surface integrin expression promotes enhanced neutrophil-endothelial cell adhesion. This hyperadhesive state impedes neutrophil extravasation into lung tissue and alveoli, thereby impairing bacterial clearance in infected lungs. Therefore, hyperadhesive neutrophils in SAP are a key mechanism underlying impaired host anti-infective defenses.

[0176] The formation of neutrophil extracellular traps (NETs) was quantitatively assessed by Sytox Green fluorescence. Figure 6 E and Figure 7 C). Although mouse models have demonstrated that inflammatory mediator-driven NETosis can exacerbate pancreatitis, SAP neutrophils paradoxically exhibit reduced NET production. This phenomenon is caused by two interdependent mechanisms: (i) enhanced bacterial clearance activity of SAP neutrophils (phagocytic index: 4.7 ± 1.2 vs. HC 2.1 ± 0.6) causes elastase and myeloperoxidase to be retained in phagolysosomes, blocking the nuclear translocation required for chromatin decondensation; (ii) impaired NADPH oxidase activity limits ROS-dependent activation of peptidylarginine deiminase 4 (PAD4), a prerequisite for histone citrullination. The bactericidal ability of neutrophils in SAP patients is significantly impaired ( Figure 6 F), and is strongly correlated with defective ROS production. While SAP neutrophils exhibit diminished overall bactericidal capacity, their plasma bactericidal activity is paradoxically enhanced (data not shown). This paradox suggests that the bactericidal effect of SAP plasma primarily depends on the membrane attack complex (MAC) produced during complement activation: increased C4d deposition on peripheral blood cells in SAP patients indicates significantly enhanced complement cascade activity.

[0177] Signaling pathway analysis ( Figure 6 G) Display: SAP neutrophils show reduced p47, p40 and ERK phosphorylation, while AKT phosphorylation is enhanced. Impaired p47phox / p40phox phosphorylation suggests NADPH oxidase complex assembly disorder, directly explaining the ROS deficiency observed in functional experiments. ERK dephosphorylation reflects the interruption of the MAPK-ROS positive feedback loop, while AKT over-activation is a compensatory activation of the PI3K pathway. Elevated p105 phosphorylation suggests TLR4-driven classical NF-κB signaling activation, which is associated with circulating pathogen-associated molecular patterns (such as LPS) caused by bacterial translocation in SAP patients.

[0178] LPS-stimulated whole blood leukocyte TNFα production experiments show that peripheral monocytes in SAP patients exhibit an immunosuppressive phenotype, characterized by reduced TNFα secretion ( Figure 6 H and Figure 7 D). Notably, neutrophils in some SAP patients exhibit paradoxically high activation characteristics, which reflects the phase fluctuation of disease progression - neutrophil over-activation is gradually exhausted. Neutrophil dysfunction is mediated by multiple immunosuppressive stresses (C5a, LPS, cytokines and immunometabolism), but no significant difference in total CD11b expression on the surface of neutrophils is observed ( Figure 7 E).

[0179] The present application proposes that by dynamically monitoring the plasma CFP concentration, individual baseline differences can be corrected, and the progression of immune dysfunction can be longitudinally tracked, providing a new strategy for clinical evaluation of disease severity.

[0180] Experimental Example 5 Establishment and pathological characteristics of a SAP mouse model

[0181] In order to more extensively exploit the clinical application potential of CFP, the present application conducts research on a mouse acute pancreatitis model. The present application uses a modeling strategy that synchronously activates complement cascade and macrophage phagocytosis by intraperitoneal injection of zymosan. Based on the biological anti-degradation characteristics of zymosan, systemic inflammatory response and multiple organ dysfunction syndrome (MODS) can be persistently induced. The CAE-zymosan mixed model precisely simulates the characteristic SIRS-MODS dual pathological process of clinical SAP by combining caerulein-induced pancreatic interstitial edema with zymosan-mediated exacerbation.

[0182] The CAE-zymosan modeling scheme ( Figure 8 A) can reproduce the characteristic biphasic mortality pattern of SAP ( Figure 8B): The acute inflammatory phase (0-72 hours) presents rapid death (20-40%), and the subacute organ failure phase (72 hours-10 days) presents gradual death (15-30%). The maximum weight loss reaches 20% of the baseline, and the lowest value is seen 3-5 days after modeling ( Figure 8 C).

[0183] Pathological results of SAP mice 12 hours after modeling showed that biochemical tests showed that compared with the AP control group, the levels of amylase (AMY) and lipase (LIP) in SAP model mice increased by 3-5 times, and liver damage markers (ALT, AST) and renal function indicators (urea, creatinine) increased significantly ( Figure 8 D); Hematological indicators showed significant leukopenia with decreased neutrophil / lymphocyte counts ( Figure 8 E); Intestinal pathological evaluation showed that the length of the small intestine was significantly shortened (18.6±1.2 vs 27.4±1.5 cm), indicating severe intestinal inflammation; the colony-forming units of the peritoneal lavage fluid increased significantly, confirming bacterial translocation ( Figure 8 F, G).

[0184] Pancreatic histopathological evaluation showed typical SAP features: extensive acinar necrosis (accounting for 43±5% of the total tissue area), interstitial edema and dense inflammatory infiltration ( Figure 8 H). Flow cytometric analysis showed that the expression of terminal complement components was upregulated, and the density of membrane attack complex (MAC / C5b-9) on the surface of circulating neutrophils increased by 4.1 times compared with the control group, and that on bone marrow cells increased by 2.9 times ( Figure 8 I).

[0185] The CAE-Zymosan model effectively reproduces the core pathophysiological characteristics of human SAP: local pancreatic inflammatory response, MODS and biphasic disease progression pattern, providing a reliable platform for subsequent mechanism research and therapeutic intervention.

[0186] Experimental Example 6 Plasma CFP Concentration for Differentiating Survival from Non-survival Mice in the CAE-Zymosan Model

[0187] To elucidate a critical knowledge gap regarding the functional differences in peripheral neutrophils between clinical survivors and non-survivors, this study employed the CAE-Zymosan model to reveal heterogeneity in neutrophil function in mice with different survival outcomes. CAE-Zymosan-induced non-survivors exhibited significantly different pathophysiological phenotypes compared to survivors: after zymosan administration, non-survivors exhibited sustained weight loss and bradykinesia, and exhibited characteristic features distinguishing them from survivors before dying, including hypothermia, bradykinesia, prone recumbency, and respiratory distress.

[0188] The experimental results showed that the growth curve showed that the non-surviving mice showed no signs of recovery ( Figure 9 A); There was a significant difference in plasma CFP concentration between the two groups of mice ( Figure 9 B). Mice in the non-survival group showed shortened small intestine, bacterial translocation, and a large number of inflammatory cells (macrophages and neutrophils) in the peritoneal lavage fluid ( Figure 9 C and Figure 10 A).

[0189] Blood biochemical tests showed that the amylase levels of both groups of mice returned to normal, but the lipase levels of the mice in the death group increased significantly ( Figure 9 D). The survival group showed significant improvement in physiological function compared to the group sacrificed 12 hours later, while the major organ functions of the mice in the death group continued to be impaired, especially in the pancreas, as evidenced by acinar cell vacuolation / necrosis and inflammatory cell infiltration ( Figure 10 B).

[0190] Further functional analysis of mouse bone marrow neutrophils revealed that the total number of bone marrow cells in the survival group was within the normal range, while that in the death group was only half of the normal value or lower ( Figure 9 E and Figure 10 C). Neutrophil function in the 83% death group was significantly reduced, as evidenced by a decrease in the number of migrating cells, a decrease in bactericidal ability, and a decrease in NETs formation ( Figure 9 FH and Figure 10 DF). Functional similarity was observed only in one pair of mice during the 3.5-day observation period. The survival difference suggests that the death group continued in the acute inflammatory phase, and its pathological mechanism of death was related to systemic inflammatory response syndrome (SIRS). It is worth noting that only 33% of the mice in the death group showed more severe peripheral monocyte immunosuppression than the survival group, suggesting that the monocyte immunosuppression state cannot fully reflect the overall function of the mouse ( Figure 9 I).

[0191] The study of bone marrow neutrophil signaling pathways between the two groups showed that the most significant differences were in NADPH oxidase signaling, apoptosis, autophagy, and NF-κB signaling ( Figure 9 J and Figure 10 G).

[0192] Mice in the death group showed a significant decrease in bone marrow cell count and impaired neutrophil function, a phenomenon that was significantly positively correlated with decreased circulating CFP levels. This pathophysiological association is highly consistent with the clinical observations in patients with severe acute pancreatitis described above, suggesting a conserved mechanism in the progression of critical illness across species. Notably, within a population of mice with the same genetic background, the progression of acute pancreatitis was highly heterogeneous, as evidenced by significant inter-individual mortality differences, suggesting that the disease process is regulated by non-genetic factors, such as the microenvironment or immune status.

[0193] Experimental Example 7: Interventional study of GM-CSF-based immunotherapy guided by plasma properdin concentration in SAP mice

[0194] GM-CSF (granulocyte-macrophage colony-stimulating factor) is a pleiotropic cytokine expressed by hematopoietic cells that regulates myeloid progenitor cell differentiation, neutrophil survival, and effector functions. Studies in canine models have demonstrated that GM-CSF reduces bacterial translocation by enhancing neutrophil activity, supporting its use in acute pancreatitis. Human clinical trials have further demonstrated that GM-CSF, alone or in combination with interferon gamma (IFN-γ), can reverse monocyte dysfunction and enhance LPS-induced TNFα production, demonstrating its therapeutic potential.

[0195] These experimental results demonstrate significant inhibitory dysfunction in both neutrophils and monocytes from clinical SAP patients, and a similar phenotype is observed in non-surviving SAP mice. GM-CSF, due to its dual regulatory effects on both cell types, has become a potential therapeutic candidate. To implement precision medicine strategies, this study systematically evaluated circulating CFP levels combined with whole-blood-stimulated TNFα production as biomarkers to guide GM-CSF administration, aiming to identify the optimal therapeutic threshold that achieves immune reconstitution while minimizing the risk of cytokine overstimulation.

[0196] Combined with the pathological cycle of the CAE-Zymosan mouse model, several time points (12h, 36h, 3d, 5d, 8d, and 12d) were selected for longitudinal analysis to select the appropriate time for GM-CSF intervention.

[0197] The mice in the SAP group showed a continuous decrease in body weight ( Figure 11 A); Transient hyperamylasemia peaks at 12 hours and recovers at 72 hours ( Figure 11 B); In the acute phase (12-36 hours), the circulating neutrophil count decreased by 62% (0.9±0.2 vs 2.4±0.3 ×10³ / μL) and the lymphocyte count decreased by 47% (2.1±0.4 vs 4.0±0.5 ×10³ / μL). It is speculated that activated inflammatory cells under SAP pathological conditions were recruited to the lesions. In the recovery phase (≥3 days), the hematopoietic function of the SAP group was gradually restored.

[0198] The plasma CFP level in SAP mice showed a downward trend from 12h to 5d after induction, and rebounded in the late stage (8-12d), with individual heterogeneity ( Figure 11 CFB levels remained relatively stable throughout the experiment. The source of CFP in mice differs from that in humans. In addition to neutrophils, the liver, lymphocytes, and spleen also contribute significantly. This multi-tissue synthesis explains why the SAP-AP differences in mice are less pronounced than in human pathology.

[0199] LPS-stimulated TNFα production experiment showed that peripheral monocyte function was maintained until 36 h, followed by sustained inhibition (up to 5 d) Figure 11 D, suggesting that there is a temporal difference between monocyte and neutrophil functional impairment, reflecting the different regulatory mechanisms of the two in the SAP process.

[0200] The above longitudinal study results suggest that mice are in a critical period of illness from 0 to 36 hours after modeling, characterized by low plasma CFP concentration and / or inhibition of monocyte function. Based on this, three initial administration time points of 0 hours, 12 hours and 36 hours were screened. Subsequently, the treatment regimen was optimized to determine two key variables: (1) the treatment initiation time window; (2) the subcutaneous GM-CSF administration dose. The results showed that the 36-hour initiation of treatment can achieve the optimal therapeutic effect, which is manifested in high survival rate (100%), rapid recovery of body weight and small intestine length (2.5 days), improved intestinal barrier function (FITC-dextran leakage experiment) and inhibition of peritoneal bacterial translocation (bacterial colony count) Figure 12 A-E and Figure 13 B). In contrast, the 0-hour treatment group had a lower survival rate and sustained weight loss, and the 12-hour treatment group showed intermediate efficacy. The ranking of bactericidal activity was 36-hour group > 12-hour group > 0-hour group ( Figure 12 F). At 0 hours after modeling, immune cell function was intact (normal plasma CFP, sufficient bone marrow cells, and no impairment of neutrophil function) ( Figure 12 K-M); at 12 hours after modeling: plasma CFP decreased, bone marrow cells decreased, suggesting that neutrophil function may be impaired, but monocyte activity (TNFα secretion) was still preserved. This transient immune suppression state characterized by neutrophil dysfunction but preserved monocyte activity induced a secondary inflammatory exacerbation after GM-CSF administration. Treatment with GM-CSF needs to be based on real-time immune monitoring to avoid administration during the monocyte high response period to reduce the risk of proinflammatory. At 36 hours after modeling, SAP mice showed a synergistic immune suppression phenotype, with further decreased plasma CFP levels, continued progression of bone marrow depletion, and significant impairment of monocyte function (reduced TNFα secretion). Administration of GM-CSF at this stage can achieve the maximum therapeutic effect, and its mechanism is achieved by simultaneously repairing the function of myeloid cell lineages (neutrophils and monocytes). In summary, combined with the optimization of the treatment window based on plasma CFP levels and LPS-stimulated whole blood TNFα response, a theoretical framework for precise immune intervention in SAP is provided.

[0201] Three GM-CSF administration regimens were further tested: 200 ng single daily administration (3-day course), 100 ng twice daily administration (3-day course), and 100 ng twice daily administration (5-day course). Although there was a decrease in body weight at the beginning of the 5-day course, it was associated with significantly reduced bacterial translocation rate and enhanced bactericidal activity ( Figure 12GJ). The final optimized plan was to start treatment 36 hours after model establishment and continue administering 100 ng twice daily for 5 days ( Figure 13 A).

[0202] In the formal treatment experiment, the most significant effect of the GM-CSF group was the improvement of the mouse survival rate (100% vs. 40%) ( Figure 13 C). However, no significant differences were detected in body weight, bone marrow neutrophil function, and cell signaling pathways ( Figure 13 DK), which was due to the fact that the surviving mice in the untreated group had recovered on their own.

[0203] The intestine is a key hub for the progression of SAP, and pancreatic injury exacerbates intestinal barrier damage, driving bacterial translocation and systemic inflammatory cascades. This study found that GM-CSF repairs intestinal barrier integrity by enhancing neutrophil-mediated bactericidal activity, suggesting that intestinal bacterial dissemination is a core mechanism of intestinal damage in SAP. GM-CSF has a dual regulatory effect on immunosuppressive neutrophils and monocytes, and its efficacy depends on precise timing of administration. In summary, preclinical validation studies have confirmed the practical value of plasma CFP quantitative analysis in guiding immunotherapy regimens and established a translational research framework for CFP as a biomarker-driven neutrophil / monocyte targeted intervention strategies.

[0204] The foregoing description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the form disclosed herein and should not be construed as excluding other embodiments. Rather, the present invention can be used in various other combinations, modifications, and environments and can be modified within the scope of the concept described herein through the above teachings or techniques or knowledge in the relevant field. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention are intended to be protected by the appended claims.

Claims

1. Application of a reagent for detecting blood properdin concentration in the preparation of a product for diagnosing the severity of acute pancreatitis.

2. Use of the reagent for detecting blood properdin concentration according to claim 1 in preparing a product for diagnosing the severity of acute pancreatitis, characterized in that: The diagnosis is performed by measuring the concentration of properdin in the blood of the subject. When the properdin concentration in the blood decreases by -40% to 50% relative to the properdin concentration in the normal group, it is judged as non-severe acute pancreatitis; when the properdin concentration in the blood decreases by more than 50% relative to the properdin concentration in the normal group, it is judged as severe acute pancreatitis.

3. Use of the reagent for detecting blood properdin concentration according to claim 2 in the preparation of a product for diagnosing the severity of acute pancreatitis, characterized in that: The normal group refers to a group of people with the same ethnic background as the subject and matching some demographic characteristics.

4. Use of the reagent for detecting blood properdin concentration according to claim 2 in preparing a product for diagnosing the severity of acute pancreatitis, characterized in that: The properdin concentration in the blood of the subjects and the normal group was measured using the same detection method.

5. Use of the reagent for detecting blood properdin concentration according to claim 4 in preparing a product for diagnosing the severity of acute pancreatitis, characterized in that: The detection methods include electrochemiluminescence technology, enzyme-linked immunosorbent assay, liquid phase chip technology, single molecule immune array technology, fully automatic capillary digital Western Blot, proximity extension analysis technology, immunoelectrophoresis, high performance liquid chromatography tandem mass spectrometry or microfluidic chip.

6. Use of the reagent for detecting blood properdin concentration according to claim 2 in preparing a product for diagnosing the severity of acute pancreatitis, characterized in that: The types of properdin in the blood include one or more of plasma properdin, serum properdin and whole blood properdin.

7. Use of the reagent for detecting blood properdin concentration according to claim 1 in preparing a product for diagnosing the severity of acute pancreatitis, characterized in that: The uses of the product for diagnosing the severity of acute pancreatitis include single use or combined use to assist in guiding the formulation of a dosing regimen for acute pancreatitis.

Images