Identification of novel biomarkers for sepsis diagnosis via serum proteomic analysis using iTRAQ‐2D‐LC‐MS/MS

Abstract Background Sepsis is a common cause of morbidity and mortality in the ICU patients. Early diagnosis and appropriate patient management is the key to improve the patient survival and to limit disabilities in sepsis patients. This study was aimed to find new diagnostic biomarkers of sepsis. Methods In this study, serum proteomic profiles in sepsis patients by iTRAQ2D‐LC‐MS/MS. Thirty seven differentially expressed proteins were identified in patients with sepsis, and six proteins including ApoC3, SERPINA1, VCAM1, B2M, GPX3, and ApoE were selected for further verification by ELISA and immunoturbidimetry in 53 patients of non‐sepsis, 37 patients of sepsis, and 35 patients of septic shock. Descriptive statistics, functional enrichment analysis, and ROC curve analysis were conducted. Results The level of ApoC3 was gradually decreased among non‐sepsis, sepsis, and septic shock groups (p = 0.049). The levels of VCAM1 (p = 0.010), B2M (p = 0.004), and ApoE (p = 0.039) were showing an increased tread in three groups, with the peak values of B2M and ApoE in the sepsis group. ROC curve analysis for septic diagnosis showed that the areas under ROC curve (AUC) of ApoC3, VCAM1, B2M, and ApoE were 0.625, 0.679, 0.581, and 0.619, respectively, which were lower than that of PCT (AUC 0.717) and CRP (AUC 0.706), but there were no significant differences between each index and PCT or CRP. The combination including four validated indexes and two classical infection indexes for septic diagnosis had the highest AUC‐ROC of 0.772. Conclusion Proteins of ApoC3, VCAM1, B2M, and ApoE provide a supplement to classical biomarkers for septic diagnosis.


| INTRODUC TI ON
Sepsis might lead to poor organ function or insufficient blood flow when the body responds to an infection, 1 which are common causes of morbidity and mortality in the intensive care unit (ICU) patients. It is estimated that there are annually 19.4 million sepsis cases with potentially high deaths in the high-income countries as North America, Europe, Asia, and Australia. Considering a higher prevalence of sepsis in the low-and middle-income countries, 2 it is suspected that the global epidemiological burden of sepsis is still difficult to be relieved.
Early diagnosis and appropriate patient management are the key to improve the survival and to limit disabilities in sepsis patients. 3 Sepsis is a heterogeneous syndrome, as a consequence, it is challenging to identify at early course of the disease, even based on the sepsis 2.0 or sepsis 3.0 diagnostic criteria. Proved by the extensive laboratory and clinical studies on sepsis, biomarkers are able to provide adjunctive information about the pathogenesis of sepsis and guide rapid diagnosis. Procalcitonin (PCT) level rises quickly after the onset of infection, serving as a superior biomarker for evaluating suspected septic patients. 4 C-reactive protein (CRP), interleukin-27 , and neutrophil CD64 (nCD64), with different sensitivity and specificity, were infection markers for early diagnose of sepsis. 5 Besides serving as infection markers, heparin-binding protein (HBP) and sphingosine-1-phosphate, which could induce endothelial cell dysfunction, were also involved in the pathogenesis and development of sepsis. [6][7][8] It is worth mentioning that the increase of HBP level can be detected 10.5 h prior to the development of organ dysfunction, which provides a reference for the early diagnosis and clinical management of sepsis. 9 Host immune system response is invoked early in sepsis to regulate both pro-inflammatory and anti-inflammatory responses. HLA-DR and PDL1 expression on monocytes, TNF production by LPS-stimulated whole blood cells, were potential biomarkers for innate immunity, and PD1 expression on CD4 + or CD8 + cells, IFNγ production by T cells, and number of circulating regulatory T cells were potential biomarkers for adaptive immunity in sepsis. 10 In 2010, Pierrakos and Vincent 11 summarized that at least 178 different sepsis biomarkers have been reported.
Only 16 factors were evaluated specifically for the early diagnosis of sepsis, and 5 of these, IL-12, interferon-induced protein 10 (IP-10), Group II phospholipase 2 (PLA2-II), CD64, and neutrophil CD11b with high sensitivity and specificity up to 90%. However, only few biomarkers have been used in the clinical testing so far.
As been called "one of the oldest and most elusive syndromes in medicine," 12 different biochemical and immunological pathways are involved in sepsis, and a variety levels of proteins have changed in human serum. In this sense, a set of biomarkers works superior to a single biomarker. Isobaric tag for relative and absolute quantification labeling coupled with two-dimensional liquid chromatographytandem mass spectrometry (iTRAQ-2DLC-MS/MS) is a proteomics method comparing samples between different groups to screen out a batch of differentially expressed proteins for further identification. 13 In recent years, several iTRAQ-2D-LC-MS/MS studies on sepsis were performed. Su et al. 14 brought insights into the prognosis of sepsis using iTRAQ-2D-LC-MS/MS. They identified seven urinary proteins and verified that downregulated LAMP-1 level may be useful for prognostic assessment of sepsis. Cao et al. 15 proved that proteins involved in the acute phase response, coagulation signaling, atherosclerosis signaling, lipid metabolism, and production of nitric oxide, and reactive oxygen species were associated with community-acquired pneumonia which would induce sepsis in the elderly. Jiao et al. 16 compared serum proteins of septic rats with controls, together with different time points of the survivor and non-survivor rats, finding five proteins were tightly correlated with the presence of sepsis after verified, and four proteins were related to the prognosis of sepsis.
Bian et al. 17 found the mechanism of H2 in the treatment of sepsis mice by proteomic approaches, which might be helpful for the clinical application of H2 in sepsis patients. The development of experimental techniques and methods provides a technical basis for exploring effective biomarkers for the early diagnosis of sepsis.
In the study, the changes in serum proteome between sepsis and non-sepsis groups were investigated by iTRAQ-2D-LC-MS/MS, and differentially expressed proteins were quantitative identified and validated in patients with non-sepsis and sepsis. The candidate protein biomarkers were validated as predictors for the diagnosis of sepsis.
Our study provides potential biomarkers for the early diagnosis of sepsis.

| Patients and control subjects
In order to study the early warning and standardized diagnosis and treatment system of sepsis, "database of the whole process management of early warning and diagnosis and treatment of sepsis," which was jointly developed by Zhejiang Hospital and Zhejiang University, Adult patients with clinically suspected infection exhibiting one or more of the following indicators were included (1) body temperature of<36°C or >38°C, (2) respiratory rate of >20 breaths/min, (3) pulse rate of>90 beats/min, (4) white blood cell (WBC) count of>12.0 × 10 9 /L or <4.0 × 10 9 /L, or immature granulocyte of >10%, and (5) chief complaint of fever or chills. Cases were excluded if they refused to sign the informed consent, or age<18 years, or diagnosed with malignant tumors. Patients were evaluated four times: at enrollment, the first day, the second day, and the third day according to Sepsis 2.0. 18 All the "severe sepsis" diagnosed by Sepsis 2.0 were defined as "sepsis" according to Sepsis 3.0, 1 and others were defined as "non-sepsis" in this article. Body temperature, pulse and respiratory rates, SOFA score and APACHE score, and blood samples were analyzed at the first evaluation. This study was approved by the Ethics Committee of Zhejiang Hospital (No. 2015-94K).
The levels of PCT (Roche Diagnostics, USA) and CRP (Beckman coulter, Brea, CA, USA) were measured by Zhejiang Hospital according to the manufacturer's instructions.

| Abundant protein depletion and protein digestion
To reduce the influence of individual variation and increase pre-

| LC-MS/MS analysis
The SCX fractions were analyzed using Easy-nLC1000HPLC system (Thermo Fisher Scientific) connected to Q-Exactive mass spectrometer (Thermo Fisher Scientific). Mobile phase A was 0.1% formic acid solution, and mobile phase B was 0.1% formic acid-acetonitrile solution (contain 84% acetonitrile). The chro- The protein-protein interaction network of these DEPs was constructed by Search Tool for the Retrieval of Interacting Genes database (STRING, https://www.strin g-db.org/) and visualized by Cytoscape software. The PPI pairs were extracted with a minimum required interaction score: >0.38. The degree of each protein node was calculated by Cytoscape software. Serum B2M protein and ApoE levels were detected on AU5821 and AU5421 Automatic biochemical analyzer (Beckman coulter).

| Differential proteins validation
B2M was agglutinated with latex particles coated with B2M antibody, and the turbidity was directly proportional to the concentration of B2M by immunoturbidimetry (AUTEC Diagnostica). ApoE was also detected by immunoturbidimetric method (Saike Biotechnology).

| Statistical analysis
Continuous variables were presented as mean ± standard deviation (SD) or the median with interquartile range (IQR), and categorical variables as numbers and percentages. Comparisons among nonsepsis, sepsis, and septic shock groups were performed using ANOVA test for means, chi-square test for the numbers, and Kruskal-Wallis

| Characteristics of the patients
A total of 125 patients, containing 53 patients of non-sepsis, 37 patients of sepsis, and 35 patients of septic shock, were enrolled in the study. The characteristics of the patients were compared in Table 1.
Compared to non-sepsis patients, sepsis patients and septic shock patients presented increased temperature, pulse rate, SOFA score and APACHE score, and decreased mean arterial pressure (p < 0.05).   Infection marker of PCT was showing increased trend, and CRP was shoving a peak value in sepsis group (p < 0.05).

| DISCUSS ION
Sepsis lacks effective early diagnostic biomarkers, which could lead to rapidly disease progresses and poor prognosis. With the development of proteomics, genomics, transcriptomics, and metabolomics, the researchers attempt to set up a batch of biomarkers for the diagnosis and prognosis of sepsis, which provide higher diagnostic specificity and sensitivity than a single biomarker. In this study, 37 differential proteins were screened between sepsis and non-sepsis  with an AUC of 0.772 in diagnosing sepsis, which was higher than the AUC of PCT (0.717) and CRP (0.706).
Endothelium dysfunction causes impaired perfusion, tissue hypoxia, organ dysfunction, and subsequent sepsis. VCAM1 appears to be associated with sepsis. VCAM was upregulated in sepsis and downregulated in non-survivors. 21 Elevated serum level of VCAM1 was a more powerful predictor for septic encephalopathy in adult community-onset sepsis on admission. 22 We found that the level of VCAM1 was showing an increased tread among non-sepsis, sepsis, and septic shock groups, with a peak value in the sepsis group.
Numerous studies demonstrated thatβ2-microglobulin (B2M) was a biomarker of renal function, whose serum levels were associated with glomerular filtration rate. 23 The level of serum B2M often in combination with the increase of cystatin C and urea ni- diseases, which might be an important indicator for diagnostic, prognostic, and therapeutic status. 35 In addition, this study showed that the serum level of ApoE was increased in sepsis and septic shock group, with a peak value in the sepsis group. Previous studies had shown that ApoE genotype was associated with sepsis. Wild-type (ε3) was associated with decreased incidence of sepsis, 36 and disease-associated (ε4) alleles were also found increased in coagulation system failure in human sepsis, which was a determinant of the human innate immune response to multiple TLR ligands. 37 Changes of lipid metabolism in sepsis were also re- Infection markers such as PCT and CRP are commonly used for the adjuvant diagnosis of sepsis, 4,5 and their role in the diagnosis and prediction of sepsis is still being explored. 45,46 In this study, a diagnostic model by a combination of the four validated indexes and two infection markers was established, with the highest AUC-ROC of 0.772 than that of PCT or CRP, although there were no significant differences between each combination AUC and AUC of PCT or CRP. B2M, VCAM1, and ApoE, besides ApoC3, already can be detected in clinical laboratory, which provided the clinical application possibility for septic diagnosis.

| CON CLUS IONS
In this study, the combination of ApoC3, VCAM1, B2M, and ApoE proteins were screened and identified as biomarkers for sepsis by using iTRAQ-2D-LC-MS/MS method. This is a new combination and a supplement to classical biomarkers such as procalcitonin or C-reactive protein.

ACK N OWLED G M ENT
The Authors thank Dr. Tingting Jiang from South China University of Technology School of Medicine for discussing and editing the article.

CO N FLI C T O F I NTE R E S T
No conflict of interest exists in the submission of this article.

AUTH O R CO NTR I B UTI O N S
Jing YAN conceived the study, reviewed the draft, and commented on it; Mo-lei YAN and Shang-zhong CHEN enrolled patient and established the database; Meng LI and Chen CHEN conducted the experiments; Meng LI and Rong-rong REN analyzed the data and wrote the draft. All authors reviewed the article and approved the final article.

DATA AVA I L A B I L I T Y S TAT E M E N T
The authors certify that this article reports original clinical trial data. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.