S100P is a core gene for diagnosing and predicting the prognosis of sepsis

Abstract

Sepsis, characterized as a severe systemic inflammatory response syndrome, typically originates from an exaggerated immune response to infection that gives rise to organ dysfunction. Serving as one of the predominant causes of death among critically ill patients, it's pressing to acquire an in-depth understanding of its intricate pathological mechanisms to strengthen diagnostic and therapeutic strategies. By integrating genomic, transcriptomic, proteomic, and metabolomic data across multiple biological levels, multi-omics research analysis has emerged as a crucial tool for unveiling the complex interactions within biological systems and unraveling disease mechanisms in recent years. Samples were collected from 23 cases of sepsis patients and 10 healthy volunteers from January 2019 to December 2020. The protein components in the samples were explored by independent data acquisition (DIA) analysis method, while Circular RNA (circRNA) categories were usually identified by RNA sequencing (RNA-seq) technology. Subsequent to the above steps, data quality monitoring was performed by employing software, and unqualified sequences were excluded, and conditions were set for differential expression network analysis (protein group and circRNA group were separately used log₂ |FC|≥ 1 and log₂ |FC|≥ 2, P < 0.050). Gene Ontology (GO) enrichment analysis and gene set enrichment analysis (GSEA) analysis were performed on common differentially expressed proteins, followed by protein-protein interaction between common differentially expressed genes and cytoscape software enrichment analysis, and subsequently its association with associated diseases (Disease Ontology (DO)) was investigated in an all-round manner. Afterwards, the distribution distinction of common differentially expressed genes in sepsis group and healthy volunteer group was displayed by heat map after Meta-analysis. Subsequent to the above procedures, pivotal targets with noticeable survival curve distinctions in two states were screened out after Meta-analysis. At last, their potential value was verified by in vitro cell experiment, which provided reference for further discussion of the diagnostic value and prognostic effect of target gene. A total of 174 DEPs and 308 DEcircRNAs were identified in the proteomics analysis, while a total of 12 common differentially expressed genes were identified after joint analysis. The protein-protein interaction (PPI) network suggested the degree of interaction between the dissimilar genes, and the heat map demonstrated their specific distribution in distinct groups. Through enrichment analysis, these proteins predominantly participated in a sequence of crucial processes such as intracellular material synthesis and secretion, changes in inflammatory receptors and immune inflammatory response. The meta-analysis identified that S100P is highly expressed in sepsis. As illustrated by the ROC curve, this gene has high clinical diagnostic value, and utimately confirmed its expression in sepsis through in vitro cell experiments. In these two groups of healthy people and septic patients, S100P demonstrated a more obvious trend of differential expression; Cell experiments also proved its value in diagnosis and prognosis judgment in sepsis; As a result, they may become diagnostic and prognostic markers for sepsis in clinical practice.

Keywords: Proteomics; RNA sequencing; S100P; Sepsis; Survival analysis.

Fig. 1

The timeline of this study is outlined as follows: (1) Selection of appropriate sample data; (2) Data analysis by employing the IDEP0.93 website; (3) Enrichment analysis of multi-omics differential genes or proteins through the Shiny GO website; (4) Intersection processing of the two datasets, with PPI adopted to demonstrate the degree of association between them; (5) Analysis of differential genes or proteins; (6) Verification of S100P by utilizing external validation datasets and Q-PCR.

Fig. 2

Data processing. (A–B) PCA plots of the proteome and circRNA groups. (C, D) Volcano plots of differentially expressed genes in circRNA and proteome groups. (E, F) The heat map distributions of the proteome and circRNA.

Fig. 3

Enrichment analysis of DEGs by employing the Shiny GO 0.77 tool. (A–C) Enrichment analysis typically involves three biological processes, namely biological process, Cellular Component and Molecular Function.

Fig. 4

GSEA Enrichment Analysis. (A–D) Represents the clustering range in the two groups, severally.

Fig. 5

Cross-linking of differentially expressed proteins (DEP) with circular RNA-associated proteins. (A) The twelve shared proteins formed in the intersection of both; (B) The interconnections between co-expressed differentially regulated proteins in the two datasets.

Fig. 6

The functional analysis of differentially expressed proteins. Section (A) demonstrates the distribution properties of diverse proteins; (B) elucidates their associations with respective diseases; sections (C–D) discuss the functional characteristics of distinct proteins from the perspective of functional effects.

Fig. 7

S100P was expressed at a high level in the sepsis group on the basis of meta-analysis (P < 0.010).

Fig. 8

Survival Analysis and ROC Curve for Core Genes. (A) In the survival curve, patients with low S100P expression demonstrate a higher 28-day survival rate compared to those with high expression; (B) The ROC curve reveals high sensitivity and diagnostic specificity of S100P.

Fig. 9

Q-PCR. Expression levels of S100P mRNA, with green representing the normal control group and red representing the sepsis group. *, p < 0.050* *, p < 0.010. * * *, p < 0.001.

Fig. 10

Spatial distribution of core genes at the single-cell level. (A) The horizontal and vertical axes represent the first and second principal components, separately, after dimensionality decrement. (B) Cells from distinct sample sources are differentiated by distinct colors. (C) The proportion of each cell group is displayed on the basis of inter-sample cell distribution variations. (D) The lowered-dimensionality cell groups are distinguished by dissimilar colors. (E) A heatmap illustrates the degree of correlation between cell populations. (F) The UMAP plot shows the annotation results for cell types. (G–H) S100P expression and distribution in human peripheral blood mononuclear cells (PBMCs).