Identification of the critical genes and signaling pathways in subcutaneous adipose tissue after bariatric surgery based on the GEO database

doi:10.21203/rs.3.rs-4127363/v1

Download PDF

Article

Identification of the critical genes and signaling pathways in subcutaneous adipose tissue after bariatric surgery based on the GEO database

https://doi.org/10.21203/rs.3.rs-4127363/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

This study aimed to explore potential biomarkers and mechanisms following bariatric surgery. Two gene expression profiles from the Gene Expression Omnibus (GEO) were analysed to identify differentially expressed genes (DEGs) in subcutaneous adipose tissue (AT) post-bariatric surgery. Subsequently, Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), Gene-Set Enrichment Analysis (GSEA), and Protein-Protein Interaction (PPI) network analyses were employed to identify hub genes and associated pathways. Among the DEGs, 29 genes were downregulated. Enrichment analysis revealed that the downregulated DEGs significantly participated in inflammatory responses. GSEA provided comprehensive evidence that most genes were enriched in pro-inflammatory pathways before surgery, while after surgery, most genes were enriched in metabolism. In the PPI network, five key genes, including TREM2, MNDA, HP, C5AR1, and S100A8, were identified, with most validated as highly expressed in obesity by the Attie Lab Diabetes and another dataset, GSE72158. Bariatric surgery induces a significant shift from an obesity-promoting inflammatory state to an anti-inflammatory state, accompanied by improvements in adipocyte metabolic function. This represents a key mechanism for the enhancement of adipose tissue function following bariatric surgery. This study deepens the understanding of the benefits of bariatric surgery and provides potential biomarkers or therapeutic targets.

Health sciences/Biomarkers

Health sciences/Diseases

Health sciences/Gastroenterology

Bariatric surgery

obesity

adipose tissue

differentially expressed genes

PPI network

Obesity has emerged as a global epidemic [1], directly contributing to the onset of cardiovascular risk factors, including dyslipidemia, type 2 diabetes, hypertension, and sleep disorders [2]. According to the World Health Organization (WHO), projections estimate that by 2025, 2.7 billion adults will be overweight, with over 1 billion classified as obese [3]. Adipose tissue, exhibiting remarkable plasticity and activity, possesses functional pleiotropism and a high capacity for remodeling [4]. Beyond increased fat mass, obesity triggers inflammation in white adipose tissue (WAT), resulting in both local and systemic metabolic dysfunctions, notably insulin resistance (IR) [5]. The societal health and economic burdens imposed by obesity underscore the imperative to explore effective weight-loss treatment plans for implementation in clinical and public health contexts. Current drug therapies for obesity are limited in number and efficacy. Investigational drugs target various systems and tissues, including gastrointestinal hormones and adipose tissue. In addition to pharmacotherapy, noninvasive anti-obesity strategies such as novel drug delivery systems, vaccines, modulation of the gut microbiome, and gene therapy are being explored [6]. Common bariatric approaches in Asia, including sleeve gastrectomy (SG), Roux-en-Y gastric bypass (RYGB), and one-anastomosis gastric bypass (OAGB), performed laparoscopically, have shown more extensive and often more durable weight loss compared to lifestyle modification and obesity pharmacotherapy [7]. Studies in Asian and other populations indicate that bariatric surgery is more effective than conventional medical therapy in the long-term control of type 2 diabetes [8].

The principal mechanisms through which bariatric surgery influences obesity and its complications involve eating behaviours, altered gut hormone release, and changes in microbiota [9]. A prospective observational study identified an association between weight loss after bariatric surgery and changes in specific fatty acids [10]. Further research is warranted to elucidate the impact of bariatric surgery on lipid metabolism. Given the ongoing global health crisis posed by obesity, exploring potential biomarkers and mechanisms associated with the benefits of bariatric surgery is crucial for the development of individualized precision treatment in obese patients.

With the advancement and widespread application of high-throughput system (HTS) technologies and the development of bioinformatics, numerous potential biomarkers have been uncovered. Chen obtained the dataset from the GEO database and analysed bariatric surgery using Weighted Gene Co-expression Network Analysis (WGCNA), Least Absolute Shrinkage and Selection Operator (LASSO), and Support Vector Machine Recursive Feature Elimination (SVM-RFE) algorithms [11]. However, most bioinformatics analyses on bariatric surgery have not integrated 12 algorithms in CytoHubba in Cytoscape software and Upset.

In this study, we identified Differentially Expressed Genes (DEGs) in subcutaneous adipose tissue after bariatric surgery by analysing two mRNA expression profiles from the GEO database. Employing 12 algorithms in CytoHubba combined with the Upset R package enabled us to identify key genes regulating adipose differentiation. Finally, the Attie Lab Diabetes and the external dataset GSE72158 were employed for further validation.

1 Identification of Differentially Expressed Genes (DEGs) after Bariatric Surgery

In the pursuit of discerning DEGs pre- and post-bariatric surgery, and to standardise the dataset, 250 and 66 DEGs were respectively extracted from the GSE59034 and GSE53376 datasets (Fig. 1). The volcano plot illustrates the DEGs, revealing 250 in GSE59034, comprising 32 upregulated genes and 218 downregulated genes. In GSE53376, 66 DEGs were identified, encompassing 14 upregulated genes and 52 downregulated genes. Subsequently, heatmaps were constructed, with the GSE53376 heatmap based on DEGs with |logFC|>1 and the GSE59034 heatmap based on the top hundred DEGs with |logFC|>1. A Venn diagram was then utilised to identify 29 commonly expressed DEGs (Table 1) (Fig. 2).

Table 1

Differentially expressed genes after bariatric surgery.
DEGs	Gene symbol
Down-regulated	SPP1,CHI3L1,SFRP4,PRG4,C3AR1,MSC,FPR3,NPR3,CD163,RANBP3L,HP,AQP9, RGS1,HSD11B1,CCR1,TREM2,IL18,GPR183,LYZ,CD180,DHRS9,C5AR1,CXCL16, CTSS,S100A8,CD53,EVI2A,MNDA,EVI2B

2 Gene Ontology and KEGG Pathway Enrichment Analyses after Bariatric Surgery

The Gene Ontology (GO) enrichment results for biological processes (BP) indicate that differentially expressed genes are significantly enriched in the inflammatory response and cell activation involved in immune response. Results for cellular components (CC) reveal significant enrichment in the intrinsic component of the membrane, secretory granules, and secretory vesicles. In terms of molecular function (MF), enrichment is observed in G protein-coupled receptors and molecular transduction activity. The KEGG pathway enrichment results demonstrate that differentially expressed genes are enriched in the Staphylococcus aureus infection signaling pathway, cytokine-cytokine receptor interaction signaling pathway, complement and coagulation cascade lysosomes, and neuroactive ligand-receptor interaction signaling pathway (Fig. 3).

3 Gene Set Enrichment Analysis (GSEA)

The objective of the GSEA analysis is to elucidate potential mechanisms associated with bariatric surgery. Samples were categorised into the pre-bariatric surgery group and the post-bariatric surgery group. The analysis reveals (Fig. 4) that the most significantly enriched pathways in the pre-bariatric surgery group encompass the chemokine signalling pathway, toll-like receptor signalling pathway, hematopoietic cell, lysosome, B cell receptor signalling pathway, cytokine-cytokine receptor interaction, natural killer cell-mediated cytotoxicity, NOD-like receptor signalling pathway, leukocyte transendothelial migration, T cell receptor signalling pathway, cell adhesion molecules, and Fc gamma receptor-mediated phagocytosis, among others. Additionally, pathways exhibiting a positive correlation with the post-bariatric surgery group and displaying significant differences include propionate metabolism, fatty acid metabolism, citrate cycle, glycerophospholipid metabolism, alanine, aspartate, and glutamate metabolism, as well as linoleic acid metabolism, among others.

4 Protein-Protein Interaction (PPI) Network Analysis and Key Gene Selection

Based on disparities identified in STRING, the Protein-Protein Interaction (PPI) network of differentially expressed genes comprises 22 downregulated genes, forming a cluster of 22 nodes and 50 edges (Fig. 5). Protein interaction network graph data of differentially expressed genes, acquired from the STRING online database, were exported and subsequently analysed in Cytoscape software, resulting in an in-depth analysis of the protein interaction network graph by Cytoscape. The gene ranking methodology utilises 12 algorithms within the CytoHubba plugin to select the top 15 genes for each algorithm (Table 2 and Table 3). Gene selection utilises the UpsetR package in R software, as depicted in the gene set visualisation graph. Ultimately, MNDA, TREM2, C5AR1, S100A8, and HP are identified as key genes. Expression results of key genes before and after bariatric surgery indicate a varied reduction in postoperative expression compared to pre-surgery. Pearson correlation analysis of key genes reveals a positive correlation between the five genes in pairs.

Table 2

CytoHubba plugin 12 algorithms get the top 15 genes(1)
MCC	DMNC	Stress	Radiality	MNC	EPC
CTSS	MNDA	CD53	CCR1	CTSS	CD163
MNDA	C5AR1	CD163	CD163	CD163	CTSS
CD53	TREM2	CCR1	CTSS	CCR1	CD53
LYZ	C3AR1	CTSS	C3AR1	MNDA	CCR1
C3AR1	LYZ	C3AR1	CD53	CD53	LYZ
CCR1	S100A8	LYZ	LYZ	C3AR1	MNDA
CD163	EVI2B	MNDA	MNDA	LYZ	C3AR1
C5AR1	CD53	IL18	TREM2	C5AR1	TREM2
TREM2	AQP9	TREM2	IL18	TREM2	C5AR1
S100A8	GPR183	S100A8	C5AR1	IL18	S100A8
EVI2B	HP	C5AR1	S100A8	S100A8	EVI2B
IL18	CXCL16	CHI3L1	EVI2B	EVI2B	IL18
AQP9	CD180	EVI2B	HP	AQP9	AQP9
HP	EVI2A	HP	AQP9	GPR183	HP
CHI3L1	CHI3L1	SPP1	CXCL16	HP	CHI3L1

Table 3

CytoHubba plugin 12 algorithms get the top 15 genes(2)
Degree	EcCentricity	ClusteringCoefficient	Closeness	BottleNeck	Betweenness
CD163	CD163	GPR183	CD163	CTSS	CD53
CTSS	IL18	CXCL16	CTSS	CCR1	CD163
CD53	CTSS	CD180	CD53	CD163	CCR1
CCR1	S100A8	EVI2A	CCR1	CD53	CTSS
MNDA	C5AR1	RGS1	MNDA	IL18	IL18
C3AR1	MNDA	EVI2B	C3AR1	MNDA	C3AR1
LYZ	C3AR1	AQP9	LYZ	HP	MNDA
IL18	CD53	C5AR1	TREM2	S100A8	LYZ
S100A8	AQP9	TREM2	IL18	EVI2B	TREM2
C5AR1	CCR1	C3AR1	C5AR1	CHI3L1	S100A8
TREM2	TREM2	LYZ	S100A8	TREM2	C5AR1
EVI2B	LYZ	MNDA	EVI2B	GPR183	CHI3L1
HP	GPR183	S100A8	HP	FPR3	HP
AQP9	FPR3	HP	AQP9	CXCL16	EVI2B
CHI3L1	HP	CTSS	CHI3L1	C5AR1	SPP1

5 Cell Type Enrichment Analysis

To comprehensively investigate alterations in immune cell infiltration levels post-bariatric surgery, this study employed xCell analysis to examine the relative abundance of immune cells, as well as changes in adipocytes and pre-adipocytes before and after surgery (Fig. 6). The findings unveiled a noteworthy decrease in adipocytes, pre-adipocytes, B cells, dendritic cells, macrophages, M1 macrophages, M2 macrophages, monocytes, and neutrophils post-surgery. The expressions of TREM2, HP, MNDA, and C5AR1 demonstrated positive correlations with macrophages, M1 macrophages, M2 macrophages, dendritic cells, and monocytes. Furthermore, the expression of S100A8 exhibited positive correlations with neutrophils and monocytes. Additionally, all five key genes demonstrated positive correlations with adipocytes or pre-adipocytes (Fig. 7).

6 Preliminary Validation of Key Genes

The mRNA levels of hub genes in obesity underwent validation using the Attie Lab Diabetes Database. The findings revealed that, in comparison to the lean group, the expression of TREM2, HP, C5AR1, and S100A8 was up-regulated in the 10-week obese mice (Fig. 8). Subsequently, the GSE72158 dataset from the GEO database was employed to confirm the expression of MNDA, TREM2, HP, C5AR1, and S100A8 before and after surgery. The results (Fig. 9) indicated that following surgery, the expression of TREM2, MNDA, HP, C5AR1, and S100A8 exhibited varying degrees of reduction compared to pre-surgery levels.

7 Diagnostic Efficacy of Key Genes for Obesity

Through a collective analysis of the obese and healthy populations in the GSE59034 dataset, ROC curves for hub genes were drawn using the pROC package. Subsequently, the area under the curve (AUC) was calculated to assess the diagnostic effectiveness of the hub genes (Fig. 10). The Area Under the Curve (AUC) for TREM2 stands at an impressive 0.918, indicating superior diagnostic performance for obesity. HP follows with a respectable AUC of 0.848, while MNDA exhibits a value of 0.828. C5AR1 shows a slightly lower but still commendable AUC of 0.879. S100A8, with an AUC of 0.762, indicates a moderate level of diagnostic utility. In light of these AUC values, it is evident that TREM2 is highly efficacious in diagnosing obesity, with HP, MNDA, C5AR1, and S100A8 also contributing to the diagnostic process, albeit to varying degrees.

8 Clinical Data Validate Changes in Patient Inflammatory Status Before and After Bariatric Surgery

8.1 Basic Information

Clinical data from 200 obese patients, selected based on inclusion and exclusion criteria from 2020 to 2023, were analysed. The demographic details were as follows: a total of 200 cases, comprising 168 female patients (84%) and 32 male cases (16%). The mean body weight was (101.20 ± 16.59) kg, and the mean BMI was (36.81 ± 4.69) kg/m2. One hundred and forty patients completed a one-year follow-up.

8.2 Inclusion and Exclusion Criteria

(1) Inclusion criteria: According to the "2019 Chinese Guidelines for the Surgical Treatment of Obesity and Type 2 Diabetes," patients with simple obesity with BMI ≥ 32.5 kg/m² or BMI ≥ 27.5 kg/m² and two metabolic syndromes or type 2 diabetes were included. Patients with no apparent surgical contraindications who agreed to participate in the clinical study were also considered.

(2) Exclusion criteria: Non-compliance with medical advice and poor compliance, BMI below 25kg/m2, type 1 diabetes patients, and individuals unable to tolerate laparoscopic surgery.

8.3 Data Collection

Clinical data, including BMI, neutrophil count (NEU), lymphocyte count (LYM), neutrophil-to-lymphocyte ratio (NLR), triglycerides (TG), high-density lipoprotein (HDL), and TG/HDL, were collected both before and one year after bariatric surgery(Table 4). Changes in BMI and markers of low-grade inflammation (NLR, TG/HDL) were observed before and after surgery.

Table 4

Comparison of the observation indicators before and 1 year after surgery
Index	Pre-surgery	Post-surgery	P
Example(n)	200	140	-
height(m)	1.65 ± 0.07	1.65 ± 0.07	> 0.05
weight(kg)	101.20 ± 16.59	73.11 ± 9.68	< 0.01
BMI(Kg/m²)	36.81 ± 4.69	26.62 ± 2.92	< 0.01
TG(mmol/L)	1.98(1.64)	1.50(0.11)	< 0.01
HDL(mmol/L)	1.26(0.04)	1.40(0.10)	0.766
TG/HDL	1.50(1.31)	1.04(0.28)	< 0.01
WBC(10⁹/L)	8.95 ± 2.03	6.25 ± 1.85	< 0.01
NEU(10⁹/L)	5.38 ± 1.57	3.91 ± 0.71	< 0.01
LYM(10⁹/L)	2.70(1.02)	2.70(0.35)	< 0.01
NLR	1.86(0.95)	1.50(0.03)	< 0.01
CRP(mg/L)	5.64(7.07)	2.95(1.00)	0.138

In the contemporary context of the global obesity pandemic, accumulating evidence indicates that obesity contributes to various complications, including cardiovascular disease, insulin resistance, type 2 diabetes (T2DM), hypertension, and hyperlipidemia [12]. Bariatric surgery, known for its effectiveness in weight loss, mitigating T2DM, improving cardiovascular risk factors, alleviating other obesity-related comorbidities, and enhancing long-term survival, stands as a potent therapeutic intervention. The primary mechanisms through which weight-loss surgery impacts obesity and its complications encompass dietary behavior modification, alterations in intestinal hormone release, and changes in the microbiota [13]. The molecular intricacies underlying the effects of bariatric surgery remain an area of clinical exploration. This study delves into the gene expression data of subcutaneous adipose tissue in obese patients, employing bioinformatics analyses to gain insights into the potential mechanisms of weight loss and metabolic improvement, with the aim of contributing findings for clinical treatment.

Drawing from the GSE53376 and GSE59034 datasets, this study uncovers potential adipose tissue-related pathways both before and after bariatric surgery. Employing bioinformatics methodologies, we identify 29 downregulated differentially expressed genes. Gene Ontology (GO) enrichment analysis highlights that genes with reduced expression post-surgery are enriched in inflammatory response and cellular activation involved in the immune response, aligning with existing literature [14–16]. G protein-coupled receptors play a pivotal role in regulating various adipocyte functions, influencing processes such as lipolysis, heat production, glucose processing, and adipokine factor secretion. Their signalling significantly contributes to the regulation of adipose tissue metabolism, particularly in obesity, type 2 diabetes, and related comoridities [17–19]. G protein-coupled receptors are integral in maintaining energy homeostasis, influencing food intake, energy expenditure, glycemic control, and participating in cell division, growth, repair, protein synthesis, and immune responses [20].

Gene Set Enrichment Analysis (GSEA) outcomes support the notion that surgery-induced weight loss triggers a shift from a state of pro-inflammation to anti-inflammation, ultimately enhancing metabolic functions. GSEA reveals that, before bariatric surgery, most genes are predominantly enriched in chemokines, TOLL-like receptor, and B cell receptor signaling pathways—mechanisms integral to adipose tissue inflammation. Adipocytes release chemokines that intensify tissue inflammation, with specific chemokines like CCL2 and CCL5 exacerbating systemic inflammation by recruiting proinflammatory M1 macrophages, thereby contributing to obesity and its chronic complications [21]. Toll-like receptor signaling induces fatty acid synthesis, fostering chronic inflammation in adipose tissue and contributing to the metabolic syndrome associated with obesity [22]. B cells, through regulating T cell function and inflammatory factor secretion, can exacerbate the inflammatory response in obesity and T2DM [23].

GSEA provides additional support, revealing that the enrichment outcomes post-bariatric surgery are intricately linked with fatty acid and amino acid metabolism. Numerous obese individuals grapple with lipid metabolism disorders, notably abdominal obesity, elevating cardiometabolic risk due to atherogenic dyslipidemia [24]. Nutrient composition, particularly fatty acids, can significantly impact lipid levels. Adiponectin, exclusively expressed by adipocytes, exerts anti-inflammatory effects and enhances insulin sensitivity by promoting fatty acid oxidation. Most studies consistently report a substantial increase in adiponectin levels following laparoscopic sleeve gastrectomy [25]. A recent extensive meta-analysis demonstrated significant adiponectin increases at 6, 12, and 18 months post-weight loss [26]. Mounting evidence suggests that bariatric surgery contributes to enhanced production of anti-inflammatory adipokines and a concurrent reduction in proinflammatory adipokine production [27]. Branched-chain amino acid levels, including leucine, isoleucine, and valine, exhibit significant elevation in obese individuals with type 2 diabetes or insulin resistance [28]. Numerous studies affirm a decrease in the levels of most amino acids, including branched-chain amino acids, post-bariatric surgery. However, the levels of specific amino acids, such as glycine and serine, increase uniformly, irrespective of the surgical type [29]. Research underscores that metabolites like glucose, fatty acids, glutamine, and succinate, functioning as energy sources, can influence the activity of macrophages, neutrophils, and T cells. Obesity is characterized by a continuous supply of these metabolites. The observed metabolic shifts in nutrient supplies after bariatric surgery have the potential to modulate the immune system, playing a pivotal role in immune differentiation. Bariatric surgery, by curbing pathological immune activation and thereby diminishing the supply of nutrient metabolites, directly regulates inflammation. Consequently, it establishes an anti-inflammatory state, thereby enhancing metabolic function [30]. These mechanisms contribute to the comprehension of how bariatric surgery orchestrates a significant transition from inflammation to metabolism. TREM2, a membrane protein expressed on immune cells, is implicated in neurodegenerative diseases and cancer [31]. Metabolic symptoms, including adipocyte hypertrophy, hypercholesterolaemia, fat accumulation, and glucose intolerance, are observed in TREM2 knockout mice [32]. Exploring TREM2's role in metabolic diseases and obesity may unveil additional biological functions. MNDA governs bone marrow cell differentiation and myelodysplastic syndrome development [33]. It has been demonstrated that MNDA participates in inflammatory signaling pathways among proteins [34]. Subsequent studies indicate MNDA as an inflammatory gene with elevated expression in T1DM and hyperglycemia [35]. C5AR1 is closely linked to inflammation [36]. Adipose tissue reduction is noted in C5AR knockout mice irrespective of dietary changes [37]. Administering C5AR selective antagonists to diet-induced obese rats not only reduces body weight but also ameliorates insulin resistance and adipose tissue inflammation [38]. HP, a recognized inflammation marker [39], exhibits a positive correlation with body mass index in certain studies [40]. During early obesity phases, HP interacts with CCR2, facilitating monocyte entry into white adipose tissue. White adipose tissue macrophages enhance IL-6 and TNF α production, further inducing HP expression. Absence of HP diminishes adipose histiocytic inflammation, reducing obesity-related comorbidities. HP emerges as a potential therapeutic target [41]. S100A8 expedites macrophage migration, induces proinflammatory factor production, and triggers inflammatory responses in adipocytes. It is a pivotal component regulating the initial steps of the inflammatory cascade in obese adipose tissue. S100A8 antibodies effectively inhibit diet-induced inflammatory changes and play a crucial role in improving insulin resistance [42].

Collectively, these five key genes significantly contribute to obesity development. This study introduces potential therapeutic targets and biomarkers for obesity, offering a theoretical foundation for subsequent investigations.

Key gene expression data were extracted from the Attie Lab database to confirm the association of these genes with obesity. Results revealed higher expression of most key genes in the obese group.

The study presented in this paper possesses certain limitations. Firstly, the dissimilar follow-up durations of the two datasets (12 months for GSE53376 and 2 years for GSE59034) may have overlooked some biological information due to inadequate consideration of this temporal variation. Secondly, while the method employed by the Attie Lab diabetes database to confirm the elevated expression levels of key genes in obesity may not be optimal, it does provide partial confirmation of the correlation between key genes and obesity. However, the results lack more rigorous validation due to the absence of experimental verification. Thirdly, the data utilised in this study were sourced from public databases, and the data quality assessment was unfeasible. Additionally, the methods involving Affymetrix gene expression arrays, used in the data analysis of this study, have not been extensively employed. Lastly, the relatively small sample size of the dataset may impact the analysis of gene expression before and after bariatric surgery in obese patients.

In summary, this study concludes that bariatric surgery induces a substantial transformation in obesity, shifting it from a pro-inflammatory state to an anti-inflammatory state, thereby enhancing adipocyte metabolic function. Five key genes, identified as highly expressed in obesity, shed light on potential mechanisms underlying adipose tissue function changes post weight-reducing surgery. These findings offer novel therapeutic targets for ameliorating adipocyte inflammation, suggesting alternative approaches for the non-surgical treatment of obesity and associated comorbidities. The study provides a theoretical foundation for future investigations. The modified gastrointestinal anatomy following surgery serves as an ideal model for observing changes in the signal transmission chain from the gut environment to the brain, aiding in elucidating the brain's activity relationship with weight loss. Real-time studies are essential to further unravel the physiological mechanisms of bariatric surgery, with the overarching aim of expanding obesity prevention and treatment strategies.

1 Microarray Dataset

Three datasets pertaining to alterations in the subcutaneous adipose tissue (AT) transcriptome following ba retrieved from the Gene Expression Omnibus (GEO) website (Table 5). GSE53376 and GSE59034 were selected for primary analyses, while the GSE72158 dataset served for subsequent validation. The platform for GSE53376 utilised GPL16686[HuGene-2_0-st] Affymetrix Human Gene 2.0 ST Array [transcript (gene) version]. This dataset comprised 16 samples of subcutaneous AT obtained from obese subjects, alongside an additional 16 subcutaneous AT samples collected shortly after bariatric surgery. On the other hand, the platform for GSE59034 employed GPL11532 [HuGene-1_1-st] Affymetrix Human Gene 1.1 ST Array [transcript (gene) version]. This dataset included subcutaneous AT samples from obese subjects before (n = 16) and after short-term bariatric surgery (n = 16), as well as samples from individuals maintaining a never-obese status (n = 16). Lastly, the platform for GSE72158 was GPL10558[Illumina HumanHT-12 V4.0], encompassing subcutaneous AT samples from obese subjects before (n = 40) and after short-term bariatric surgery (n = 40).riatric surgery were

Table 5

Basic information of the dataset
dataset	year	tissue	sample	platform
GSE53376	2014	Subcutaneous adipose tissue	before surgery (n = 16) and after surgery (n = 16)	GPL16686[HuGene-2_0-st] Affymetrix Human Gene 2.0 ST Array [transcript (gene) version]
GSE59034	2017	Subcutaneous adipose tissue	before surgery (n = 16), after surgery (n = 16) and never obese people(n = 16)	GPL11532[hugene-11.1-st] Affymetrix Human Gene 1.1 ST Array [transcript (gene) version]
GSE72158	2015	Subcutaneous adipose tissue	before surgery (n = 40) and after surgery (n = 40)	GPL10558[Illumina HumanHT-12 V4.0]

2 Identification of Differentially Expressed Genes

The R packages "GEOquery" (V2.66.0) and "limma" (V3.54.0) within the interactive online tool GEO2R (https://www.ncbi.nlm.nih.gov/geo/geo2r) were utilised for data normalisation and screening of Differentially Expressed Genes (DEGs) (|fold change| ≥ 1 and adj. p < .05). The resultant DEGs were presented visually through Heatmaps. The Draw Venn Diagram tool (http://bioinformatics.psb.ugent.be/webtools/Venn/) was employed to identify common DEGs across the two datasets.

3 Functional and Pathway Enrichment Analysis

Gene Ontology (GO) analysis was applied to discern biological attributes [43], encompassing biological processes (BP), cellular components (CC), and molecular functions (MF). KEGG analysis provided more comprehensive information on genome sequences and protein interactive networks [44]. Enrichment analysis was conducted using the SangerBox platform[45], with gene set sizes ranging from 5 to 5,000, and a threshold of statistical significance set at p-Value < .05 and FDR < 0.25.

4 Gene Set Enrichment Analysis

The GSEA [46] software (4.3.2) was employed for further analysis of the genes identified in the GSE59034 dataset, offering a comprehensive screening of biological functions post-surgery. Each analysis included 1000 gene set permutations, and significance was established at a p-value < 5% and a false discovery rate (FDR) < 25%.

5 Construction of PPI Network and Identification of Key Genes

The STRING [47] database (http://www.string-db.org/) was utilised to construct a Protein-Protein Interaction (PPI) network of differentially expressed genes, identifying and predicting interactions between genes or proteins. The PPI network of differentially expressed genes was visualised using Cytoscape V3.9.1. Core nodes, indicative of greater impact in protein interaction networks, were determined using Cytohubba [48], employing 12 statistical analysis methods. The top 15 genes ranked by each algorithm were selected, filtered using the Upset package to identify key genes, and subjected to Pearson correlation analysis.

6 Immune Infiltration Analysis

This study employed xCell[49] analysis to detect the distribution of immune cells in the tissue based on gene expression profiles, predicting the relative abundance of cell types.

7 Validation of Key Genes

The Attie Lab Diabetes Database (http://diabetes.wisc.edu) facilitated a search for gene expression in six key tissues (islet, liver, adipose, hypothalamus, gastrocnemius, and soleus) in the mouse model of Type 2 diabetes, considering genetic obesity (lean vs. OB/OB), strain (B6 vs. BTBR), and age (4 vs. 10 weeks) (significance set at p < 0.05). This enabled verification of key gene expression in adipose tissue of both lean and obese mice at 4 and 10 weeks. Additionally, the dataset GSE72158 was utilised to validate the expression of key genes selected before and after surgery.

8 Assessment of the Diagnostic Effectiveness of Key Genes

To validate the diagnostic efficacy of the key genes in identifying obesity within the GSE59034 gene sets, ROC logistic regression analysis was conducted. The Area Under the Receiver Operating Characteristic Curve (ROC-AUC) served as the criterion for assessment.

9 Clinical Data to Verify Changes in Inflammatory Status Before and After Bariatric Surgery

Clinical data were collected both before and one year after bariatric surgery, encompassing Body Mass Index (BMI), neutrophil count (NEU), lymphocyte count (LYM), Neutrophil-to-Lymphocyte Ratio (NLR), triglycerides (TG), high-density lipoprotein (HDL), and the TG/HDL ratio. Observations focused on alterations in BMI and markers of inflammation (NLR, TG/HDL). SPSS 27.0 software (IBM SPSS Statistics 27 V27.0.1) facilitated data processing. A paired sample t-test was applied for normally distributed variables in the comparison of the two groups before and after surgery, while a paired sample Wilcoxon test was employed for non-normally distributed variables. A significance level of P < 0.05 denoted a statistically significant difference.

Acknowledgement

We are grateful to our team members for valuable discussion

Author contributions statement

H.H. and C.H. designed the research. C.H. conducted the search. Y.W. , Z.Z. and H.W. prepared figures. J.W. and Y.X. revised the article critically. W.W., Y.X. and H.Z. collected and analysed the clinical data. All the authors contributed to the manuscript writing, read and approved the final manuscript.

Competing interests

Each author declares that he or she has no commercial associations (e.g. consultancies, stock ownership, equity interest, patent/licensing arrangement etc.) that might pose a conflict of interest in connection with the submitted article.

Data availability

Any data and R script in this study can be obtained from the corresponding author upon reasonable request. The final manuscript was read and approved by all authors. In this study, publicly available datasets were analyzed. These are available on GEO (https://www.ncbi.nlm.nih.gov/). Attie Lab Database (http://diabetes.wisc.edu/) , STRING database (http://www.string-db.org/) and Sangerbox (http://sangerbox.com/home.html) are also freely available.

Ethical Statement

The study was approved by the Scientific research IRB of Wannan Medical College Yijishan Hospital. Written informed consent was obtained from all participants prior to their inclusion in the study. The authors affirm that participants' identities have been anonymized and that any potentially identifying information has been removed. The study adhered to the principles of ethical research, including the protection of participants' rights and welfare. The Ethics Approval Number is (2023)伦审研第(148).

Roberto, C. A. et al. Patchy progress on obesity prevention: emerging examples, entrenched barriers, and new thinking. Lancet (London, England), 385(9985), 2400–2409. https://doi.org/10.1016/S0140-6736(14)61744-X (2015).
Powell-Wiley, T. M. et al. American Heart Association Council on Lifestyle and Cardiometabolic Health; Council on Cardiovascular and Stroke Nursing; Council on Clinical Cardiology; Council on Epidemiology and Prevention; and Stroke Council. Obesity and Cardiovascular Disease: A Scientific Statement From the American Heart Association. Circulation, 143(21), e984–e1010. https://doi.org/10.1161/CIR.0000000000000973 (2021).
Boutari, C., & Mantzoros, C. S. A 2022 update on the epidemiology of obesity and a call to action: as its twin COVID-19 pandemic appears to be receding, the obesity and dysmetabolism pandemic continues to rage on. Metabolism: clinical and experimental, 133, 155217. https://doi.org/10.1016/j.metabol.2022.155217 (2022).
Cypess A. M. Reassessing Human Adipose Tissue. The New England journal of medicine, 386(8), 768–779. https://doi.org/10.1056/NEJMra2032804 (2022).
Marcelin, G., Gautier, E. L., & Clément, K. Adipose Tissue Fibrosis in Obesity: Etiology and Challenges. Annual review of physiology, 84, 135–155. https://doi.org/10.1146/annurev-physiol-060721-092930 (2022).
Angelidi, A. M. et al. Novel Noninvasive Approaches to the Treatment of Obesity: From Pharmacotherapy to Gene Therapy. Endocrine reviews, 43(3), 507–557. https://doi.org/10.1210/endrev/bnab034 (2022).
Ohta, M. et al. Bariatric/Metabolic Surgery in the Asia-Pacific Region: APMBSS 2018 Survey. Obesity surgery, 29(2), 534–541. https://doi.org/10.1007/s11695-018-3539-7 (2019).
Mingrone, G. et al. Metabolic surgery versus conventional medical therapy in patients with type 2 diabetes: 10-year follow-up of an open-label, single-centre, randomised controlled trial. Lancet (London, England), 397(10271), 293–304. https://doi.org/10.1016/S0140-6736(20)32649-0 (2021).
Akalestou, E. et al. Mechanisms of Weight Loss After Obesity Surgery. Endocrine reviews, 43(1), 19–34. https://doi.org/10.1210/endrev/bnab022 (2022).
Wrzosek, M. et al. Impact of Fatty Acids on Obesity-Associated Diseases and Radical Weight Reduction. Obesity surgery, 32(2), 428–440. https://doi.org/10.1007/s11695-021-05789-w (2022).
Chen, S. et al. Bioinformatics analysis to obtain critical genes regulated in subcutaneous adipose tissue after bariatric surgery. Adipocyte, 11(1), 550–561. https://doi.org/10.1080/21623945.2022.2115212 (2022).
Wu, H., & Ballantyne, C. M. Metabolic Inflammation and Insulin Resistance in Obesity. Circulation research, 126(11), 1549–1564. https://doi.org/10.1161/CIRCRESAHA.119.315896 (2020).
Akalestou, E. et al. Mechanisms of Weight Loss After Obesity Surgery. Endocrine reviews, 43(1), 19–34. https://doi.org/10.1210/endrev/bnab022 (2022).
Schmidt, V. et al. Obesity-Mediated Immune Modulation: One Step Forward, (Th)2 Steps Back. Frontiers in immunology, 13, 932893. https://doi.org/10.3389/fimmu.2022.932893 (2022).
Artemniak-Wojtowicz, D., Kucharska, A. M., & Pyrżak, B. Obesity and chronic inflammation crosslinking. Central-European journal of immunology, 45(4), 461–468. https://doi.org/10.5114/ceji.2020.103418 (2020).
Taylor E. B. The complex role of adipokines in obesity, inflammation, and autoimmunity. Clinical science (London, England : 1979), 135(6), 731–752. https://doi.org/10.1042/CS20200895 (2021).
Wang, L. et al. Adipocyte G signaling is essential for maintaining whole-body glucose homeostasis and insulin sensitivity. Nature communications, 11(1), 2995. https://doi.org/10.1038/s41467-020-16756-x_i (2020).
Ceddia, R. P., & Collins, S. A compendium of G-protein-coupled receptors and cyclic nucleotide regulation of adipose tissue metabolism and energy expenditure. Clinical science (London, England : 1979), 134(5), 473–512. https://doi.org/10.1042/CS20190579 (2020).
London, E., & Stratakis, C. A. The regulation of PKA signaling in obesity and in the maintenance of metabolic health. Pharmacology & therapeutics, 237, 108113. https://doi.org/10.1016/j.pharmthera.2022.108113 (2022).
Jamaluddin, A., & Gorvin, C. M. RISING STARS: Targeting G protein-coupled receptors to regulate energy homeostasis. Journal of molecular endocrinology, 70(4), e230014. https://doi.org/10.1530/JME-23-0014 (2023).
Chan, P. C., & Hsieh, P. S. The Chemokine Systems at the Crossroads of Inflammation and Energy Metabolism in the Development of Obesity. International journal of molecular sciences, 22(24), 13528. https://doi.org/10.3390/ijms222413528 (2021).
Li, B. et al. A global perspective on the crosstalk between saturated fatty acids and Toll-like receptor 4 in the etiology of inflammation and insulin resistance. Progress in lipid research, 77, 101020. https://doi.org/10.1016/j.plipres.2019.101020 (2020).
DeFuria, J., Belkina, A. C., Jagannathan-Bogdan, M., Snyder-Cappione, J., Carr, J. D., Nersesova, Y. R., Markham, D. et al. B cells promote inflammation in obesity and type 2 diabetes through regulation of T-cell function and an inflammatory cytokine profile. Proceedings of the National Academy of Sciences of the United States of America, 110(13), 5133–5138. https://doi.org/10.1073/pnas.1215840110 (2013).
Nussbaumerova, B., & Rosolova, H. Obesity and Dyslipidemia. Current atherosclerosis reports, 25(12), 947–955. https://doi.org/10.1007/s11883-023-01167-2 (2023).
Liu, L. et al. Adiponectin/SIRT1 Axis Induces White Adipose Browning After Vertical Sleeve Gastrectomy of Obese Rats with Type 2 Diabetes. Obesity surgery, 30(4), 1392–1403. https://doi.org/10.1007/s11695-019-04295-4 (2020).
Huang, J. et al. Change in Adipokines and Gastrointestinal Hormones After Bariatric Surgery: a Meta-analysis. Obesity surgery, 33(3), 789–806. https://doi.org/10.1007/s11695-022-06444-8 (2023).
Cibičková, Ľ. Et al. Effect of Laparoscopic Sleeve Gastrectomy on Serum Adipokine Levels. Physiological research, 72(Suppl 2), S165–S172. https://doi.org/10.33549/physiolres.935053 (2023).
Yoneshiro, T. et al. BCAA catabolism in brown fat controls energy homeostasis through SLC25A44. Nature, 572(7771), 614–619. https://doi.org/10.1038/s41586-019-1503-x (2019).
Cuomo, P. et al. Role of Branched-Chain Amino Acid Metabolism in Type 2 Diabetes, Obesity, Cardiovascular Disease and Non-Alcoholic Fatty Liver Disease. International journal of molecular sciences, 23(8), 4325. https://doi.org/10.3390/ijms23084325 (2022).
Villarreal-Calderón, J. R. et al. Interplay between the Adaptive Immune System and Insulin Resistance in Weight Loss Induced by Bariatric Surgery. Oxidative medicine and cellular longevity, 2019, 3940739. https://doi.org/10.1155/2019/3940739 (2019).
Reich, T. et al. TREM2 has a significant, gender-specific, effect on human obesity. Scientific reports, 13(1), 482. https://doi.org/10.1038/s41598-022-27272-x (2023).
Jaitin, D. A. et al. Lipid-Associated Macrophages Control Metabolic Homeostasis in a Trem2-Dependent Manner. Cell, 178(3), 686–698.e14. https://doi.org/10.1016/j.cell.2019.05.054 (2019).
Briggs, R. C. et al. Dysregulated human myeloid nuclear differentiation antigen expression in myelodysplastic syndromes: evidence for a role in apoptosis. Cancer research, 66(9), 4645–4651. https://doi.org/10.1158/0008-5472.CAN-06-0229 (2006).
Fairbrother, W. J. et al. The PYRIN domain: a member of the death domain-fold superfamily. Protein science : a publication of the Protein Society, 10(9), 1911–1918. https://doi.org/10.1110/ps.13801 (2001).
Friedrichs, P. et al. Hyperglycaemic memory affects the neurovascular unit of the retina in a diabetic mouse model. Diabetologia, 60(7), 1354–1358. https://doi.org/10.1007/s00125-017-4254-y (2017).
Reis, E. S. et al. New insights into the immune functions of complement. Nature reviews. Immunology, 19(8), 503–516. https://doi.org/10.1038/s41577-019-0168-x (2019).
Roy, C. et al. C5a receptor deficiency alters energy utilization and fat storage. PloS one, 8(5), e62531. https://doi.org/10.1371/journal.pone.0062531 (2013).
Lim, J. et al. C5aR and C3aR antagonists each inhibit diet-induced obesity, metabolic dysfunction, and adipocyte and macrophage signaling. FASEB journal : official publication of the Federation of American Societies for Experimental Biology, 27(2), 822–831. https://doi.org/10.1096/fj.12-220582 (2013).
Schmidt, M. I. et al. Markers of inflammation and prediction of diabetes mellitus in adults (Atherosclerosis Risk in Communities study): a cohort study. Lancet (London, England), 353(9165), 1649–1652. https://doi.org/10.1016/s0140-6736(99)01046-6 (1999).
Chiellini, C. et al. Serum haptoglobin: a novel marker of adiposity in humans. The Journal of clinical endocrinology and metabolism, 89(6), 2678–2683. https://doi.org/10.1210/jc.2003-031965 (2004).
Maffei, M. et al. The Multifaceted Haptoglobin in the Context of Adipose Tissue and Metabolism. Endocrine reviews, 37(4), 403–416. https://doi.org/10.1210/er.2016-1009 (2016).
Sekimoto, R. et al. Visualized macrophage dynamics and significance of S100A8 in obese fat. Proceedings of the National Academy of Sciences of the United States of America, 112(16), E2058–E2066. https://doi.org/10.1073/pnas.1409480112 (2015).
Gene Ontology Consortium, Aleksander, S. A. et al. The Gene Ontology knowledgebase in 2023. Genetics, 224(1), iyad031. https://doi.org/10.1093/genetics/iyad031 (2023).
Ogata, H. et al. KEGG: Kyoto Encyclopedia of Genes and Genomes. Nucleic acids research, 27(1), 29–34. https://doi.org/10.1093/nar/27.1.29 (1999).
Shen, Weitao et al. “ Sangerbox: A Comprehensive, Interaction-Friendly Clinical Bioinformatics Analysis Platform.” iMeta 1, e36. https://doi.org/10.1002/imt2.36 (2022).
Subramanian, A. et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proceedings of the National Academy of Sciences of the United States of America, 102(43), 15545–15550. https://doi.org/10.1073/pnas.0506580102 (2005).
Szklarczyk, D. et al. The STRING database in 2021: customizable protein-protein networks, and functional characterization of user-uploaded gene/measurement sets. Nucleic acids research, 49(D1), D605–D612. https://doi.org/10.1093/nar/gkaa1074 (2021).
Chin, C. H. et al. cytoHubba: identifying hub objects and sub-networks from complex interactome. BMC systems biology, 8 Suppl 4(Suppl 4), S11. https://doi.org/10.1186/1752-0509-8-S4-S11 (2014).
Aran, D. et al. xCell: digitally portraying the tissue cellular heterogeneity landscape. Genome biology, 18(1), 220. https://doi.org/10.1186/s13059-017-1349-1 (2017).

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Identification of the critical genes and signaling pathways in subcutaneous adipose tissue after bariatric surgery based on the GEO database

Status:

Version 1

Abstract

Figures

Introduction

Results

1 Identification of Differentially Expressed Genes (DEGs) after Bariatric Surgery

2 Gene Ontology and KEGG Pathway Enrichment Analyses after Bariatric Surgery

3 Gene Set Enrichment Analysis (GSEA)

4 Protein-Protein Interaction (PPI) Network Analysis and Key Gene Selection

5 Cell Type Enrichment Analysis

6 Preliminary Validation of Key Genes

7 Diagnostic Efficacy of Key Genes for Obesity

8 Clinical Data Validate Changes in Patient Inflammatory Status Before and After Bariatric Surgery

8.1 Basic Information

8.2 Inclusion and Exclusion Criteria

8.3 Data Collection

Discussion

Methods

1 Microarray Dataset

2 Identification of Differentially Expressed Genes

3 Functional and Pathway Enrichment Analysis

4 Gene Set Enrichment Analysis

5 Construction of PPI Network and Identification of Key Genes

6 Immune Infiltration Analysis

7 Validation of Key Genes

8 Assessment of the Diagnostic Effectiveness of Key Genes

9 Clinical Data to Verify Changes in Inflammatory Status Before and After Bariatric Surgery

Declarations

Acknowledgement

Author contributions statement

Competing interests

Data availability

Ethical Statement

References

Additional Declarations

Status:

Version 1