Identifying of immune‐associated genes for assessing the obesity‐associated risk to the offspring in maternal obesity: A bioinformatics and machine learning

Yanxing Shang, Xueqin Wang, Sixuan Su, Feng Ji, Donghai Shao, Chengwei Duan, Tianpeng Chen, Caixia Liang, Dongmei Zhang, Hongjian LuCNS 2024 · DOI

TL;DR

Maternal obesity significantly increases the risk of metabolic diseases in offspring. This study uses bioinformatics and machine learning to identify two key immune-associated genes, Sytl4 and Kcnc2, that can predict this risk. A diagnostic nomogram based on these genes shows high accuracy, offering a potential tool for early identification and intervention.

What Problem Does This Paper Address?

Obesity is a growing global health crisis, and emerging evidence points to a critical role for maternal health during pregnancy in shaping offspring's long-term metabolic health. Specifically, maternal obesity can "program" offspring for increased susceptibility to obesity and related metabolic disorders later in life. The hypothalamus, a key brain region regulating energy balance, is implicated in this process, with evidence suggesting that immune dysregulation within the hypothalamus plays a crucial role. However, the specific molecular mechanisms and candidate genes involved in this maternal obesity-induced risk transmission to offspring remain unclear. This paper aims to identify these key immune-associated genes and develop a predictive tool for offspring obesity risk stemming from maternal obesity.

Methods

The researchers employed a multi-pronged bioinformatics approach. They started by identifying differentially expressed genes (DEGs) in the hypothalamus of mice exposed to maternal obesity using the Limma package. To uncover gene networks and modules associated with maternal obesity, they utilized Weighted Gene Co-expression Network Analysis (WGCNA). Key genes from these modules and DEGs were then intersected to find common candidates. Functional enrichment analysis (GO and KEGG) was performed to understand the biological roles of these genes. Crucially, they applied a machine learning algorithm, Least Absolute Shrinkage and Selection Operator (LASSO) regression, to pinpoint the most significant "hub" genes for predicting offspring obesity risk. Finally, a nomogram was constructed using these hub genes, and its diagnostic accuracy was evaluated using Receiver Operating Characteristic (ROC) curves. Immune cell infiltration analysis was also conducted to explore the role of immune cells in the hypothalamus.

Datasets

The study utilized two publicly available gene expression datasets from the Gene Expression Omnibus (GEO) database:

  • GSE135830: Gene expression data from the hypothalamus of mouse offspring from mothers fed either a high-fat diet (mHFD) or a control diet (mChow).
  • GSE127056: Hypothalamus microarray data from young adult mice with diet-induced obesity (HFD) versus control (Chow) groups.

Key Insights and Ideas

  • Immune Dysregulation is Key: The study strongly suggests that immune system dysregulation in the offspring's hypothalamus is a central mechanism linking maternal obesity to offspring metabolic risk. This goes beyond just the direct effects on neuronal development.
  • Two Genes Tell a Story: Sytl4 (Synaptotagmin-like 4) and Kcnc2 (Potassium voltage-gated channel Shaw-related subfamily member 2) emerged as the most promising candidate hub genes. Their increased expression in models of hypothalamic inflammation and obesity highlights their potential role.
  • A Powerful Predictive Tool: The nomogram constructed from Sytl4 and Kcnc2 demonstrated near-perfect diagnostic accuracy (AUC of 1.000), suggesting its potential as a reliable biomarker for identifying offspring at risk.
  • Immune Cell Infiltration Patterns Differ: The study observed distinct patterns of immune cell infiltration in the hypothalamus of offspring exposed to maternal obesity, with an increase in certain immune cells like naive B cells and macrophages, and a decrease in others like CD8 T cells. This suggests a complex immune response is triggered.

Key Results

  • DEGs and Modules: Identified 2868 DEGs in offspring of maternal obesity and 2627 key module genes associated with maternal obesity.
  • Common Genes: Found 13 common genes between adult obesity DEGs and maternal obesity module genes, enriched in neurodevelopment and immune response pathways.
  • Candidate Hub Genes: LASSO regression identified Sytl4, Kcnc2, and Rnf213 as potential biomarkers.
  • Diagnostic Power: Sytl4 and Kcnc2 showed high diagnostic value (AUCs of 0.967 and 1.000, respectively).
  • Nomogram Performance: The nomogram combining Sytl4 and Kcnc2 achieved a perfect AUC of 1.000, indicating excellent diagnostic specificity and sensitivity.
  • In Vivo/In Vitro Validation: Expression of Sytl4 and Kcnc2 was validated in HFD-induced hypothalamic inflammation mouse models and palmitic acid-stimulated microglial models, confirming their upregulation in inflammatory conditions.

Limitations and Future Work

The authors acknowledge several limitations:

  • Small Sample Size: The number of hypothalamic samples used for analysis was low, which may have contributed to the high diagnostic value of the nomogram. They emphasize the need for larger-scale studies to confirm these findings.
  • Unelucidated Mechanisms: While Sytl4 and Kcnc2 were identified as hub genes, their specific functions and underlying molecular mechanisms in the context of maternal obesity-induced offspring metabolic disorders require further investigation.

Future work should focus on validating these findings in larger human cohorts and conducting detailed mechanistic studies to understand how Sytl4 and Kcnc2 contribute to hypothalamic programming and inflammation, paving the way for targeted therapeutic interventions.

Summary updated

Abstract

Two candidate hub genes (Sytl4 and Kcnc2) were identified and a nomogram was developed to predict obesity risk in offspring with maternal obesity. These findings offer potential diagnostic candidate genes for identifying obesity-associated risks in the offspring of obese mothers.
BibTeX
@inproceedings{shang2024identifying,
  author = {Yanxing Shang and Xueqin Wang and Sixuan Su and Feng Ji and Donghai Shao and Chengwei Duan and Tianpeng Chen and Caixia Liang and Dongmei Zhang and Hongjian Lu},
  title = {{Identifying of immune‐associated genes for assessing the obesity‐associated risk to the offspring in maternal obesity: A bioinformatics and machine learning}},
  booktitle = {CNS},
  year = {2024},
  doi = {10.1111/cns.14700},
}
Figure