Integrated bioinformatic analysis and experimental validation for exploring the key molecular of brain inflammaging

Zhixin Du, Yaohui Wang, Liping Yang, Tong Zhang, Yu Jiang, Zhenqiang ZhangFrontiers in Immunology 2023 · DOI

TL;DR

This study integrates bioinformatics and experimental validation to uncover key molecular mechanisms driving "inflammaging" in the aging brain. They identified CX3CL1 as a critical biomarker that, by modulating macrophage polarization, could be a therapeutic target to delay brain aging and immunosenescence.

What Problem Does This Paper Address?

Aging is a complex process characterized by a gradual decline in physiological integrity and function. In the brain, this aging process is exacerbated by "inflammaging"—a chronic, low-grade inflammatory state. This inflammaging disrupts the delicate balance of pro- and anti-inflammatory factors, making the aging brain more susceptible to disease and cognitive decline. While the general concept of inflammaging is recognized, the specific molecular players and pathways driving this process in different brain regions, and how they can be targeted, remain areas requiring deeper investigation. This paper aims to identify these key molecular drivers and potential therapeutic targets for brain aging.

Methods

The researchers employed a multi-pronged approach:

  • Bioinformatic Analysis:
    • Differential Gene Expression (DEG) Analysis: Used the "limma" package to identify genes with significantly different expression levels between aged and young brain samples across four regions (hippocampus, entorhinal cortex, superior frontal gyrus, postcentral gyrus).
  • Functional Enrichment Analysis: Utilized Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways to understand the biological processes and molecular functions associated with the identified DEGs.
  • Drug Prediction: Employed the Drug Gene Interaction Database (DGIdb) to identify potential small-molecule drugs targeting inflammatory response-related DEGs.
  • Machine Learning for Biomarker Identification: Used Least Absolute Shrinkage and Selection Operator (LASSO) and Random Forest (RF) algorithms to screen for diagnostic biomarkers.
  • Artificial Neural Network (ANN) Construction: Developed an ANN model for diagnosing aged individuals based on the identified biomarkers.
  • Single-Gene Gene Set Variation Analysis (GSVA): Assessed the enrichment of KEGG pathways based on the expression levels of individual biomarkers.
  • Consensus Clustering: Categorized aged samples into distinct molecular subtypes to identify differences in aging patterns.
  • Weighted Gene Co-expression Network Analysis (WGCNA): Identified modules of co-expressed genes and their correlation with aging and specific brain regions.
  • Immuno-infiltration Analysis: Assessed the composition of immune cells within the brain tissue.
  • Single-cell RNA sequencing Analysis: Utilized the AgeAnno tool to characterize the roles of biomarkers at the single-cell level and across different cell types.
  • Experimental Validation:
    • RT-qPCR: Quantified the expression levels of key genes (ADM, APLNR, C3AR1, CX3CL1) in animal models to validate bioinformatic findings.
  • Animal Experiments: Used male Wister rats (young and aged) to experimentally validate gene expression changes.

Datasets

  • GSE11882: A dataset from the Gene Expression Omnibus (GEO) database, comprising 173 postmortem human brain tissue samples from four regions (hippocampus, entorhinal cortex, superior frontal gyrus, postcentral gyrus), divided into young (20-59 years) and aged (60-99 years) groups. This was the primary dataset for analysis.
  • GSE1572: A validation dataset from GEO, including 22 aged and 24 young samples, used to test the performance of the developed ANN diagnostic model.
  • MSigDB (Molecular Signatures Database): Used as a source for inflammatory response-related genes.

Key Insights and Ideas

  • Inflammaging as a Central Driver: The study strongly supports the notion that inflammaging is a fundamental mechanism underlying brain aging, with DEGs consistently enriched in inflammatory response pathways across multiple brain regions.
  • CX3CL1 as a Key Regulator and Biomarker: CX3CL1 emerged as a critical player. Its expression declines with age, particularly in excitatory neurons, and its dysregulation is linked to increased inflammatory responses. This makes it a promising biomarker for brain aging.
  • Macrophage Polarization and CX3CL1: CX3CL1 appears to modulate macrophage polarization. Its reduced expression in aging might lead to a shift towards pro-inflammatory M1 macrophages, exacerbating inflammaging.
  • Therapeutic Potential of Targeting CX3CL1: The findings suggest that interventions targeting CX3CL1 could potentially mitigate inflammaging and delay brain aging and immunosenescence.
  • Multi-Regional Brain Aging: The study highlights that aging processes and their molecular underpinnings can vary across different brain regions, necessitating region-specific analyses.

Key Results

  • 126 Differentially Expressed Genes (DEGs): Identified between aged and young brain samples, with significant enrichment in inflammatory response, cytokine production, and immune regulation pathways.
  • 12 Inflammatory Response-Related DEGs: Further refined from the initial DEGs, these genes were central to the study's focus.
  • Four Robust Diagnostic Biomarkers: ADM, APLNR, C3AR1, and CX3CL1 were identified through LASSO and RF analysis and validated as robust biomarkers for distinguishing aged from young individuals.
  • ANN Diagnostic Model Performance: Achieved an AUC of 0.865 on the GSE11882 dataset and 0.696 on the GSE1572 validation set, demonstrating good diagnostic capability.
  • CX3CL1 Downregulation in Aged Brains: Experimental validation (RT-PCR) confirmed significantly lower CX3CL1 mRNA expression in aged rat hippocampal tissue compared to young rats.
  • CX3CL1 Decline in Specific Cell Types: Single-cell analysis revealed a significant decrease in CX3CL1 expression in inhibitory neurons, excitatory neurons, and astrocytes in aged brains, with the most pronounced reduction in excitatory neurons.
  • Macrophage Increase in Aged Brains: Immuno-infiltration analysis showed a significant increase in macrophages in aged individuals.

Limitations and Future Work

The authors acknowledge several limitations and suggest future directions:

  • Sample Size: Increasing the sample size for brain aging evaluation and prediction could improve accuracy.
  • In Vitro Validation: Further validation of the identified biomarkers using in vitro cell lines is recommended to provide practical evidence for clinical applications.
  • In Vivo and In Vitro Experiments: Conducting experiments with animal models that knock down or overexpress CX3CL1, or performing in vitro molecular experiments, would provide substantial evidence for the proposed mechanisms.
  • Further Exploration of CX3CL1: While progress has been made, more research is needed to fully understand the relationship between brain inflammaging and CX3CL1 levels.
  • Macrophage Polarization and Inflammaging: Further investigation into the correlation between macrophage polarization and brain inflammaging is warranted.
  • Longevity Populations: Exploring the molecular characteristics of longevity populations could offer valuable insights into biological processes and mechanisms associated with extended healthspan.

Summary updated

Abstract

In summary, targeting CX3CL1 can potentially delay inflammaging and immunosenescence in the brain.
BibTeX
@inproceedings{du2023integrated,
  author = {Zhixin Du and Yaohui Wang and Liping Yang and Tong Zhang and Yu Jiang and Zhenqiang Zhang},
  title = {{Integrated bioinformatic analysis and experimental validation for exploring the key molecular of brain inflammaging}},
  booktitle = {Frontiers in Immunology},
  year = {2023},
  doi = {10.3389/fimmu.2023.1213351},
}
Figure