Ligand-Receptor Analysis of Brain Cell Type Marker Data Reveals Intricate Endothelial Interaction

Abstract

Background: Brain endothelial interaction with neurons, astrocytes, oligodendrocytes and microglial cells is critical for brain physiology; it is still far from being mapped. Understanding of the endothelial communication with other brain cell type could unravel novel insight into neurovascular homeostasis.

Purpose: This study aims to construct neurovascular interaction network, focusing on brain endothelial cell interactome using brain cell marker gene dataset and ligand-receptor (LR) pair.

Methods: We curated brain marker gene list from McKenzie et al.'s brain cell type top 1000 marker list of endothelial, microglia, astrocyte, neuron, oligodendrocyte and oligodendrocyte progenitor cell (OPC) and extracted LR interaction between them. Subsequently, using Cytoscape, endothelial cell interaction map was constructed and top interaction and hub gene were derived. Moreover, we performed Kyoto encyclopedia of genes and genomes (KEGG) pathways enrichment (p value < .1) to infer biological information hidden.

Results: Neurovascular LR interaction showed endothelial cells as the top network having 25.34% of total interaction with 176 outgoing and 171 incoming interactions. A considerable portion of signalling (11%) is involved in autocrine signalling functionally related to vascular tone, angiogenesis and others. Paracrine signalling between endothelial cells with microglia, astrocytes, neurons and OPC constituted 13.5%, 8.9%, 5.8% and 4.9% of total interactions, respectively. Functional enrichment of LR interaction in endothelial-microglia, endothelial-astrocyte and endothelial-neuron networks constitutes 49, 45 and 36 significant KEGG pathways (p value < .1) respectively. These pathways include extracellular matrix (ECM) receptor, axon guidance, chemokine, nuclear factor kappa B (NF-kB) and signalling pathways, among others. Hub gene analysis showed ITGB1 in endothelial cells, ITGA4 in microglia, NOTCH2 in astrocytes and LAMC2 in neurons having maximum interaction in the endothelial network.

Conclusion: This study recapitulated not only previously known gene interactions using a markers gene list but also identified novel interactions between endothelial and other brain cell types. In conclusion, this analysis underscores the critical role of endothelial cell interactions in brain physiology.

Keywords: Function; genetics; neurology; physiology.

Figure 1.. The Total Number of Ligands, Receptors and Ligand–receptor Pairings Identified in the Top 1000 Dataset. Cells Expressing the Ligands Are Represented in Rows, and Cells Expressing the Receptors Are Represented in Columns. The Network-graph View Indicates the Number of Ligand–receptor Pairs Between and Within Each Cell Type. The Colour of the Edge Is Coded According to the Number of Interactions, from Green at the Lowest to Red at the Highest. The Arrows at the End of the Edge Show the Ligand–receptor direction. Circos View for the Same Network in a Circular Fashion. The Percentage of Each Cell Type’s Interaction Is Derived from the Total Ligand–receptor Interaction.

Figure 2.. The Putative Paracrine and Autocrine Ligand–receptor Signalling Networks of Endothelial Cells with Microglia, Astrocytes, Neurons, Oligodendrocytes and OPC. Network Nodes Indicate Ligand or Receptor Pairs, Differentially Coloured for Each Cell Type. Arrows Show Signals from Ligands to Receptors. Only Ligand–receptor Interactions Equal to or Greater Than 0.5 Weight Are Used for Interaction Analysis.

Figure 3.. The Bar Graph Shows the KEGG Pathways That Were Significantly Enriched After Multiple Testing Adjustments (p < .1) in the Identified Ligand–receptor Pair of (A) Endothelial–endothelial and (B) endothelial–microglia. (C) Endothelial–astrocytes, (D) neuronal endothelial–endothelial oligo, (E) endothelial OPC, (F) endothelial OPC, (F) endothelial OPC and (F) endothelial. The Y-axis Indicates the Total Number of Ligands or Receptors Modulated in Each Pathway in Each Group. For Pathway Analysis, Only Ligand–receptor Interactions Equal or Greater Than 0.5 Weight Are Used.

Figure 4.. NeuroExpresso Confirms Marker Gene Expression Patterns of Ligand–receptor Obtained from Top Weighted Ligand–receptor Interaction in (A) Endothelial Ligand–microglia Receptor, (B) Microglia Ligand–endothelial Receptor, (C) Endothelial Ligand–astrocytes Receptor, (D) Astrocytes Ligand–endothelial Receptor, (E) Endothelial Ligand–neuron receptor, (F) Neuron Ligand–endothelial Receptor from the Cortex of Each Points Indicates Expression in the Microarray or RNA seq Dataset Curated in This Software.

Figure 5.. Endothelial Autocrine and Paracrine Ligand–receptor Interactions Within the Neurovascular Unit. Top Ligand–receptors Interaction Having High Communication Scores (2.3–24.2) Is Shown in Each Interaction. E. CELL: Endothelial Cell, OPC: Oligodendrocyte Progenitor Cell.