Charting new territory: the molecular landscape of vascular cells in the human brain

Abstract

A study using advanced single-cell technologies has broadened our understanding of the diversity and complexity of brain endothelial cells by uncovering new endothelial subtypes and transcriptional patterns. These findings offer insights into potential therapeutic targets and emphasize the need for further research on vascular lineages and neurovascular interactions.

Situated between the blood and the neural parenchyma, the brain vasculature serves as the brain’s indispensable guardian and catalyst. These vascular cells, including endothelial and mural cells, simultaneously maintain the blood brain barrier, rejuvenate stem cell niches and facilitate differentiation, and coordinate neural signals to metabolic demand in neurovascular coupling. Accordingly, vascular cells are increasingly recognized to influence brain function both in homeostasis and disease. Early studies harnessing single-cell technologies in the human brain understandably focused on neural cells because of their overwhelming abundance. These technologically pioneering studies paved the way for the molecular and cellular characterization of the brain vascular network^1–3. A new study⁴ now broadens our knowledge on human brain endothelial biology by profiling these cells from development to adulthood and in pathological conditions including meningioma, glioblastoma, arteriovenous malformation and brain metastasis.

Building on the classical arteriovenous axis that was first molecularly characterized in the adult mouse brain⁵, Wälchli and colleagues discovered 14 major endothelial cell clusters, thereby demonstrating a tremendous diversity of brain endothelial cells and unprecedented transcriptional resolution of arteriovenous zonation. The investigators show that endothelial cells from glioblastoma and arteriovenous malformation can recapitulate gene programmes from prenatal endothelial cells. In addition, the pathological samples contained endothelial cell subtypes that transcriptionally resembled early developmental stages, including actively proliferating cells. Furthermore, the detailed characterization of endothelial subtypes in this study highlights new cell states with major potential functional implications. Brain vascular samples from individuals with brain disease had an abundance of cells that the researchers identify as endothelial cells from ‘angiogenic capillaries’ (Fig. 1). This aberrant endothelial state, which is broader and different to normal development, was overactivated in pathological samples and was repeatedly associated in their analyses as displaying the most cell–cell communication, ‘central nervous system signature’ and ‘blood–brain barrier dysfunction’ modules. In general, capillaries had the highest transcriptional heterogeneity when compared by species, tissue specificity and blood–brain barrier transcriptome⁴. Although RNA expression is only the first step to defining cellular function, pinpointing the specific expression patterns in each vascular subtype offers potential therapeutic insights. Future studies can build on this foundational work by manipulating these expression patterns in different experimental settings, including in in vivo models and a growing array of in vitro human vascular mimetics. Although research into brain vascular cell biology still has a distance to travel to achieve the same level of knowledge as our current understanding of the functional, molecular and cellular diversity of neuronal cell types, the depth and breadth of endothelial cell types described in this study⁴ provides a solid foundation.

Fig. 1. Endothelial cell heterogeneity in the human brain.

The arteriovenous axis in the brain contains several subtypes of endothelial cells, including arterial, venous and capillary endothelial cells, endothelial cells of angiogenic capillaries, and endothelial-to-mesenchymal transition (EndoMT) clusters. SMC, smooth muscle cell.

As with any impactful publication, this work prompts new questions. The spatial organization of vascular cells, their subtypes and the intriguing possibility of regional specification in the neurovascular unit remain incompletely understood. In the human cortex, the mechanisms of neural arealization include strong, mutually exclusive frontal and occipital gene expression signatures and continue to inspire new hypotheses. Owing to surgical availability, Wälchli et al. obtained all control adult human brain samples from the temporal lobe. A greater diversity of both cortical and non-cortical brain region representation would be a welcome addition in the future. To investigate different types of cellular interactions, Wälchli et al. performed ligand–receptor interaction maps, both within endothelial cells and among the varied cell types of the neurovascular unit. Given the spatial complexity of the arteriovenous axis and the use of dissociated cells in their analysis, the in vivo validity of the reported cellular interactions remains uncertain. The researchers used spatial transcriptomics to validate some of their results, a logical and reasonable next step after the extensive dissociated single-cell experiments in their study. However, vascular cells have remained under-represented in brain spatial transcriptomic studies so far. As new discoveries emerge about the diverse and important functional capacities of the brain vasculature, the possibility that unique neuronal subtypes have equally distinct vascular partners arises. The research field eagerly awaits a more complete picture of vascular arealization.

Classic mouse developmental studies have shown that vascular cells dive into the growing brain parenchyma⁶. Notably, the specific identify of these vascular progenitors remains unclear in any species. In the human brain, different bioinformatic pipelines have been used to infer cell lineage relationships, but these algorithms have caveats. Wälchli et al. also employ different bioinformatic strategies in their uniform manifold approximation and projection analyses of human brain vascular cells in development, adulthood and disease, with unclear results. To clearly address the stages of vascular lineage, new technologies and approaches are needed. These approaches might include sequencing strategies (for example, to identify somatic single-nucleotide variants), non-invasive imaging methods and realistic in vitro models.

Although the exact origins of brain vascular cells are still unknown, Wälchli and colleagues identify new vascular stem cell states in one of their most exciting findings. These cell states include stem-to-endothelial cell transdifferentiating clusters and proliferating endothelial-to-mesenchymal transition clusters. In both homeostasis and disease, the remarkable therapeutical, functional and transcriptional plasticity of cells in different vascular beds is emerging. Endothelial cells can secrete factors that support neural stem cells throughout the brain, express transcripts that mimic their cardiomyocyte neighbours in the heart, and can be infused into patients to rejuvenate stem cell niches after bone marrow transplantation^7–9. Pioneering studies in glioblastoma established that cancer cells create new, transformed endothelium that probably fuels the notorious malignance of these tumours¹⁰. These disparate findings point to one theme: that the thousands of miles of vascular beds are a workhorse of health-span in multiple organs, but especially the brain. This study is an important step towards deciphering the individual endothelial cellular decisions that enable thriving vascular niches while selecting against cancer-causing ones.

In summary, it is a thrilling time in brain vascular biology. The available tools, including in vivo imaging and omics technologies, enable an unprecedented understanding of both intrinsic neurovascular biology as well as multicellular communications in the neurovascular unit. Building on this atlas and others, we look forward to many future explorations and discoveries in this field.