Angiogenesis: To branch or to expand?
- Weatherall Institute of Molecular Medicine, John Radcliffe Hospital, United Kingdom
The network of blood vessels in the body needs to be continuously repaired and extended to enable it to supply sufficient oxygen and nutrients to tissues. New blood vessels are created from existing ones in a complex process called angiogenesis. Normally, this process is tightly controlled because it can also promote the progression of many diseases, such as inflammatory arthritis and cancer. However, little is known about the mechanisms that determine whether a vessel should create a new branch, or widen existing vessels (a process that is known as expansion).
A well established model for the control of blood vessel branching involves a feedback loop between two key signal proteins called vascular endothelial growth factor A (VEGF) and Delta-like four (Dll4) (Blanco and Gerhardt, 2013). Tumour cells or pro-inflammatory cells that are starved of oxygen release VEGF, which activates the branching process by binding to a receptor on so-called tip cells (Figure 1A). These cells are a type of endothelial cell found on the inner surface of blood vessels. They have outgrowths called filopodia that sense the VEGF signals and respond by moving towards the source.
Figure 1
VEGF and Dll4 regulate blood vessel branching and expansion.
(A) Tumour cells in areas of low oxygen levels release VEGF, which is detected by receptors (VEGFR2) on tip cells (yellow). This triggers the tip cells to display more Dll4 proteins on their surface. Dll4 binds to Notch receptors on neighbouring cells (green), which leads to them becoming stalk cells. This results in a new branch of the vessel (sprout) forming with a mosaic pattern of tip and stalk cells, where the tip cells are separated by stalk cells. (B) The mosaic pattern of tip and stalk cells is not permanent and changes as the new sprout grows towards the source of the VEGF signal. Ubezio, Blanco et al. used fluorescent reporters to show that some stalk cells have previously been tip cells. The levels of VEGF increase as the sprout expands and high levels of VEGF (which also lead to high levels of Dll4) disrupt the mosaic pattern and lead to stunted sprouting, reduced branching and allow the vessel to expand.
VEGF triggers an increase in the amount of Dll4 protein on the surface of the tip cells. Dll4 can bind to a receptor protein called Notch that is found on neighbouring endothelial cells. This reduces the amount of VEGF receptor on these so-called stalk cells, preventing them from responding to VEGF signals. This process – referred to as Dll4/Notch lateral inhibition – generates an evenly scattered “mosaic” arrangement of tip cells that are separated by no more than two stalk cells (Bentley et al., 2014; Claxton and Fruttiger, 2004).
The mosaic pattern is not a permanent feature. As the tip cells move out from the existing blood vessel, the stalk cells follow to create the new branch (Figure 1B). The stalk cells then multiply so that the tip cells are now separated by larger numbers of stalk cells, which enables the new branch of the vessel (the sprout) to grow towards the source of the signal. VEGF and Dll4/Notch lateral inhibition are responsible for the continuous competition between these cell types as the sprout grows towards the VEGF gradient. Now, in eLife, Katie Bentley and Holger Gerhardt at the London Research Institute and colleagues – including Benedetta Ubezio and Raquel Blanco as co-first authors – report that Dll4 levels fluctuate in individual endothelial cells as the sprouts form (Ubezio et al., 2016).
Ubezio, Blanco et al. – who are based at the London Research Institute and Vesalius Research Centre – combined video microscopy data with computer simulations to predict and demonstrate how a network of blood vessels could develop. By monitoring the levels of Dll4 messenger ribonucleic acid (mRNA) in cultured single layers of endothelial cells, they found that the production of DII4 in these cells oscillated in response to VEGF. In cells exposed to low levels of VEGF, Dll4 mRNA fluctuated up and down in 4-5 hour cycles. In contrast, high levels of VEGF resulted in a much larger peak of Dll4 mRNA at about 8 hours that was dependent on Notch signalling. In more complex models, high VEGF led to the formation of shorter vessels with wider diameters. The vessels contained tip cells arranged in clusters instead of an alternating mosaic pattern.
Cell culture experiments using fluorescent reporters that respond to Dll4 confirmed that the mosaic pattern of tip and stalk cells seen in animals could be reproduced under low levels of VEGF. Next, Ubezio, Blanco et al. manipulated the developing blood vessels within the eyes of newborn mice by artificially increasing VEGF levels, and blocking the multiplication of stalk cells. This enabled them to exclude cell division as an explanation for the change in vessel size. This confirmed that high VEGF levels synchronise the Dll4 signal across many cells and drive vessel expansion. Furthermore, the fluorescent reporters show that some of the stalk cells in the expanding vessels have previously been tip cells. In a mouse model of glioblastoma – a type of brain tumour – vessel branching was associated with a scattered pattern of Dll4 expression, indicating that the tip and stalk cells were arranged in a mosaic pattern. On the other hand, increasing the levels of Dll4 or VEGF in the eyes disrupted vessel branching and the scattered Dll4 pattern, and promoted vessel expansion.
Why is it important to understand how blood vessels decide to branch or expand in tumours? Drugs that inhibit angiogenesis have been developed to treat cancer and although they slow the progression of the disease, they rarely increase the lifespan of the patient (Bridges and Harris, 2011). The loss of filopodia from VEGF-induced sprouts due to high VEGF levels and the synchronisation of Dll4 signals may lead to vessels expanding in tumours. This could be an aspect of vascular normalisation, where blood vessel walls become less permeable to blood plasma proteins and the flow of blood along the vessel increases (Wong et al., 2015; Jain, 2013). However, many tumours do not show this effect and the enlarged vessels that remain may still be detrimental to the patient because of the increased delivery of oxygen to the tumour and the clearance of waste molecules (Li et al., 2011).
Some studies report that inhibiting VEGF signalling can improve the delivery of drugs that block the growth of tumours (Batchelor et al., 2013), while other studies report the opposite effect (Pham et al., 2015). By understanding how Dll4/Notch lateral inhibition and VEGF signalling regulate angiogenesis, new therapies can be developed to target essential phases of blood vessel growth.
Many questions about angiogenesis remain. For example, how do the tip and stalk cells move towards the source of the VEGF signal? What are the mechanisms that control their ability to adhere to each other (Dejana and Lampugnani, 2014)? How are other types of cells recruited to the tip and stalk cells to form the outside wall of the vessel, and what effect does nutrient starvation have on angiogenesis (Bridges and Harris, 2011; Pham et al., 2015)? With the model developed by Ubezio, Blanco et al., we can expect a much deeper understanding in the near future.
References
-
Improved tumor oxygenation and survival in glioblastoma patients who show increased blood perfusion after cediranib and chemoradiationProceedings of the National Academy of Sciences of the United States of America 110:19059–19064.https://doi.org/10.1073/pnas.1318022110
-
The role of differential VE-cadherin dynamics in cell rearrangement during angiogenesisNature Cell Biology 16:309–321.https://doi.org/10.1038/ncb2926
-
VEGF and Notch in tip and stalk cell selectionCold Spring Harbor Perspectives in Medicine 3:.https://doi.org/10.1101/cshperspect.a006569
-
The angiogenic process as a therapeutic target in cancerBiochemical Pharmacology 81:1183–1191.https://doi.org/10.1016/j.bcp.2011.02.016
-
Periodic Delta-like 4 expression in developing retinal arteriesGene Expression Patterns 5:123–127.https://doi.org/10.1016/j.modgep.2004.05.004
-
Differential adhesion drives angiogenesisNature Cell Biology 16:305–306.https://doi.org/10.1038/ncb2943
-
Normalizing tumor microenvironment to treat cancer: bench to bedside to biomarkersJournal of Clinical Oncology 31:2205–2218.https://doi.org/10.1200/JCO.2012.46.3653
Article and author information
Author details
Publication history
- Version of Record published: June 6, 2016 (version 1)
Copyright
© 2016, Bridges et al.
This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.
Metrics
-
- 2,047
- views
-
- 262
- downloads
-
- 1
- citations
Views, downloads and citations are aggregated across all versions of this paper published by eLife.
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Angiogenesis: To branch or to expand?
eLife 5:e17079.
https://doi.org/10.7554/eLife.17079
Further reading
-
- Computational and Systems Biology
- Genetics and Genomics
We propose a new framework for human genetic association studies: at each locus, a deep learning model (in this study, Sei) is used to calculate the functional genomic activity score for two haplotypes per individual. This score, defined as the Haplotype Function Score (HFS), replaces the original genotype in association studies. Applying the HFS framework to 14 complex traits in the UK Biobank, we identified 3619 independent HFS–trait associations with a significance of p < 5 × 10−8. Fine-mapping revealed 2699 causal associations, corresponding to a median increase of 63 causal findings per trait compared with single-nucleotide polymorphism (SNP)-based analysis. HFS-based enrichment analysis uncovered 727 pathway–trait associations and 153 tissue–trait associations with strong biological interpretability, including ‘circadian pathway-chronotype’ and ‘arachidonic acid-intelligence’. Lastly, we applied least absolute shrinkage and selection operator (LASSO) regression to integrate HFS prediction score with SNP-based polygenic risk scores, which showed an improvement of 16.1–39.8% in cross-ancestry polygenic prediction. We concluded that HFS is a promising strategy for understanding the genetic basis of human complex traits.
-
- Computational and Systems Biology
Revealing protein binding sites with other molecules, such as nucleic acids, peptides, or small ligands, sheds light on disease mechanism elucidation and novel drug design. With the explosive growth of proteins in sequence databases, how to accurately and efficiently identify these binding sites from sequences becomes essential. However, current methods mostly rely on expensive multiple sequence alignments or experimental protein structures, limiting their genome-scale applications. Besides, these methods haven’t fully explored the geometry of the protein structures. Here, we propose GPSite, a multi-task network for simultaneously predicting binding residues of DNA, RNA, peptide, protein, ATP, HEM, and metal ions on proteins. GPSite was trained on informative sequence embeddings and predicted structures from protein language models, while comprehensively extracting residual and relational geometric contexts in an end-to-end manner. Experiments demonstrate that GPSite substantially surpasses state-of-the-art sequence-based and structure-based approaches on various benchmark datasets, even when the structures are not well-predicted. The low computational cost of GPSite enables rapid genome-scale binding residue annotations for over 568,000 sequences, providing opportunities to unveil unexplored associations of binding sites with molecular functions, biological processes, and genetic variants. The GPSite webserver and annotation database can be freely accessed at https://bio-web1.nscc-gz.cn/app/GPSite.
