The zebrafish hindbrain vascular network assembles in a stereotyped sequence: the basilar artery (BA) forms by medial sprouting of endothelial cells from the bilateral primordial hindbrain channels (PHBCs), a hybrid angiogenesis/vasculogenesis process, and a second wave of dorsal PHBC sprouts forms the central arteries (CtAs) that penetrate the hindbrain and connect to the BA. Fujita and colleagues (2011) showed that chemokine signaling (ligand Cxcl12b in the ventral neural keel, receptor Cxcr4a on migrating endothelial cells) is required for BA assembly and CtA targeting, and PI3K acts downstream; that work explicitly left one mechanistic issue unresolved — whether the CtA targeting defect reflects direct chemokine guidance or a secondary consequence of failed BA assembly — and identified three further gaps that it did not resolve: the effectors acting downstream of CXCR4 beyond PI3K, the cues generating metameric CtA positioning (including the negative cues implied by CtA avoidance of Olig2+ clusters), and the absence of any detailed examination of this assembly process in other vertebrates. As of 2026 these issues remain open despite substantial partial progress: (i) whether chemokine signaling directly guides CtA sprouts to the BA or acts indirectly through BA assembly or BA-derived cues has still not been disentangled by tissue-selective perturbation; (ii) the molecular effectors converting CXCR4 activation into endothelial alignment, detachment from the PHBC, and lumenization remain undefined beyond PI3K (candidate effectors eNOS and Eph/Ephrin are untested in this context, and Ca2+ dynamics have only recently been implicated in a later capillary-network setting); (iii) the cues generating the metameric, rhombomere-centered positioning of CtA emergence are only partially identified (pax6a+ neuroprogenitor-derived Vegfaa and glial Vegf sources have been implicated as positive cues), and the proposed negative cues from Olig2+ neuronal clusters that CtAs avoid remain unidentified; and (iv) no comparative developmental analysis has established whether the zebrafish hybrid BA assembly mode generalizes to other vertebrates, where only gross-anatomical paired-precursor-to-midline-fusion descriptions exist.

Background

The vertebrate brain is vascularized by a stereotyped, evolutionarily conserved program whose cellular and molecular logic is only partly understood. In zebrafish (Danio rerio) embryos, where the optically accessible hindbrain can be imaged live in transgenic lines labelling endothelial cells (for example Tg(fli1a:EGFP) and Tg(kdrl:GFP); developmental time is expressed as hours post-fertilization, hpf), Fujita et al. (2011) reconstructed the assembly of the hindbrain vascular network step by step. The basilar artery (BA) — the single midline artery running along the ventral keel of the hindbrain, the main arterial supply of the brainstem in all vertebrates — was shown to form by a previously undescribed composite process: endothelial sprouts emerge from the medial walls of the bilateral primordial hindbrain channels (PHBCs, a pair of primitive veins flanking the hindbrain), migrate to and align along the ventral neural midline, detach from their PHBC connections, and coalesce into an independent lumenized arterial tube. Because this combines sprouting from a pre-existing vessel (angiogenesis) with local de-facto assembly of a new vessel from migrating cells (vasculogenesis), it was called a hybrid angiogenesis/vasculogenesis mode. A temporally distinct second wave of sprouts from the dorsal PHBC walls gives rise to the central arteries (CtAs), which penetrate the hindbrain parenchyma and interconnect the PHBC-derived perineural vessels with the BA; the BA itself emits no sprouts and serves as the target for CtA connection. Patterning of the CtAs is metameric — matched to the segmental organization of the hindbrain into rhombomeres (transient neuromeres that assign segmental identity to hindbrain neurons): CtAs grow through the centers of individual rhombomeres rather than along rhombomere boundaries, in a fixed sequence beginning with rhombomere 3, and in rhombomeres 5 and 6 the growing CtAs visibly turn away from and avoid nearby clusters of Olig2+ neuroepithelial precursors (motor neuron/oligodendrocyte progenitors labelled by Tg(olig2:DsRed2)), suggesting undefined negative guidance cues. Mechanistically, the chemokine ligand cxcl12b is expressed in the ventral neural keel (the solid rod of neural tissue that forms the fish embryonic brain before the neural-tube lumen appears) adjacent to the assembling BA, and its receptor cxcr4a is expressed in the migrating PHBC endothelial cells; knockdown of either disrupts BA formation and CtA connection to the BA. Downstream, pharmacological inhibition of phosphoinositide 3-kinase (PI3K, a lipid-kinase signal transducer acting downstream of many receptor tyrosine and G-protein-coupled receptors) with LY294002 arrests PHBC medial migration and BA formation in vivo, and CXCL12-stimulated migration of cultured human umbilical vein endothelial cells is blocked by the CXCR4 antagonist AMD3100 and by PI3K inhibition. Two results from other systems suggested candidate additional effectors: endothelial progenitor-cell migration was reported to proceed through a PI3K/Akt/eNOS pathway (eNOS, endothelial nitric oxide synthase, the enzyme generating the vasodilator nitric oxide), and EphB2/EphB4 forward signaling (Eph receptors and their ephrin membrane-bound ligands, a contact-dependent signaling system controlling vascular remodeling) was reported to promote SDF-1/CXCL12-induced endothelial chemotaxis and branching. In the companion studies of the same period, Ulrich et al. (2011) provided cellular-resolution documentation that hindbrain angiogenic sprouts reproducibly enter through rhombomere centers where differentiated neurons reside, and Bussmann et al. (2011) showed that cxcr4a expression in brain capillaries is negatively regulated by blood flow, so that unperfused vessels maintain cxcr4a expression. Comparative anatomy (Gilland and colleagues) establishes that across jawed vertebrates the BA derives ontogenetically from paired precursors that usually fuse at the midline, while the unpaired midline penetrating arteries of teleosts are a derived trait relative to the circumferential penetrating branches of other vertebrates — but this is adult-anatomical knowledge, not a cellular-mechanistic comparison of the assembly process. The unresolved question is thus set within a field (developmental angiogenesis of the CNS) in which guidance of brain vessel sprouting by tissue-derived cues, and its downstream endothelial transduction, are central open problems.

Problem Statement

The common objective is a mechanistic account of how the zebrafish hindbrain vascular network is assembled and patterned — and whether that program is conserved — building on the description and chemokine requirement established by Fujita et al. (2011). Four linked questions, the first explicitly posed as unresolved by that work and the other three grounded in its own observations, cited candidate pathways and acknowledged gaps, jointly constitute the problem: (1) Direct versus indirect chemokine action in CtA targeting: determine whether the CtA targeting defects caused by loss of cxcl12b or cxcr4a reflect a direct requirement for chemokine guidance of CtA sprouts, or an indirect consequence of failed basilar artery assembly or of altered BA-derived targeting cues — a discrimination that global perturbation alone cannot deliver, since chemokine signaling would have to be altered separately in CtA-forming endothelial cells and in the BA/neural-keel target. (2) Downstream transduction: identify the molecular mechanism(s) by which CXCR4 activation is converted into the coordinated endothelial processes of alignment along the ventral neural keel, detachment from the primordial hindbrain channels, and lumenization of the basilar artery — going beyond the implicated PI3K pathway to establish or exclude candidate effectors such as eNOS, Eph/Ephrin-mediated remodeling, and Ca2+ dynamics in this context. (3) Metameric positioning cues: identify the cues that establish the metameric, rhombomere-centered positioning and emergence sequence of the CtAs, including the nature of the positive signals (of which pax6a+ neuroprogenitor-derived Vegfaa is a recently identified candidate) and of the putative negative signals from hindbrain neuronal clusters such as the Olig2+ clusters that CtAs turn away from, and link them mechanistically to the observed CtA trajectories. (4) Evolutionary conservation: determine the extent to which the zebrafish hybrid angiogenesis/vasculogenesis mode of basilar artery assembly generalizes to other vertebrates, and how species-specific differences in transient plexus formation relate mechanistically. The four questions are facets of one mechanistic account of hindbrain vascular assembly and patterning.

References

  1. Fujita M, Cha YR, Pham VN, Sakurai A, Roman BL, Gutkind JS, Weinstein BM (2011) Assembly and patterning of the vascular network of the vertebrate hindbrain. Development 138(9):1705-1715. doi:10.1242/dev.058776 https://doi.org/10.1242/dev.058776
  2. Ulrich F, Ma L-H, Baker RG, Torres-Vázquez J (2011) Neurovascular development in the embryonic zebrafish hindbrain. Developmental Biology 357(1):134-151. doi:10.1016/j.ydbio.2011.06.037 https://doi.org/10.1016/j.ydbio.2011.06.037
  3. Bussmann J, Wolfe SA, Siekmann AF (2011) Arterial-venous network formation during brain vascularization involves hemodynamic regulation of chemokine signaling. Development 138(9):1717-1726. doi:10.1242/dev.059881 https://doi.org/10.1242/dev.059881
  4. Zheng H, Fu G, Dai T, Huang H (2007) Migration of endothelial progenitor cells mediated by stromal cell-derived factor-1α/CXCR4 via PI3K/Akt/eNOS signal transduction pathway. Journal of Cardiovascular Pharmacology 50(3):274-280. doi:10.1097/fjc.0b013e318093ec8f https://doi.org/10.1097/fjc.0b013e318093ec8f
  5. Salvucci O, de la Luz Sierra M, Martina JA, McCormick PJ, Tosato G (2006) EphB2 and EphB4 receptors forward signaling promotes SDF-1-induced endothelial cell chemotaxis and branching remodeling. Blood 108(9):2914-2922. doi:10.1182/blood-2006-05-023341 https://doi.org/10.1182/blood-2006-05-023341
  6. Wang H, Luo L, Yang D (2017) Loss of Gspt1l disturbs the patterning of the brain central arteries in zebrafish. Biochemical and Biophysical Research Communications 486(1):156-162. doi:10.1016/j.bbrc.2017.03.018 https://doi.org/10.1016/j.bbrc.2017.03.018
  7. Tao B, Hu H, Mitchell K, Chen J, Jia H, Zhu Z, Trudeau VL, Hu W (2018) Secretogranin-II plays a critical role in zebrafish neurovascular modeling. Journal of Molecular Cell Biology 10(5):388-401. doi:10.1093/jmcb/mjy027 https://doi.org/10.1093/jmcb/mjy027
  8. Umans RA, Pollock C, Mills WA, Clark KC, Pan YA, Sontheimer H (2021) Using Zebrafish to Elucidate Glial-Vascular Interactions During CNS Development. Frontiers in Cell and Developmental Biology 9:654338. doi:10.3389/fcell.2021.654338 https://doi.org/10.3389/fcell.2021.654338
  9. Eberlein J, Herdt L, Malchow J, Rittershaus A, Baumeister S, Helker CSM (2021) Molecular and Cellular Mechanisms of Vascular Development in Zebrafish. Life 11(10):1088. doi:10.3390/life11101088 https://doi.org/10.3390/life11101088
  10. Trindade M, Silva N, Rodrigues J, Kawakami K, Campinho MA (2025) Maternal thyroid hormone is required to develop the hindbrain vasculature in zebrafish. Communications Biology 8:960. doi:10.1038/s42003-025-08404-1 https://doi.org/10.1038/s42003-025-08404-1
  11. Gui N, Sako K, Fukumoto M, Mochizuki N, Nakajima H (2025) Chemokine induces phase transition from non-directional to directional migration during angiogenesis. Cell Structure and Function 50(1):91-101. doi:10.1247/csf.24081 https://doi.org/10.1247/csf.24081
  12. Shen X, Wen Z, Deng S, Qiu Y, Ma W, Dong X, Gong J, Zhang Y, Liu D, Xu B (2025) Regulation of Hindbrain Vascular Development by rps20 in Zebrafish. Cells 14(14):1070. doi:10.3390/cells14141070 https://doi.org/10.3390/cells14141070
  13. Rahmat S, Gilland E (2014) Comparative Anatomy of the Carotid-Basilar Arterial Trunk and Hindbrain Penetrating Arteries in Vertebrates. The Open Anatomy Journal 6:1-26. doi:10.2174/1877609401406010001 https://doi.org/10.2174/1877609401406010001
  14. Rahmat S, Gilland E (2019) Hindbrain neurovascular anatomy of adult goldfish (Carassius auratus). Journal of Anatomy 235(4):783-793. doi:10.1111/joa.13026 https://doi.org/10.1111/joa.13026
  15. Chiu S, Zhou Q, Xiao C, Bai L, Zhu X, Ding W, Xiong J-W (2026) Glial betaPix is essential for blood vessel development in the zebrafish brain. eLife 14:RP106665. doi:10.7554/eLife.106665 https://doi.org/10.7554/eLife.106665