Suppression of transcytosis regulates zebrafish blood-brain barrier development

As an optically transparent model organism with an endothelial blood-brain barrier (BBB), zebrafish offer a powerful tool to study the vertebrate BBB. However, the precise developmental profile of functional zebrafish BBB acquisition and the subcellular and molecular mechanisms governing the zebrafish BBB remain poorly characterized. Here we find a spatiotemporal gradient of barrier acquisition. Moreover, we capture the dynamics of developmental BBB leakage using live imaging, revealing a combination of steady accumulation in the parenchyma and sporadic bursts of tracer leakage. Electron microscopy studies further reveal that this steady accumulation results from high levels of transcytosis that are eventually suppressed, sealing the BBB. Finally, we demonstrate a key mammalian BBB regulator Mfsd2a, which inhibits transcytosis, plays a conserved role in zebrafish. Mfsd2aa mutants display increased larval and adult BBB permeability due to increased transcytosis. Our findings indicate a conserved developmental program of barrier acquisition between zebrafish and mice.


37
Blood vessels in the vertebrate brain are composed of a single layer of endothelial cells that 38 possess distinct functional properties that allow the passage of necessary nutrients yet prevent 39 unwanted entry of specific toxins and pathogens into the brain. This specialized endothelial layer 40 forms the blood-brain barrier (BBB) and restricts the passage of substances between the blood 41 and the brain parenchyma via two primary mechanisms: 1) specialized tight junction complexes 42 between apposed endothelial cells to prevent intercellular transit (Reese and Karnovsky, 1967; 43 Brightman and Reese, 1969) and 2) suppressing vesicular trafficking or transcytosis to prevent 44 transcellular transit (Ben-Zvi et al., 2014;Andreone et al., 2015;Andreone et al., 2017). BBB 45 selectivity is further refined with expression of substrate-specific transporters that dynamically 46 regulate the influx of necessary nutrients and efflux of metabolic waste products (Sanchez-47 Covarrubias et al., 2014). While the BBB is comprised of endothelial cells, the surrounding 48 perivascular cells including pericytes and astroglial cells, play a critical role in forming and 49 maintaining barrier properties (Janzer and Raff, 1987;Armulik et al., 2010;Bell et al., 2010; 50 Daneman et al., 2010;Wang et al., 2014). Collectively, endothelial cells and the surrounding 51 perivascular cells form the neurovascular unit.

61
Additionally, with the advent of CRISPR-Cas9 technology, zebrafish provide an efficient genetic 4 toolkit for targeted mutagenesis (Hwang et al., 2013;Gagnon et al., 2014;Ablain et al., 2015; 63 Varshney et al., 2015;Albadri et al., 2017;Hogan and Schulte-Merker, 2017). However, the 64 subcellular and molecular mechanisms governing the formation and maintenance of the zebrafish 65 BBB remain poorly characterized. Expanding our understanding of the zebrafish BBB can thus 66 reveal the mechanistic similarities between the zebrafish and mammalian BBB to further elevate 67 the position of zebrafish as a model organism for studying the BBB.

69
Barrier properties of brain endothelial cells are induced by extrinsic signals from other cells in the 70 surrounding microenvironment during development (Stewart and Wiley, 1981). In rodents, the 71 BBB becomes functionally sealed in a spatiotemporal gradient, with the hindbrain and midbrain 72 barriers becoming functional before the cortical barrier (Daneman et al., 2010;Ben-Zvi et al., 73 2014). Within the cortex, barrier function is acquired along a ventral-lateral to dorsal-medial 74 gradient (Ben-Zvi et al., 2014). In the zebrafish, existing studies have disagreed over the timing 75 of zebrafish barrier formation, with some suggesting that BBB maturation occurs at 3 days post 76 fertilization (dpf) (Jeong et al., 2008;Umans et al., 2017) and others providing a wide range 77 beginning at 3 dpf and extending to 10 dpf (Fleming et al., 2013). These conflicting reports may 78 be due to regional developmental gradients of barrier acquisition or differences in the 79 experimental approaches used to assess BBB permeability such as the molecular weight of 80 tracers and circulation time. To date, a thorough regional characterization of functional barrier 81 acquisition has been lacking in zebrafish. 94 Andreone et al., 2017). The subcellular and molecular mechanisms of zebrafish BBB acquisition 95 have yet to be elucidated.

97
Here in zebrafish, we find a spatiotemporal gradient of barrier acquisition, and capture the 98 dynamics of developmental BBB leakage using time lapse live imaging. We further find a 99 conserved role for transcytosis suppression in determining barrier properties, both during normal 100 development and in mfsd2aa mutants.

103
Posterior-Anterior Gradient of Zebrafish BBB Development

104
To determine when and how the zebrafish BBB becomes functional in different brain regions, we 105 performed intracardiac injections of fluorescently-conjugated tracers (1 kDa NHS and 10 kDa 106 Dextran) simultaneously at different developmental stages and imaged live fish after one hour of 107 tracer circulation ( Figure 1A and 1B). We used a combination of different molecular weight tracers 108 to tease apart potential avenues of leakage, as tight junctional defects result specifically in the 109 leakage of low molecular weight tracers 1 kDa and below into the brain parenchyma (Nitta et al., 110 2003;Campbell et al. 2008;Sohet et al., 2015;Yanagida et al. 2017). At 3 dpf, we observed a 111 sealed barrier in the hindbrain as previously described (Jeong et al., 2008), with only a few of the 112 parenchymal cells taking up the injected tracer (average of 2 ± 0.3 cells/embryo with NHS and 2 113 ± 0.4 cells/embryo with Dextran; Figure 1 -Supplement 1), which we quantify as a proxy of tracer 6 leakage into the brain. However, in the midbrain we observed an increased number of 115 parenchymal cells that accumulated the circulating tracers (average of 24 ± 1 cells/embryo with 116 NHS and 24 ± 1 cells/embryo with Dextran; Figure 1C and 1D), indicating that the tracers leaked 117 out of the blood vessels into the brain and that the midbrain barrier is not yet functional. In addition 118 to the use of exogenous injected fluorescent tracers, we also assayed BBB permeability with an 119 endogenous transgenic serum DBP-EGFP fusion protein (Tg(l-fabp:DBP-EGFP)) to account for 120 injection artifacts (Xie et al., 2010). At 3 dpf, we observed similar leakage patterns with the 121 transgenic serum protein as we did with the injected tracers (average of 24 ± 1 cells/embryo in  Figure 1 -Supplement 1). However, the midbrain BBB remains leaky (average 126 of 23 ± 1 cells/embryo with NHS and DBP-EGFP and 24 ± 1 cells/embryo with Dextran; p=0.68 127 compared to 3 dpf, one-way ANOVA; Figure 1C and 1D). However, at 5 dpf the number of 128 midbrain parenchymal cells that uptake the tracers is dramatically reduced (average of 9 ± 1 129 cells/embryo with NHS, Dextran, and DBP-EGFP; p<0.0001, one-way ANOVA; Figure 1B and 130 1C) and no change was observed in the hindbrain (Figure 1 -Supplement 1). No significant 131 change was observed in midbrain permeability from 5 to 6 dpf (average of 8 ± 1 cells/embryo with 132 NHS, Dextran, and DBP-EGFP; p=0.904, one-way ANOVA; Figure 1C and 1D), indicating that 133 the midbrain barrier becomes sealed at 5 dpf. We did not assay the forebrain at any of these 134 stages due to the fact that it remains avascularized until 5 dpf. Of note, all three tracers showed 135 nearly indistinguishable patterns of uptake, with similar numbers of cells and many of the same

234
To investigate whether either paralogue was necessary for barrier formation, we performed NHS 235 tracer injection assays as we did to identify the developmental timeline of barrier formation and 236 assayed for midbrain leakage at 5 dpf. In addition to the use of the exogenous injected fluorescent

255
Given the leakage phenotype in mfsd2aa mutants at 5 dpf, we next wanted to examine whether 256 the leakage phenotype persisted into adulthood. To address this, we performed retro-orbital 257 injections of HRP, which has been shown to be confined within the adult zebrafish brain 258 vasculature (Jeong et al., 2008), and allowed the HRP to circulate for 30 minutes. As expected, 259 the wildtype siblings retained the HRP within their blood vessels ( Figure 5A). However, mfsd2aa 260 mutants exhibited HRP extravasation into the brain parenchyma ( Figure 5B), suggesting that the 261 leakage phenotype was not limited to larval fish. Finally, to determine whether this increased 262 permeability was due to increased transcytosis as in Mfsd2a knockout mice, we measured     (Armulik et al., 2010;Bell et al., 2010;Daneman et al., 2010). Similarly, pericyte deficient 321 notch3 fh332 fish display increased BBB permeability in a tight junction independent manner like 322 pericyte-deficient mice (Wang et al., 2014). Our TEM data show that zebrafish pericytes are in 323 close contact with brain endothelial cells (Figures 2 and 5), even at the earliest stage examined 324 (3 dpf; Figure 2), and are embedded within the endothelial basement membrane as in mammals.

325
Our subcellular localization data is in line with a growing body of evidence for the conserved role 326 of pericytes in the zebrafish BBB (Wang et al., 2014;Lei et al., 2017). While zebrafish lack stellate 327 astrocytes, they possess radial glia that express several astrocytic markers, such as Gfap,    16

544
Functional and developmental analysis of the blood-brain barrier in zebrafish. Brain Res.

545
Bull. 75 (     Dextran intensity in the brain parenchyma over time at 3 dpf (green) and 5 dpf (