Hübner et al., 2017 - Wnt Signaling Positively Regulates Endothelial Cell Fate Specification in the Fli1a-Positive Progenitor Population via Lef1. Developmental Biology   430(1):142-155 Full text @ Dev. Biol.

Fig. 1

Wnt signaling is active in the fli1a-positive progenitor cell population of the posterior LPM, consisting of medial kdrl-positive cells and more lateral gata1-positive cells. (A-D) Position of the progenitor cell population (A, Tg(fli1a:GFP)y1), endothelial cells (B, Tg(kdrl:GFP)s843, ECs) and erythrocytes (C, Tg(gata1:GFP)la781) in the posterior LPM at 14 hpf (10ss). (a-c, a’-c’) Magnified dorsal images of the left side of the posterior LPM region caudal to the 9th somite. Scale bars 30 µm. The fli1a-positive cell population has a width of 4–5 cells (A), the kdrl-positive cell population has a width of 1–2 cells (B) and the gata1-positive cell population has a width of 3–4 cells (C). Double labeling of fli1a-positive and kdrl-positive cells (D, left panel, Tg(kdrl:GFP)s843;(fli1a:dsRed)y1 with immuno-labeling of dsRed) or of gata1-positive erythrocytes and kdrl-positive ECs (D, right panel, Tg(kdrl:GFP)s843;(gata1:dsRed)sd2 with immuno-labeling of dsRed) indicated, that kdrl-positive ECs are sorted more medially within the fli1a-positive cell population than gata1-positive cells. (E) Schematic of the posterior LPM at 14 hpf illustrating the localization of the kdrl-positive (ECs, green) and the gata1-positive (erythrocytes, red) cell populations within the fli1a-positive progenitor cell population (purple) as shown in A-D. (F-H) Time course of the migration of fli1a-positive cells from 12 to 16 hpf. Dorsal views of the posterior LPM. Images were processed either as surface rendering projections (F,G) or displayed as maximum intensity projections (H, upper panel) or using inverted colors (H, lower panel) for better visualization. Tg(fli1a:GFP)y1 is expressed in bilateral stripes that migrate towards the midline (F). Wnt responding cells from Tg(axin2BAC:Venus-Pest)mu288 (G) or cre mRNA injected Tg(14TCF:loxP-STOP-loxP-dGFP)mu202 (H, upper panel) show mesodermal expression in the developing somites and expression in bilateral stripes similar to Tg(fli1a:GFP)y1. Double transgenic embryos for Tg(14TCF:loxP-STOP-loxP-dGFP)mu202 and Tg(fli1a:cre)mu225 exhibit GFP expression in a subset of fli1a-positive cells at 14 hpf (H, lower panel). Grey lines surround the region of the migrating fli1a-positive cell population.

Fig. 2

Tg(axin2BAC:Venus-Pest)mu288 and Tg(14TCF:loxP-STOP-loxP-dGFP)mu202 lines represent bona fide β-catenin dependent Wnt signaling reporters. (A,B) Domain structure of Tg(axin2BAC:Venus-Pest)mu288 (A) and Tg(14TCF:loxP-STOP-loxP-dGFP)mu202 (B), and whole mount in situ hybridization using antisense axin2, Venus and GFP probes. axin2 and Venus and GFP are expressed in similar domains: at 13 hpf hindbrain-midbrain boundary, posterior neuroectoderm and mesoderm; at 24 hpf in parts of the developing CNS, the skin and slightly in the tail tip with the GFP expression at 24hpf being limited to smaller domains in the CNS. (C) Tg(axin2BAC:Venus-Pest)mu288 (left panel) and Tg(14TCF:loxP-STOP-loxP-dGFP)mu202 (right panel) embryos were exposed to a Wnt signaling inhibitor (IWR-1) or a Wnt signaling activator (BIO) from 9 to 13 hpf or to 24 hpf and analyzed by confocal microscopy. Treatment with IWR-1 strongly reduced the fluorescence signal of both reporter lines compared to DMSO controls. Vice versa, treatment with Wnt signaling activator BIO resulted in strongly increased expression of both reporters compared to the respective controls. Hence, Tg(axin2BAC:Venus-Pest)mu288 and Tg(14TCF:loxP-STOP-loxP-dGFP)mu202 faithfully respond to Wnt signaling stimulation. For analysis, Tg(14TCF:loxP-STOP-loxP-dGFP)mu202 embryos were injected with cre mRNA at single cell stage.

Fig. 3

β-catenin dependent Wnt signaling positively regulates the number of ECs, while negatively regulating the number of erythrocytes. (A) Schematic representation of the experiment. (B) EC volume and cell numbers were analyzed by GFP expression in Tg(kdrl:GFP)s843 embryos. (C) Erythrocyte volume and cell numbers were analyzed by GFP expression in Tg(gata1:GFP)la781 embryos. Confocal images with the insets representing the overview, and the larger images displaying the specific volume used for quantification. Nuclei were visualized using TOPRO-3 (white). Inhibition of Wnt signaling by IWR treatment led to a reduction in EC volume (B, n = 10) and EC cell numbers (B, n = 19), and to an increase in erythrocyte volume (C, n = 6) and erythrocyte cell numbers (C, n = 22) compared to DMSO treated control embryos (volumes: B, n = 11; C, n = 8; cell numbers: B, n = 30; C, n = 25). Wnt activation (BIO) increased the EC volume (B, n = 8) and the EC number (B, n = 10), while decreasing the erythrocyte volume (C, n = 8) and the erythrocyte number (C, n = 36) compared to DMSO treated control embryos. Values represent mean±SD. *p<0.05, **p<0.01, ***p<0.001; Student's t-test. (D,E) Alternative analysis of the EC population by GFP expression of Tg(etv2:GFP)ci1 (D) and of the erythrocyte cell population using whole mount in situ hybridization with an antisense tal1 probe (E). As observed by kdrl or gata1 expression (B,C), Wnt inhibitor treatment (IWR-1) resulted in reduced EC volume, but in an expanded erythrocyte cell population. Treatment with Wnt activator (BIO) increased the EC volume and decreased the erythrocytes population, respectively. Note, that for EC volume analysis of Tg(etv2:GFP)ci1 the GFP staining in the muscle tissue was cropped and not included in the analysis. (F) Analysis of the EC volume using Tg(kdrl:GFP)s843 and erythrocyte volume using Tg(gata1:GFP)la781 in embryos treated with either Wnt inhibitor (IWR-1), proliferation inhibitor (AHU) or both (AHU+IWR-1). Inhibition of proliferation by AHU decreased the volume of kdrl:GFP and gata1:GFP compared to untreated control embryos, but had no effect on the ratio of both populations (indicated as gata1:GFP/kdrl:GFP). Hence, Wnt signaling specifically affects cell fate specification of ECs and erythrocytes and does not act via asymmetric cell division. (F, n ≥ 8), Values represent mean±SEM. *p<0.05, **p<0.01, ***p<0.001; Student's t-test.

Fig. 4

EC specification is induced by Wnt3a and mediated via Lef1. (A,B) Morpholino (MO)-mediated knockdown of wnt3a reduced the EC number (A, n = 14) in Tg(kdrl:GFP)s843 embryos compared to control (ctr.) MO (A, n = 19). Knockdown of other Wnt ligands, expressed in the zebrafish posterior LPM, did not significantly change EC numbers compared to ctr. MO injected embryos (A): wnt9a (n = 13), wnt9b (n = 14), wnt8 (n = 19) or wnt3 (n = 7). (C) MO-mediated knockdown or transient CRISPR-Cas9-mediated knockdown of wnt3a significantly reduced the EC volume in Tg(kdrl:GFP)s843 embryos (ctr. MO: n = 5; wnt3a MO: n = 8; ctr. CRISPR: n = 5; wnt3a CRISPR: n = 8). Agarose gel electrophoresis> of wnt3a CRISPR genotyping showed an undigested putatively mutated PCR product (wnt3a CRISPR fragment). Non-mutated wild type fragments were cut by PstI digest. (D,E) MO-mediated knockdown of lef1 reduced the EC number (D, n = 10) compared to ctr. MO (D, n = 38) in Tg(kdrl:GFP)s843 embryos. Knockdown of any of the other Tcf-transcription factors did not significantly change EC numbers compared to ctr. MO injected embryos (D): tcf7 (n = 30), tcf7l1a (n = 23), tcf7l1b (n = 28) or tcf7l2 (n = 32). (F) Heatshock induced Wnt ligand overexpression using Tg(hsp70l:wnt8-GFP)w34 (indicated as hs:wnt8) increased the EC volume in Tg(kdrl:mCherry)s896 embryos (F, n = 7) compared to control siblings (F, n = 12). In contrast, heatshock induced overexpression of Wnt signaling inhibitor Dkk1 using Tg(hsp70l:dkk1-GFP)w32 (indicated as hs:dkk1) decreased the EC volume (F, n = 5) compared to control siblings. (G) EC and erythrocyte cell volume analysis in lef1u767 mutant embryos. Embryos carrying heterozygous and homozygous lef1u767 mutations showed reduced EC volumes (G, upper panel, wt sibling: n = 14; lef1u767/+: n = 18; lef1u767/u767: n = 10), but increased erythrocyte volumes (G, lower panel, wt sibling: n = 14; lef1u767/+: n = 19; lef1u767/u767: n = 9) compared to wild type siblings. (H) MO-mediated knockdown of lef1 decreased the EC volume (H, upper panel, n = 6) compared to ctr. MO (H, upper panel, n = 7) in Tg(kdrl:GFP)s843 embryos, but increased the erythrocyte volume (H, lower panel, n = 7) compared to ctr. MO (H, lower panel, n = 7) in Tg(gata1:GFP)la781 embryos. Overactivation of Wnt signaling in lef1-deficient embryos using BIO restored EC volume and erythrocyte volume nearly to control levels (H, upper and lower panel, n = 7). All values represent mean±SD. *p<0.05, **p<0.01, ***p<0.001; Student's t-test.

Fig. 5

wnt3a is expressed in the medial mesoderm, flanked by fli1a-positive cells, while lef1 is expressed in a subpopulation of fli1a-positive cells. (A) Position of the fli1a-positive progenitor cell population in a 14 hpf old zebrafish embryo (blue). Left: lateral view, Right: dorsal view, with the red box at the posterior side indicating the region of analysis. (B-C) Double in situ hybridization for fli1a (red) and wnt3a (blue). Wild type embryos showed wnt3a expression in the medial mesoderm, flanked by fli1a expressing tissue (B, higher magnification b). In homozygous> cloche mutant embryos lacking ECs and erythrocytes, fli1a expression was absent, while wnt3a expression could be detected in the medial mesoderm in a patchier pattern compared to the wild type siblings (C). (D-G) In situ hybridization for fli1a or/and lef-1. fli1a was expressed bilaterally in posterior LPM (D, higher magnification d). In comparison, lef1 was also expressed bilaterally in the posterior LPM, but in a narrower region (E, higher magnification e) and additionally in the tailbud (asterisk). Moreover, the bilateral expression of fli1a (red) and lef1 (blue) was absent homozygous cloche mutant embryos (G), compared to the wild type siblings (F, higher magnification f). (H,I) Confocal maximum intensity projections of fluorescent double in situ hybridization of fli1a (red) and lef1 (green) (H, higher magnification h) and subsequent software based co-localization analysis (I, higher magnification i) indicated co-expression of lef1 in a subset of fli1a expressing cells.

Fig. 6

Wnt signaling promotes EC differentiation in human ES cells. (A) Differentiation protocol for investigating the effects of Wnt inhibition after an initial treatment with BMP, Wnt, and FGF to induce a mesodermal progenitor state. (B) RT-qPCR analysis of endothelial markers PECAM1 and CDH5 on day 10 of differentiation, using distinct differentiation conditions illustrated in A. (C) Immunocytochemical analysis of PECAM1 expression on day 10 using distinct differentiation conditions illustrated in A. (D-F) Gene expression time-course analysis of the primitive streak marker MIXL1 (E) as well as LEF1 (F) (microarray data). Note the decrease in LEF1 expression by inhibition of autocrine WNT signaling on days 2 and 3.

Acknowledgments:
ZFIN wishes to thank the journal Developmental Biology for permission to reproduce figures from this article. Please note that this material may be protected by copyright.

Reprinted from Developmental Biology, 430(1), Hübner, K., Grassme, K.S., Rao, J., Wenke, N.K., Zimmer, C.L., Korte, L., Mu Ller, K., Sumanas, S., Greber, B., Herzog, W., Wnt Signaling Positively Regulates Endothelial Cell Fate Specification in the Fli1a-Positive Progenitor Population via Lef1, 142-155, Copyright (2017) with permission from Elsevier. Full text @ Dev. Biol.