Dll4, a novel Notch ligand expressed in arterial endothelium

Abstract

We report the cloning and characterization of a new member of the Delta family of Notch ligands, which we have named Dll4. Like other Delta genes, Dll4 is predicted to encode a membrane-bound ligand, characterized by an extracellular region containing several EGF-like domains and a DSL domain required for receptor binding. In situ analysis reveals a highly selective expression pattern of Dll4 within the vascular endothelium. The activity and expression of Dll4 and the known actions of other members of this family suggest a role for Dll4 in the control of endothelial cell biology.

Figure 1

Murine Dll4 nucleotide sequence with murine and human-deduced amino acid sequence shown. The DSL domain is framed, the eight EGF-like repeats are shaded, and the predicted transmembrane region is underlined. For the human amino acid sequence, only the residues differing from mouse are shown. Human and mouse amino acid sequences are ∼86% identical. The GenBank accession nos. for human and mouse are AF253468 and AF253469, respectively.

Figure 2

Comparison of amino acid identity between vertebrate Delta genes reveals Dll4 is a novel family member. Murine Dll4 is ∼30%–50% identical to known Delta genes, whereas homologs (i.e., X-Delta1 and murine Dll1) are >70% identical. Dll3 remains the most divergent family member.

Figure 3

Murine Dll4 is expressed in most tissues. Northern analysis shows a transcript of ∼4.3 kb in several adult tissues, with lung expressing the highest levels. Expression is also seen in embryos, with increasing levels from embryonic day 9 through day 17. GAPDH hybridization provides for variations in loading and helps explain the lack of signal in spleen and testis.

Figure 4

Murine Dll4 is highly specific for arterial endothelial cells. In situ analysis was performed on murine tissues with an antisense murine Dll4 riboprobe. Dll4 expression in (a) choroid plexus; (b) brain; (c) kidney; (d) lung; (e) umbilical cord. The bright field and dark field views are displayed on the left and right, respectively, for each tissue section. Endothelial cells expressing Dll4 are indicated with arrows. See text for details.

Figure 5

Dll4 acts as a Notch ligand in Xenopus embryos. One blastomere of 2-cell albino embryos was injected with either (a) nLacZ RNA and 0.1–0.25 ng of Dll4 RNA, or (b) nLacZ RNA only. At stage 14, the embryos were fixed, stained in whole mount with X-gal (light blue reaction product), and then labeled by in situ hybridization for the expression of N-tubulin (dark blue). Embryos are presented anterior to the left with the injected side on the top. Note that N-tubulin expression is suppressed in the area of Dll4 injection (as visualized with lacZ). (c) Results of RNase protection assay to measure the levels of ESR-1 and ESR-7, targets of Notch signaling, in neural ectoderm isolated from embryos injected with the indicated RNAs. Ectoderm was neuralized by injection of Noggin. RNAs were extracted at stage 13 and the ability of Dll4 to activate Notch targets were examined. For quantification, specific band intensities in each lane were normalized to the amount of EF1-α RNA to control for sample recovery, and the gene expression results are represented relative to the uninjected controls. Note that Dll4 weakly activates ESR-1 and ESR-7 expression via endogenous X-Notch activation (column 3), whereas ESR gene activation by Dll4 is strongly potentiated by coinjection of murine Notch1 (column 4) to reach the levels induced by X-Delta1 (column 5). In a separate experiment (columns 7–10), the ability of Dll4 to signal through the murine Notch4 receptor was tested. The activation of ESR-1 and ESR-7 expression by Dll4 (column 9) is strongly potentiated by coinjection with murine Notch4 (column 10). In both experiments, activation of targeted genes was increased >3.5-fold upon coinjection of Dll4 with either murine Notch1 or Notch4.

Figure 5

Dll4 acts as a Notch ligand in Xenopus embryos. One blastomere of 2-cell albino embryos was injected with either (a) nLacZ RNA and 0.1–0.25 ng of Dll4 RNA, or (b) nLacZ RNA only. At stage 14, the embryos were fixed, stained in whole mount with X-gal (light blue reaction product), and then labeled by in situ hybridization for the expression of N-tubulin (dark blue). Embryos are presented anterior to the left with the injected side on the top. Note that N-tubulin expression is suppressed in the area of Dll4 injection (as visualized with lacZ). (c) Results of RNase protection assay to measure the levels of ESR-1 and ESR-7, targets of Notch signaling, in neural ectoderm isolated from embryos injected with the indicated RNAs. Ectoderm was neuralized by injection of Noggin. RNAs were extracted at stage 13 and the ability of Dll4 to activate Notch targets were examined. For quantification, specific band intensities in each lane were normalized to the amount of EF1-α RNA to control for sample recovery, and the gene expression results are represented relative to the uninjected controls. Note that Dll4 weakly activates ESR-1 and ESR-7 expression via endogenous X-Notch activation (column 3), whereas ESR gene activation by Dll4 is strongly potentiated by coinjection of murine Notch1 (column 4) to reach the levels induced by X-Delta1 (column 5). In a separate experiment (columns 7–10), the ability of Dll4 to signal through the murine Notch4 receptor was tested. The activation of ESR-1 and ESR-7 expression by Dll4 (column 9) is strongly potentiated by coinjection with murine Notch4 (column 10). In both experiments, activation of targeted genes was increased >3.5-fold upon coinjection of Dll4 with either murine Notch1 or Notch4.