Zebrafish Dpr2 Inhibits Mesoderm Induction by Promoting Degradation of Nodal Receptors

See allHide authors and affiliations

Science  01 Oct 2004:
Vol. 306, Issue 5693, pp. 114-117
DOI: 10.1126/science.1100569


Nodal proteins, members of the transforming growth factor–β (TGFβ) superfamily, have been identified as key endogenous mesoderm inducers in vertebrates. Precise control of Nodal signaling is essential for normal development of embryos. Here, we report that zebrafish dapper2 (dpr2) is expressed in mesoderm precursors during early embryogenesis and is positively regulated by Nodal signals. In vivo functional studies in zebrafish suggest that Dpr2 suppresses mesoderm induction activities of Nodal signaling. Dpr2 is localized in late endosomes, binds to the TGFβ receptors ALK5 and ALK4, and accelerates lysosomal degradation of these receptors.

Genetic studies have revealed that Nodal proteins are essential for mesoderm induction in vertebrates (14). So far, few factors have been found to inhibit Nodal signaling during early embryogenesis. Lefty proteins inhibit Nodal activity by competing for binding to their common receptors (58). An extracellular protein, Cerberus, binds to Nodal and blocks its activity (9). Tomoregulin-1, a transmembrane protein, blocks Nodal signaling by interacting with the Nodal coreceptor Cripto (10). Drap1 interacts with and inhibits DNA binding of FoxH1, a transcription factor that mediates Nodal signaling (11). Here, we demonstrate that Dpr2 inhibits Nodal signaling in zebrafish embryos, targeting Nodal receptors for lysosomal degradation.

Zebrafish dpr2 was identified by a whole-mount in situ hybridization screen for tissue-specific genes. The putative Dpr2 protein, 837 residues long, has overall sequence identity of 30 and 36% to the putative human DAPPER1 and DAPPER2 proteins (12), respectively, with higher similarity in several domains (fig. S1). Several dpr2 ESTs have been mapped adjacent to the oprm1 and smoc2 loci in the zebrafish linkage group 13 (LG13), a region syntenic to the human DAPPER2 locus, which suggests that zebrafish dpr2 is an ortholog of human DAPPER2. Although Xenopus Dapper and Frodo are almost identical, Dapper is apparently a general Dishevelled antagonist (13), whereas Frodo is a positive regulator of Wnt signaling (14).

Whole-mount in situ hybridization revealed a specific expression pattern of dpr2 during zebrafish embryogenesis (fig. S2). It was first expressed on the dorsal blastoderm at the sphere stage, about 4 hours postfertilization (hpf). At the onset of gastrulation, dpr2 expression was restricted to the whole germ ring where mesoderm precursors reside, with the highest expression in the embryonic shield. The dpr2-positive cells then involute and converge during gastrulation and thus contribute to the formation of epiblast and hypoblast. When segmentation starts, dpr2 is expressed in the dorsal trunk neural tube, the lateral plate mesoderm, and the tail bud. At 24 hpf, dpr2 expression is restricted to blood progenitors and the tail bud. This expression pattern of dpr2 suggests a role in early mesoderm induction.

To study the function of dpr2, endogenous expression of dpr2 was knocked down by injecting a mix of two specific and equally effective antisense morpholinos (dpr2-MOs) (fig. S3). The injected embryos generally showed a thicker, curled down trunk at 24 hpf (fig. S3), a phenotype also observed with a lefty1 morpholino (lft1-MO) (15). The majority (53 to 95%) of morpholino-injected embryos showed increased expression of the organizer-specific marker gsc (Fig. 1A), the lateral mesodermal marker snail1 (Fig. 1B), and the axial mesodermal markers no tail (ntl) (Fig. 1C) and shh (Fig. 1D). Mutations of the oep locus, which encodes a coreceptor for the Nodal ligands squint (sqt) and cyclops (cyc) in zebrafish, give rise to reduced Nodal activity (4, 16). Overexpression of dpr2 caused partial or complete fusion of the eyes (Fig. 1F), resembling the oep (Fig. 1G) and cyc phenotypes, which also result from insufficient Nodal signaling (16, 17). Overexpression also caused partial loss of the notochord and tail reduction in about one-third of the embryos at 24 hpf. Injection of dpr2 mRNA also led to reduction or elimination of shh expression at 24 hpf (Fig. 1H), as well as decreased ntl expression at the bud stage (fig. S3). These data suggest that dpr2 functions to inhibit mesoderm formation.

Fig. 1.

Effects of dpr2 knockdown and overexpression in zebrafish embryos. Whole-mount in situ hybridization reveals that morpholino-mediated knockdown of dpr2 resulted in increase of gsc (A) and snail1 (B) at the shield stage, ntl (C) at the5-somitestage, and shh (D) at the 8-somite stage. In each panel, the left embryo was injected with 8 ng control morpholino (cMO) and the right one with 8 ng dpr2-MOs. (A, C, and D) Dorsal views. (B) Lateral view. (E) Injection of dpr2-MOs (middle) or lft1-MO (right), but not cMO (left), led to recovery of shh expression in the floor plate (arrowheads) of oep t z 257 mutant embryos at 24 hpf. (F) Injection of wild-type embryos with 200 pg dpr2 mRNA caused partial or complete fusion of eyes in the 24-hpf embryo, which then resembled the phenotype of oeptz257 embryo (G). (F and G) Head ventral views of live embryos. (H) dpr2 overexpression reduces (middle) or eliminates (right) shh expression in the 24-hpf embryo. Embryos injected with the same amount of GFP mRNA were served as controls in (F) and (H).

Because dpr2 overexpression phenocopied oep mutants and because dpr2 knockdown resulted in morphological changes that resembled lefty1 knockdown, there may be genetic interactions between Dpr2 and Nodal signaling. We found that 55% (38 out of 69) of embryos coinjected with dpr2-MOs and 0.1 pg sqt mRNA had a much wider notochord, an effect not seen in embryos injected with sqt mRNA alone (fig. S4). This notochord effect showed that knockdown of dpr2 can enhance the phenotypes caused by elevated Nodal activity. In addition, 91% (n = 135) of embryos coinjected with 4 ng dpr2-MOs and 1.25 ng lft1-MO underwent arrest of development or died at 24 hpf, whereas this percentage was usually below 10% for single injections. Like lft1 knockdown (Fig. 1E, right), dpr2 knockdown rescued shh expression in the ventral midbrain and in the floor plate (Fig. 1E, middle) in 22 out of 23 zygotic oeptz257 mutants. However, dpr2 knockdown in MZoep mutants, which completely lack Nodal signaling (4, 18), failed to rescue shh expression (19), which suggests that Dpr2 function depends on the availability of Nodal signals. Furthermore, dpr2 knockdown counteracted lft1 overexpression, and dpr2 overexpression compromised the effect of lft1 knockdown (fig. S4).

We also examined the possible involvement of Dpr2 in other signaling pathways prominent in early embryogenesis. Overexpression of dpr2 and bmp7 had similar influences on expression of certain marker genes, and this ectopic bmp7 effect was inhibited by dpr2 knockdown (fig. S4), which suggests that Dpr2 could not inhibit the ventralizing BMP signals. Also, Dpr2 may not be an antagonist of either FGF or Wnt signals, because dpr2 knockdown led to phenotypes different from those caused by ectopic expression of fgf8 and wnt8; also, increased FGF or Wnt signaling failed to rescue shh expression in oep mutant embryos (figs. S5 and S6).

The inhibitory effect of Dpr2 on Nodal-related signaling was next tested in mammalian cells with a transforming growth factor–β (TGFβ)–induced luciferase activity with an activin-responsive element–containing luciferase reporter, ARE-luciferase. The expression of this reporter was stimulated by constitutively active ALK5/TβRI, a TGFβ receptor, in Hep3B cells. This stimulation was attenuated by coexpressing fish dpr2 (Fig. 2A), whereas human DAPPER1 (hDpr1) did not interfere with TGFβ signaling (19). Unlike hDpr1, Dpr2 had little effect on Wnt signaling (Fig. 2B). Furthermore, dpr2 overexpression inhibited the receptor-mediated phosphorylation of Smad2, as well as the formation of Smad2-Smad4 and Smad3-Smad4 complexes (fig. S7). Together, these data indicate that Dpr2 specifically antagonizes TGFβ signaling in mammalian cells.

Fig. 2.

(A) Dpr2 inhibits TGFβ receptor ALK5-mediated expression of the ARE-luciferase reporter in human hepatoma Hep3B cells. (B) Human Dpr1 (hDpr1) attenuates Wnt signaling, whereas Dpr2 does not. Data were from three independent experiments and were expressed as mean with standard deviations. (C) Dpr2 binds to ALK5 in HEK293T cells, as revealed by immunoprecipitation (IP) and immunoblotting (IB). Note that stronger binding was detected for the constitutively active (ca) ALK5. (D) Dpr2 binds to endogenous ALK5. A5, anti-ALK5 antibody; nsp, nonspecific IgG. (E to G) Dpr2 colocalizes with the late endosomal marker Rab7 in HeLa cells. Scale bars, 10 μm.

To determine the mechanism of Dpr2 action on TGFβ signaling, we next tested whether Dpr2 associates with TGFβ receptors. Dpr2 bound to overexpressed ALK5 in HEK293T cells and interacted with the active forms of ALK5 with high affinity (Fig. 2C), which indicates that Dpr2 preferentially associates with the activated receptors. This notion is supported by the observation that Dpr2 associated with endogenous ALK5 upon TGFβ stimulation (Fig. 2D). In contrast, Dpr2 associated with ALK6 (BMPR-IB) and TβRII weakly, if at all, and not at all with ActRII (19).

Immunofluorescence analysis indicated a punctate subcellular localization of Dpr2 (Fig. 2E), which overlapped with the lateendosomal marker Rab7 (Fig. 2F and 2G), but not with the early-endosomal markers EEA1 (fig. S8), Rab5 and transferrin (19), the lysosomal marker LAMP3 (fig. S8), or the caveolae marker Caveolin (fig. S8). TGFβ receptors are internalized into the cell via clathrin-coated pits or lipid rafts and/or caveolae (2023), and are down-regulated by Smad7-Smurf or by β-arrestin-2 (21, 24). The localization of Dpr2 in late endosomes suggests that it may play a role in sorting endocytosed receptors into late endosomes and/or facilitating their transport to lysosomes.

Overexpression of dpr2 led to reduction of receptor protein levels (Fig. 2C, bottom). Pulse-chase analysis confirmed that Dpr2 overexpression accelerated the degradation of ALK5 in HEK293T cells (Fig. 3A and fig. S9). Furthermore, fewer cell surface receptors, owing to a shorter half-life, were detected in the Dpr2-expressing cells (Fig. 3B and fig. S9), which suggests that Dpr2 facilitates degradation of receptors internalized from the cell surface. In addition, Dpr2-induced ALK5 degradation was inhibited by the lysosomal inhibitors bafilomycin A1, NH4Cl, or chloroquine, but not by the proteasome inhibitor MG132 (Fig. 3C). These results suggest that Dpr2 promotes lysosomal degradation of endocytosed receptors.

Fig. 3.

Dpr2 targets ALK5 for lysosomal degradation. (A) Degradation of active ALK5, pulse-labeled with 35S-Met/Cys for 2 hours, was accelerated by dpr2 overexpression, and this degradation was attenuated by NH4Cl. (B) Dpr2 promotes the turnover of biotinylated cell surface ALK5 molecules. (C) Dpr2-mediated ALK5 degradation is sensitive to the lysosomal inhibitors bafilomycin A1 (BF), NH4Cl (NC), and chloroquine (Chlq), but not to the proteasome inhibitor MG-132 (MG). The arrowhead indicates the position of the ALK5 receptor. The amount of myc-Dpr2 was indicated as + (0.1 μg) or ++ (0.2 μg).

Similar experiments with ALK4/ActRIB, an activin and Nodal receptor, revealed that Dpr2 was able to bind to ALK4, and Dpr2 overexpression inhibited active ALK4-mediated reporter expression and promoted degradation of ALK4 through lysosomal pathway (fig. S10). Analysis of the functional regions of Dpr2 demonstrated that the first 361 amino acids could bind to ALK5, and binding of this segment was sufficient to attenuate the reporter gene activation (fig. S11).

During embryogenesis, transcripts of dpr2 were not detected in sqt;cyc double mutants (Fig. 4A), and overexpression of sqt induced ectopic expression of dpr2 (Fig. 4B), which demonstrated an essential role of Nodal signaling in activating dpr2 expression. Knockdown of dpr2 led to a higher expression level of cyc (Fig. 4C) and sqt (Fig. 4D), consistent with a positive activation loop for Nodal signals (25). It was noteworthy that dpr2 knockdown increased its own transcription (Fig. 4, E and F), which implies that intracellular level of Dpr2 is precisely perceived and negatively autoregulated.

Fig. 4.

Reciprocal regulation of dpr2, sqt, and cyc expression. (A) dpr2 expression is not detectable in the sqtL235;cycm294 double mutants. (B) Injection with 0.1 pg sqt mRNA induces ectopic expression of dpr2. (C and D) dpr2 knockdown increases expression of cyc (C) and sqt (D). (E and F) dpr2 knockdown also enhances expression of dpr2 itself. The injected embryos were examined by whole-mount in situ hybridization for expression of the marker genes. In each panel, the control is on the left. (A to B) Animalpole views at the 30% epiboly stage, dorsal to the right. (C and D) Lateral views at the 30% epiboly stage. (E) Lateral view at the shield stage. (F) Dorsal view at the 6-somite stage.

In summary, our results indicate that Dpr2 functions as an antagonist of Nodal signaling during embryonic development, although we cannot fully exclude minor roles of Dpr2 in other signaling pathways. Our data suggest that Dpr2 binds to endocytosed TGFβ/Nodal receptors and facilitates their transport to lysosomes for degradation. Compared with phenotypes caused by lefty1 overexpression, however, phenotypes of dpr2 overexpression are mild. This may be attributed to divergent mechanisms by which Dpr2 and Lefty1 function. Dpr2 acts at late trafficking stages of internalized Nodal receptors, after the extracellular signals have passed to the downstream effectors, the Smad proteins. Thus, Dpr2 overexpression should never completely block initial transduction of Nodal signals. In contrast, Lefty proteins are able to completely shut off initial transduction of Nodal signals as they compete with the ligands for the signaling receptors.

Supporting Online Material

Materials and Methods

Figs. S1 to S11

References and Notes

References and Notes

View Abstract

Stay Connected to Science

Navigate This Article