Author response: Regulation of Nodal signaling propagation by receptor interactions and positive feedback
Year: 2022
Type: peer-review
Abstract: Article Figures and data Abstract Editor's evaluation eLife digest Introduction Results Discussion Materials and methods Appendix 1 Data availability References Decision letter Author response Article and author information Metrics Abstract During vertebrate embryogenesis, the germ layers are patterned by secreted Nodal signals. In the classical model, Nodals elicit signaling by binding to a complex comprising Type I/II Activin receptors (Acvr) and the co-receptor Tdgf1. However, it is currently unclear whether receptor binding can also affect the distribution of Nodals themselves through the embryo, and it is unknown which of the putative Acvr paralogs mediate Nodal signaling in zebrafish. Here, we characterize three Type I (Acvr1) and four Type II (Acvr2) homologs and show that – except for Acvr1c – all receptor-encoding transcripts are maternally deposited and present during zebrafish embryogenesis. We generated mutants and used them together with combinatorial morpholino knockdown and CRISPR F0 knockout (KO) approaches to assess compound loss-of-function phenotypes. We discovered that the Acvr2 homologs function partly redundantly and partially independently of Nodal to pattern the early zebrafish embryo, whereas the Type I receptors Acvr1b-a and Acvr1b-b redundantly act as major mediators of Nodal signaling. By combining quantitative analyses with expression manipulations, we found that feedback-regulated Type I receptors and co-receptors can directly influence the diffusion and distribution of Nodals, providing a mechanism for the spatial restriction of Nodal signaling during germ layer patterning. Editor's evaluation While the Nodal signaling pathway plays essential roles in germ layer induction and patterning during development, the precise function of different zebrafish Nodal receptors had not been determined. Here, the authors identify seven genes encoding Type I and Type II Nodal receptors in zebrafish and use a combination of genetic approaches to show that the Type I receptors Acvr1b-a and Acvr1b-b act redundantly as major mediators of Nodal signaling. Whereas the Acvr2 Type II receptors function partly redundantly and partially independently of Nodal in embryo patterning. Moreover, Type I receptor and co-receptor levels can modulate Nodal distribution, providing a mechanism for the spatial restriction of Nodal signaling during embryo patterning. https://doi.org/10.7554/eLife.66397.sa0 Decision letter eLife's review process eLife digest Building a body is complicated. Cells must organise themselves head-to-tail, belly-to-back, and inside-to-outside. They do this by laying down a chemical map, which is made up of gradients of molecular signals, high in some places and lower in others. The amount of signal each cell receives helps to decide which part of the body it will become. One of the essential signals in developing vertebrates is Nodal. It helps cells to tell inside from outside and left from right. Cells detect Nodal using an activin receptor and co-receptor complex, which catch hold of passing Nodal proteins and transmit developmental signals into cells. An important model to study Nodal signals is the zebrafish embryo, but the identity of the activin receptors and their exact role in this organism has been unclear. To find out more, Preiß, Kögler, Mörsdorf et al. studied the activin receptors Acvr1 and Acvr2 in zebrafish embryos. The experiments revealed that two putative Acvr1 and four Acvr2 receptors were present during early development. To better understand their roles, Preiß et al. eliminated them one at a time, and in combination. Losing single activin receptors had no effect. But losing both Acvr1 receptors together stopped Nodal signalling and changed the distribution of the Nodal gradient. Loss of all Acvr2 receptors also caused developmental problems, but they were partly independent of Nodal. This suggests that Acvr1s seem to be able to transmit signals and to shape the Nodal gradient, and that Acvr2s might have another, so far unknown, role. Nodal signals guide the development of all vertebrates. Understanding how they work in a model species like zebrafish could shed light on their role in other species, including humans. A clearer picture could help to uncover what happens at a molecular level when development goes wrong. Introduction The formation of the body plan during early embryogenesis depends on the interplay between evolutionarily conserved signaling pathways. The TGF-β superfamily member Nodal is one of the key regulators of vertebrate development and is required to specify mesoderm and endoderm (collectively termed mesendoderm) during germ layer formation (Schier, 2009). In the classical model, Nodal ligands signal through a receptor complex comprising Type I and Type II single-transmembrane serine/threonine kinase receptors (Attisano and Wrana, 2002; Shi and Massagué, 2003; Figure 1A). Unlike other members of the TGF-β superfamily, Nodal signaling additionally requires the presence of an EGF-CFC co-receptor to activate signaling. Our current understanding of Nodal signaling is that Nodal directly binds to Type II receptors and the EGF-CFC co-receptor Tdgf1, which in turn mediates the recruitment of the Type I receptors. Upon oligomerization of the receptor complex, Type II receptors phosphorylate the Type I receptors in their GS domains, leading to the recruitment and phosphorylation of the C-terminal SSXS motif of the receptor-regulated Smad (R-Smad) proteins Smad2 and Smad3 by the Type I receptor. The activated pSmad2/pSmad3 proteins associate with the co-factor Smad4 and translocate into the nucleus, where they activate target gene expression (Hill, 2018; Macías-Silva et al., 1996; Shi and Massagué, 2003; Yeo and Whitman, 2001; Figure 1A). Figure 1 with 1 supplement see all Download asset Open asset Multiple Nodal receptor candidates are expressed during early zebrafish development. (A) In the classical model, Nodal signaling requires the recruitment of a receptor complex comprising the co-receptor Oep (Tdgf1 homolog) as well as Type I and Type II Activin receptors (Acvr) to induce phosphorylation and nuclear translocation of the signal transducer pSmad2/3 for the induction of Nodal target genes. (B) Phylogenetic neighbor-joining alignment tree of Type I and Type II receptor protein sequences from human, mouse, and zebrafish. Bootstrap values are listed at the nodes and indicate evolutionary distances. (C) Temporal expression analysis of putative Nodal receptors at different developmental stages. TPM: Transcripts per million. dpf: day(s) post-fertilization. Data adapted from White et al., 2017. (D) Spatial expression analysis of Type I and Type II receptors at 2 cell and shield stages revealed by in situ hybridization. Except for acvr1c, all receptor-encoding transcripts are maternally deposited. At shield stage, acvr1b-a is the only receptor that is not uniformly expressed but restricted to the embryonic margin. (E) Nodal signaling controls the expression of acvr1b-a and oep. Fold change of Nodal receptor expression calculated from qRT-PCR experiments comparing the overexpression of 30 pg squint-GFP mRNA, 30 pg cyclops-GFP mRNA, 30 pg lefty2-Dendra2 mRNA and exposure to 50 μM SB-505124 Nodal inhibitor to untreated embryos at 6 hours post-fertilization (hpf). Each point is the mean fold change of an individual embryo compared to an untreated embryo. Error bars represent standard deviation. (F) In situ hybridization analysis of acvr1b-a with increased (+squint GFP) or decreased (+SB-505124) Nodal signaling. Scale bar represents 250 µm. See the Figure 1—source data 1 file for source data and sample size. Figure 1—source data 1 Source data for Figure 1. https://cdn.elifesciences.org/articles/66397/elife-66397-fig1-data1-v3.xlsx Download elife-66397-fig1-data1-v3.xlsx In zebrafish, mesendoderm patterning depends on the two secreted Nodal signals Squint (Sqt) and Cyclops (Cyc) (Dougan et al., 2003; Rogers and Müller, 2019; Schier, 2009; Shen, 2007). Nodal expression begins in the yolk syncytial layer at the embryonic margin during the blastula stage and then spreads into the embryo, generating a Nodal signaling gradient. This gradient is translated into different mesendodermal cell fates depending on the signaling level and target gene induction kinetics (Dubrulle et al., 2015). Loss of Nodal signaling causes absence of endoderm as well as trunk and head mesoderm, which leads to cyclopia due to a failure to separate the eye fields, resulting in embryonic lethality (Dubrulle et al., 2015; Feldman et al., 1998; Gritsman et al., 1999). Nodal signaling is antagonized by the secreted long-range feedback inhibitor Lefty, which is also produced at the margin (Meno et al., 1999; Thisse and Thisse, 1999). Establishment and maintenance of a correct signaling range is crucial for correct development, as also excess Nodal signaling – for example in lefty mutants – can cause severe patterning defects and embryonic lethality (Almuedo-Castillo et al., 2018; Rogers et al., 2017). Measurements of active GFP-tagged fusions showed that Squint and Cyclops proteins have a lower effective diffusivity than their inhibitors Lefty1 (Lft1) and Lefty2 (Lft2) (Müller et al., 2012; Rogers and Müller, 2019). It has been proposed that this mobility difference is due to interactions between Nodal ligands and membrane-bound diffusion regulators, whereas Lefty proteins move more freely in the extracellular space (Müller et al., 2012; Müller et al., 2013). Indeed, Nodal’s signaling range and distribution dramatically increase in the absence of the zebrafish Tdgf1 co-receptor homolog Oep (Lord et al., 2021), and single-molecule imaging has recently shown that the fraction of molecules in the bound state is larger for Nodal than for Lefty (Kuhn et al., 2022). Since Nodals strongly bind to the zebrafish Type II receptor Acvr2b-a in vivo (Wang et al., 2016), the main Nodal receptors themselves might also act as diffusion regulators. However, it is unclear whether this strong ligand-receptor interaction indeed influences Nodal dispersal, whether receptor binding affects Nodal diffusion or stability in the embryo, and what role other putative Type I and Type II Acvr receptors play in the propagation of Nodal signaling through the embryo. The two mouse, frog and human Type I receptors Acvr1b (also known as Alk4/TARAM-A) and Acvr1c (also known as Alk7) and the two Type II receptors Acvr2a and Acvr2b were identified using in vitro binding and target induction assays, and cause developmental defects when mutated (Gritsman et al., 1999; Gu et al., 1998; Kosaki et al., 1999; Matzuk et al., 1995; Oh and Li, 1997; Reissmann et al., 2001). Surprisingly, except for the zebrafish co-receptor Oep (Gritsman et al., 1999), no zebrafish Nodal receptor mutants are known to recapitulate Nodal loss-of-function phenotypes; and although zebrafish are widely used to investigate Nodal signaling during development, it is unknown which of the receptor paralogs mediate endogenous Nodal signaling during germ layer formation. To understand the role of the zebrafish receptor homologs in Nodal distribution and signaling, we generated several loss-of-function mutants and used them together with combinatorial morpholino knockdown and CRISPR F0 knockout (KO) approaches to assess compound loss-of-function phenotypes. Due to the severity of single receptor knock-outs in mice (Gu et al., 1998; Kosaki et al., 1999; Matzuk et al., 1995; Oh and Li, 1997; Reissmann et al., 2001), we expected phenotypes similar to Nodal loss-of-function mutants in zebrafish. Strikingly, loss of individual receptor function did not cause obvious patterning defects. Severe patterning phenotypes and embryonic lethality were observed with combinatorial loss of putative Type II Acvr receptors, but the defects were at least partly independently of Nodal signaling. Instead, only the combined loss of the Type I receptors acvr1b-a and acvr1b-b phenocopied known Nodal loss-of-function phenotypes (Dubrulle et al., 2015; Feldman et al., 1998; Gritsman et al., 1999), identifying these receptors as the main Type I receptors that mediate early Nodal signaling in zebrafish. Using quantitative imaging assays, we found that Type I receptor and co-receptor levels can modulate Nodal mobility and thereby directly influence the distribution of Nodal in the embryo, providing a mechanism for the spatial restriction of Nodal signaling during germ layer patterning. Results Nodal Type I and Type II receptors have several putative paralogs in zebrafish To systematically identify and characterize zebrafish Nodal receptors, we used the protein sequences of the human and mouse Type I receptors Acvr1b and Acvr1c as well as the Type II receptors Acvr2a and Acvr2b as queries for homology searches in the Uniprot database. In addition to the previously experimentally identified zebrafish Type I (Renucci et al., 1996) and Type II (Garg et al., 1999; Nagaso et al., 1999) Nodal receptor orthologs Acvr1b-a, Acvr1c, Acvr2a-a, and Acvr2b-a, our analysis yielded further potential Nodal receptor paralogous sequences named Acvr1b-b, Acvr2a-b and Acvr2b-b (Funkenstein et al., 2012; Li et al., 2019), respectively, resulting in a total of three putative Type I and four putative Type II receptors. For Acvr2a-b, the start of the gene’s coding sequence and the full amino acid sequence of the receptor was not resolved, and corresponding predictions differed between genome assemblies. Therefore, we performed 5’RACE (Rapid Amplification of cDNA Ends) and thereby mapped the start of acvr2a-b to chrUn_KN150226v1 (see Materials and methods and Figure 1—figure supplement 1A for further details). Reconstruction of a putative phylogenetic tree shows a close clustering of the zebrafish receptors with their human and mouse paralogs, and the highest sequence similarity was found between the zebrafish Type I receptors Acvr1b-a and Acvr1b-b (Figure 1B). All putative homologs have the typical features of Type I and Type II receptors, including a signal peptide, TGF-β receptor domain, transmembrane domain, cytosolic kinase domain and a GS domain in case of the Type I receptors (Figure 1—figure supplement 1B-H). Most Nodal receptor paralog transcripts are present during mesendoderm formation To determine which of the putative Nodal receptor paralogs might have roles in germ layer patterning, we first assessed their expression during early embryogenesis, focusing on early blastula and gastrula stages during which germ layer patterning takes place (Figure 1C). Analysis of a published developmental transcriptome (White et al., 2017) indicated that the transcripts of most receptor paralogs are present at these stages and before the maternal-zygotic transition, suggesting that they are maternally deposited (Figure 1C). Expression of the identified receptors persists throughout larval development up to 4 days post-fertilization (dpf). The only receptor-encoding gene that does not seem to be expressed during early development is the Type I receptor homolog acvr1c, which is first detected at 4 dpf (Figure 1C). Therefore, all putative receptors except for acvr1c are expressed at the developmental stages, during which Nodal signaling patterns the germ layers. We next used in situ hybridization analysis to characterize the spatial expression patterns of the putative receptors, and in particular to determine whether they are expressed at the embryonic margin, where Nodal signaling induces mesendoderm (Figure 1D). In agreement with the temporal analysis (Figure 1C), we found that transcripts of all putative Nodal Type I and II receptors – with the exception of acvr1c – are evenly distributed at the two-cell stage (Figure 1D), consistent with maternal deposition. During early gastrulation (shield stage), most receptors are ubiquitously expressed throughout the embryo (Garg et al., 1999; Nagaso et al., 1999) – except for acvr1c, which is not expressed, and acvr1b-a, which is constrained to the embryonic margin (Figure 1D), similar to the co-receptor oep (Renucci et al., 1996; Vopalensky et al., 2018). Together, our analyses show that, except for acvr1c all putative receptors are expressed at the right time and place to potentially act as mediators of endogenous Nodal signaling during zebrafish germ layer patterning. Nodal signaling upregulates acvr1b-a expression but does not affect other putative Nodal receptors In zebrafish, Nodal signaling induces several of its own signaling pathway components, including squint, cyclops, lefty1, lefty2, and oep (Bennett et al., 2007; Dubrulle et al., 2015; Feldman et al., 2002; Meno et al., 1999). To systematically assess potential receptor induction by Nodal signaling, we used qRT-PCR to measure receptor expression levels in embryos with increased Nodal signaling (injection of squint-GFP or cyclops-GFP mRNA [Müller et al., 2012]) or decreased Nodal signaling (injection of lefty2-Dendra2 mRNA [Müller et al., 2012] or treatment with the Nodal inhibitor SB-505124 [DaCosta Byfield et al., 2004]). acvr1c was excluded from this analysis because its spatiotemporal expression suggests that it does not mediate endogenous Nodal signaling during germ layer patterning (Figure 1D). In agreement with previous studies (Dubrulle et al., 2015), oep and acvr1b-a were upregulated by increased Nodal signaling and downregulated by decreased signaling (Figure 1E). Upon Nodal overexpression, acvr1b-a expression expanded beyond its usual domain at the margin, whereas Nodal inhibition abolished its expression (Figure 1F). In contrast, none of the other putative Nodal receptor-encoding genes exhibited a substantial change in expression upon Nodal overexpression or inhibition (Figure 1E). The Acvr2 receptors act in part redundantly to pattern the early zebrafish embryo through a partially Nodal-independent mechanism To elucidate the roles of the putative Nodal receptors in germ layer formation, we assessed the effect of their loss of function on embryonic morphologies, starting with the putative Type II receptors Acvr2a-a, Acvr2a-b, Acvr2b-a, and Acvr2b-b. These receptors are thought, and in the case of Acvr2b-a have been shown (Wang et al., 2016), to be bound by Nodals directly. To achieve receptor loss of function, we used three different approaches: (1) mutants, (2) morpholino-mediated knockdown (El-Brolosy et al., 2019; Rossi et al., 2015), and (3) CRISPR-mediated F0 knockouts (KO) (Kroll et al., 2021). To study receptor loss of function in mutants, we obtained acvr2a-aSA34654 and acvr2a-bSA18285 mutants from the European Zebrafish Resource Center (EZRC) and generated a mutant allele for acvr2b-a using CRISPR/Cas9. In all three mutants, the signal peptide or receptor domain and all downstream domains were disrupted, likely causing a complete loss of gene function (Figure 2—figure supplement 1A-D). However, surprisingly, none of the maternal-zygotic homozygous receptor mutants displayed obvious patterning defects at 1 dpf (Figure 2—figure supplement 1A-D) and all were viable. This is in stark contrast to the mouse Type IIB receptor mutants that exhibit severe malformations during early embryonic development (Gu et al., 1998; Oh and Li, 1997). Genetic compensation can mask potential patterning defects in mutants (El-Brolosy et al., 2019; Rossi et al., 2015) and may occur in some instances in the acvr2 mutants (Figure 2—figure supplement 1E, F). Therefore, we also assessed the effect of acutely knocking down gene activity using antisense morpholino oligonucleotides targeting the ATG start codons or splice sites of the putative receptor mRNAs. We found that morpholinos targeting individual receptor-encoding mRNAs had no effects or caused non-specific head or tail defects similar to a standard control morpholino (Figure 2—figure supplement 1G-L). Some morpholino treatments at high doses increased lethality (Figure 2—figure supplement 1H-L), but none of the conditions led to the Nodal-specific patterning defects observed in loss-of-function mutants of other Nodal signaling pathway components (Dubrulle et al., 2015; Feldman et al., 1998; Gritsman et al., 1999). To complement our mutant and morpholino analyses, we acutely knocked down the putative Type II Nodal receptors using CRISPR-mediated F0 KO (Kroll et al., 2021). In this approach, multiple exons of a single gene are targeted by Cas9/gRNA ribonucleoproteins (RNPs), thereby generating a biallelic zygotic KO directly in the injected embryos, the so-called F0 generation. Using this technique, we found that F0 KO of acvr2a-a, acvr2a-b, acvr2b-a, or acvr2b-b did not affect embryo morphology (Figure 2B), confirming that the disruption of individual Acvr2 receptors has no substantial influence on embryonic patterning. Figure 2 with 3 supplements see all Download asset Open asset Combinatorial removal of putative Type II Nodal receptors causes dorsal-ventral patterning defects. (A–B) Phenotypes of embryos upon single, double, triple, and quadruple CRISPR F0 KO of acvr2 receptors. (A) Lateral view of embryos of the indicated condition approximately 28–31 hpf. Arrowheads indicate the extent of ventral fin loss. Boxes indicate the phenotype class to the in Scale bar represents 250 µm. (B) of phenotypes observed in embryos of the indicated condition at 1 the of embryos. (C) Nodal and signaling by pSmad2/3 and respectively, in and quadruple acvr2 F0 KO embryos at and show with to the right. The of embryos with the phenotype is Scale bar represents µm. See the Figure data 1 file for source Figure data 1 Source data for Figure Download like zebrafish have an genome the two of and and of paralogs can the of phenotypes in single mutants et al., 1998; et al., 2019). To whether the putative Nodal Type II receptors function we used the CRISPR F0 KO to multiple acvr2 thereby combinatorial triple, and quadruple F0 KO embryos were for patterning defects. In contrast to the Nodal-specific loss-of-function embryos of all KO two (Figure However, with an of acvr2 genes disrupted, embryos displayed more severe phenotypes (Figure and KO embryos had phenotypes with the ventral tail fin partly or similar to class 1 mutant phenotypes et al., 1997). F0 KO embryos additionally showed a or tail and a yolk (Figure and for or all four acvr2 genes showed the most severe patterning including a or the absence of fin a larger yolk with yolk and head of head (Figure and and lethality at These phenotypes were in the of quadruple F0 KO embryos, and although they typical features of like a tail upon loss of signaling et al., 1997; and other like the in head and eye do not the classical To that the observed phenotypes are to the loss of the Acvr2 receptors, we additionally the acvr2 genes by quadruple morpholino-mediated knockdown and zygotic F0 KO of acvr2b-b in the of a mutant The resulting embryos were for patterning defects at 1 dpf in the case of the mutant approaches the phenotypes observed for the quadruple F0 acvr2 KO (Figure 2—figure supplement the mutants the difference between the acvr2 genes observed in the combinatorial F0 While and acvr2a-b, on their were not able to the loss of the other three receptors, the most severe phenotypes were only observed in the absence of acvr2b-a or acvr2b-b (Figure and Figure 2—figure supplement To directly determine potential of combinatorial acvr2 loss on Nodal and signaling, we performed for the Nodal signal transducer and the signal transducer in and quadruple F0 KO embryos at early and gastrulation stages and At the pSmad2/3 signal was in quadruple F0 KO embryos, the signal (Figure the of gastrulation quadruple F0 KO embryos showed an increased pSmad2/3 signal compared to embryos, but in addition a signal (Figure This that in the absence of the Acvr2 receptors Nodal signaling with some signaling To further whether the Acvr2 receptors act the pathway as we performed quadruple receptor F0 KO in mutant embryos and observed their phenotypes at 1 In addition to including these embryos also had a body and (Figure 2—figure supplement at least partially patterning defects. our show that the zebrafish receptors Acvr2a-a, Acvr2a-b, Acvr2b-a, and Acvr2b-b act partly redundantly and are essential for the formation of the embryonic body However, they do not to the function of classical Type II Nodal receptors. It will be in the to determine the mechanism by which the Type II receptors mediate partially Nodal-independent patterning The Type I receptors Acvr1b-a and Acvr1b-b redundantly mediate Nodal signaling during zebrafish germ layer patterning We next the role of the putative Type I Nodal receptors Acvr1b-a, Acvr1b-b, and Acvr1c in early zebrafish development. We first the phenotypes of single receptor loss of function using mutants, morpholino-mediated knockdown and CRISPR F0 KO as to the Acvr2 receptors, loss of individual Acvr1 receptors did not cause severe phenotypes from the non-specific head or tail defects observed for some morpholino treatments (Figure and Figure supplement and Figure 3 with 3 supplements see all Download asset Open asset Combinatorial removal of putative Type I Nodal receptors causes Nodal-specific patterning defects. Phenotypes of and oep CRISPR F0 KO embryos compared to embryos of or both acvr1b-a and acvr1b-b using morpholino CRISPR F0 and (A) Lateral view of embryos of the indicated condition approximately 28–31 hpf. point to a single to or Boxes indicate the phenotype class to the in Scale bar represents 250 µm. (B) Nodal signaling by pSmad2/3 in embryos of the indicated condition (A) at show The of embryos with the phenotype is Scale bar represents µm. (C) of phenotypes observed in embryos of the indicated condition 1 the of embryos. that one of the used for acvr1b-b F0 KO has acvr1b-a as a likely the of the phenotype (see Materials and See the Figure data 1 file for source Figure data 1 Source data for Figure Download To whether the putative Type I receptors may redundantly mediate Nodal signaling during germ layer patterning, we used combinatorial receptor loss-of-function approaches for acvr1b-a and knockdown of acvr1b-a and acvr1b-b in a loss of head mesoderm at 1 leading to the phenotype and a body with signaling (Figure supplement However, in the trunk suggesting an loss of Nodal signaling due to maternal of receptor proteins or mRNA but phenotypes were obtained with the acvr1b-a and acvr1b-b F0 Most embryos two but showed between them the zygotic F0 KO of oep (Figure and a maternal Therefore, we injected morpholinos into maternal-zygotic mutant embryos. This combination the
Source: eLife
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: gold
Estimated APC paid: $2,500
OpenAlex ID: W4312947117