August 1, 2001;
Timing of endogenous activin-like signals and regional specification of the Xenopus embryo.
Signaling by activin-like ligands is important for induction and patterning of mesoderm
. We have used an antibody that specifically recognizes the phosphorylated and activated form of Smad2
, an intracellular transducer of activin-like ligands, to examine how this signaling pathway patterns the early mesendoderm
. In contrast to the simple expectation that activin-like signaling should be highest on the dorsal side of the gastrula stage embryo
, we have found that while Smad2
phosphorylation is highest dorsally before gastrulation, signaling is attenuated dorsally and is highest on the ventral
side by mid-gastrulation. Early dorsal initiation of Smad2
phosphorylation results from cooperation between the vegetally localized maternal transcription factor VegT
and dorsally localized beta-catenin. The subsequent ventral
appearance of Smad2
phosphorylation is dependent on VegT
, but not on signaling from the dorsal side. Dorsal attenuation of Smad2
phosphorylation during gastrulation is mediated by early dorsal expression of feedback inhibitors of activin-like signals. In addition to regulation of Smad2
phosphorylation by the expression of activin-like ligands and their antagonists, the responsiveness of embryonic cells to activin-like ligands is also temporally regulated. Ectopic Vg1
all fail to activate Smad2
phosphorylation until after the midblastula transition, and the onset of responsiveness to these ligands is independent of transcription. Furthermore, the timing of cellular responsiveness differs for Xnr1
, and these distinct temporal patterns of responsiveness can be correlated with their distinctive phenotypic effects. These observations suggest that the timing of endogenous activin-like signaling is a determinant of patterning in the early Xenopus embryo
[+] show captions
VegT and activin-like ligands induce transcription-dependent positive- and negative-feedback loops of Smad2 activation. (A) Maternal VegT is required for endogenous phosphorylation of Smad2. Western blot analysis indicates that injection into oocytes of antisense oligonucleotides specific for maternal VegT (AS-VegT) eliminates Smad2 phosphorylation at stage 10+. Phosphorylation of Smad2 is rescued by co-injection of VegT (300 pg) or activin (3 pg) RNA into oocytes. Levels of Smad2 are unaffected. Actin levels serve as loading controls. (B) Maternal VegT and zygotic VegT are sufficient for Smad2 phosphorylation in animal caps. Embryos were injected into the animal region with maternal VegT (mVegT, 200 pg/embryo) or zygotic VegT (zVegT, 200 pg/embryo) RNA with or without α-amanitin at the two-cell stage, dissected at stage 9.5 and harvested for western blot analysis at stage 10.25. Smad2 and actin levels serve as loading controls. (C) Ectopic activin-like ligands induce Smad2 phosphorylation and zygotic VegT expression in animal caps. Ligand RNA (100 pg/embryo) was injected into the animal region of two cell stage embryos; animal caps were dissected at stage 9.5 and harvested at stage 10. Top: Western blot analysis of Smad2 phosphorylation. Smad2 and actin levels serve as loading controls. Bottom: Analysis of zygotic VegT expression by RT-PCR. Levels of ODC serve as loading controls. RT indicates whole embryo RNA that was (+) or was not (−) reverse transcribed. (D) Antivin and cerberus are inhibitors of activin-like signaling. Patterns of phosphoSmad2 or phosphoSmad1 from stage 9.25, 10+, 10.5 and 11.5 embryos were examined endogenously or in the presence of antivin (1 ng/embryo) or cerberus (1 ng/embryo) RNA injected supramarginally and submarginally at the two-cell stage. Actin levels serve as loading controls. (E) Attenuation of Smad2 phosphorylation is blocked by repression of FAST-regulated transcription. Fast-En RNA (F-En, 800 pg/embryo) was injected marginally and vegetally into two-cell stage embryos; embryos were harvested at stages 10+, 10.5 and 11.5. Top: Western blot analysis of Smad2 phosphorylation. Smad2 and actin levels serve as loading controls. Bottom: RT-PCR analysis of expression of Xnr1, derrière, antivin and cerberus. ODC levels serve as loading controls. −RT indicates whole embryo RNA that was not reverse transcribed.
Endogenous Smad2 activation is initiated and attenuated from the dorsal side of the Xenopus embryo. (A) Time course of early development. Developmental stages are correlated with hours post-fertilization at 23°C. Blastula, gastrula and neurula stages are indicated. Most developmental stages are from Nieuwkoop and Faber (Nieuwkoop and Faber, 1967). Intermediate stages are expressed as fractions between established stages (as stage 9.25). (B-M) Immunohistochemical analysis of phosphoSmad2 in the early Xenopus embryo. Embryos fixed at the indicated stages were bisected and then processed for immunohistochemistry with the anti-phosphoSmad2 antibody. Orientations of the axes of dissection are indicated (B,C) (Nieuwkoop and Faber, 1967). Representative sections are shown for stages 9.5 (D,E), 9.75 (F,G), 10+ (H,I), 10.5 (J,K) and 11.5 (L,M). Sectioned embryos were photographed en face. Ventral is towards the left; dorsal is towards the right. Animal is towards the top; vegetal is towards the bottom. For cross sections (B,D,F,H,J,L), fixed embryos were bisected through the dorsal-ventral midline. For longitudinal sections (C,E,G,I,K,M), fixed embryos were bisected in a horizontal plane near the equator and below the blastocoel floor; lower sections are shown. Approximate levels and angles of longitudinal sections (estimated from landmarks and the shape of the sections) are indicated with arrows on cross sections.
Dorsal-ventral prepattern directs region-autonomous temporal patterns of endogenous activin-like signaling. (A-C) Phenotype of explants and embryo at stage 30. Embryos were bisected into dorsal and ventral halves at stage 8. Explants were cultured to stage 30. Representative ventral half (A), dorsal half (B) and whole (C) embryos are shown. (D) Endogenous Smad2 phosphorylation is autonomous and distinct in isolated dorsal and ventral halves. Embryos were bisected into dorsal and ventral halves at stage 8. Cultured ventral halves, dorsal halves, and whole embryo controls were harvested at each of the indicated stages for western blot analysis. Each lane contains the same amount of embryonic material; hence, lanes containing whole embryo material should reflect the average amount of signal from ventral and dorsal halves. In general, whole embryos contain more signal, presumably owing to cell loss upon healing after dissection. (E) Autonomous expression of activin-like ligands and inhibitors is distinct in dorsal and ventral halves. Bisected embryos prepared as in D were harvested for RT-PCR analysis of expression of Xnr1, derrière, antivin and cerberus. EF1α levels serve as loading controls.
β-Catenin alters timing of Smad2 activation. (A-D) Cortical rotation is required for correct timing of initiation and attenuation of endogenous Smad2 phosphorylation in whole embryos. (A) Phenotypes of representative wild-type, u.v.-treated and u.v.-treated/β-catenin-rescued embryos at stage 40. (B) Phosphorylation of Smad2 and (C) phosphorylation of Smad1 were examined by western blot analysis from stage 8.5 to stage 12.5 in wild-type, u.v.-treated and u.v.-treated/β-catenin-rescued embryos. (D) Cytoskeletal actin is a loading control. (E) β-catenin accelerates timing of VegT-induced Smad2 phosphorylation in animal caps. Two-cell stage embryos were injected into the animal region with either β-catenin (50 pg/embryo) or VegT (50 pg/embryo) RNA or both. Animal caps were dissected between stages 8 and 9, and harvested from stage 9 to stage 11.5; whole embryos collected at these stages serve as controls for endogenous phosphorylation of Smad2. Actin levels serve as loading controls.
Activin-like ligands induce distinct temporal profiles of Smad2 activation. (A) Zygotic activin-like ligands induce Smad2 phosphorylation after, but not before, MBT. Embryos injected marginally at the two cell stage with ligand or ALK4* RNA (100 pg/embryo) were harvested at stage 6 (before MBT) or at stage 10 (after MBT) for western blot analysis. Uninjected embryos provide comparison with endogenous Smad2 phosphorylation. Levels of Smad2 and actin serve as loading controls. (B) Zygotic activin-like ligands induce Smad2 phosphorylation after MBT without requirement for zygotic transcription. Embryos injected marginally at the two-cell stage with ligand or ALK4* RNA (100 pg/embryo) alone or with α-amanitin injected vegetally at the four-cell stage were harvested at stage 10+ for western blot analysis. Uninjected embryos provide comparison with endogenous Smad2 phosphorylation. Levels of Smad2 and actin serve as loading controls. (C) Timing of ligand-induced Smad2 phosphorylation is not altered by inhibition of zygotic transcription. Patterns of Xnr1-induced Smad2 phosphorylation were compared with endogenous patterns from stages 6 to 10 in the presence or absence of α-amanitin. Xnr1 RNA (100 pg/embryo) was injected marginally at the two-cell stage; α-amanitin was injected vegetally at the four-cell stage. (D) Responsiveness for Smad2 activation is transcription independent and ligand specific. Patterns of Smad2 phosphorylation induced by activin, Xnr1 or derrière in the presence of α-amanitin were compared with each other and to the endogenous pattern without α-amanitin from stages 6 to 10. Ligand RNA (100 pg/embryo) was injected marginally at the two-cell stage; α-amanitin was injected vegetally at the four-cell stage. (E) Timing of responsiveness to ligands is not altered by increased dose. Patterns of Smad2 phosphorylation induced by Xnr1 or derrière at 100 pg/embryo or 300 pg/embryo in the presence of α-amanitin were compared with each other and with the endogenous pattern without α-amanitin from stages 6 to 10. Ligand RNA was injected marginally at the two-cell stage; α-amanitin was injected vegetally at the four-cell stage.
Dorsal β-catenin modifies timing of Smad2 phosphorylation in early developmental patterning of the Xenopus embryo. In the wild-type embryo, early initiation of Smad2 phosphorylation at stage 9.5 in the dorsal vegetal region results from cooperation between dorsal β-catenin and vegetal VegT. At stage 10.5, attenuation of Smad2 phosphorylation on the dorsal side is mediated by expression of negative feedback inhibitors (see text). In u.v.-treated embryos, Smad2 phosphorylation appears later and persists longer than in wild-type embryos. Maximal levels of Smad2 phosphorylation, however, are not affected by perturbation of the location of endogenous β-catenin, indicating that dorsal β-catenin determines the temporal patter, not the maximal level, of endogenous activin-like signaling.