A role for Vg1/Nodal signaling in specification of the intermediate mesoderm

Abstract

The intermediate mesoderm (IM) is the embryonic source of all kidney tissue in vertebrates. The factors that regulate the formation of the IM are not yet well understood. Through investigations in the chick embryo, the current study identifies and characterizes Vg1/Nodal signaling (henceforth referred to as 'Nodal-like signaling') as a novel regulator of IM formation. Excess Nodal-like signaling at gastrulation stages resulted in expansion of the IM at the expense of the adjacent paraxial mesoderm, whereas inhibition of Nodal-like signaling caused repression of IM gene expression. IM formation was sensitive to levels of the Nodal-like pathway co-receptor Cripto and was inhibited by a truncated form of the secreted molecule cerberus, which specifically blocks Nodal, indicating that the observed effects are specific to the Nodal-like branch of the TGFβ signaling pathway. The IM-promoting effects of Nodal-like signaling were distinct from the known effects of this pathway on mesoderm formation and left-right patterning, a finding that can be attributed to specific time windows for the activities of these Nodal-like functions. Finally, a link was observed between Nodal-like and BMP signaling in the induction of IM. Activation of IM genes by Nodal-like signaling required an active BMP signaling pathway, and Nodal-like signals induced phosphorylation of Smad1/5/8, which is normally associated with activation of BMP signaling pathways. We postulate that Nodal-like signaling regulates IM formation by modulating the IM-inducing effects of BMP signaling.

Fig. 1.

Vg1 induces expansion of the intermediate mesoderm. COS cells transfected with Vg1 (A-C,E-G,I-K,M-O) or control pMES (D,H,L,P) were implanted into stage 4 chick embryos adjacent to the primitive streak at 20% streak length. COS pellets were marked with DiI (A,E,I,M) to identify their location. Diagram at upper left indicates position of cell pellets at the start of the experiment. Embryos are oriented with anterior towards the left, whereas the diagram is oriented with the anterior end up. Embryos were grown for 48 hours and analyzed by in situ hybridization for Osr1 (A-D), Pax2 (E-H), Lim1 (I-L) or Paraxis (M-P). The red lines crossing the embryos in B, F, J and N indicate approximate planes of section shown in C, G, K and O, respectively. Locations of implants are indicated by arrows in whole embryos (A,E,I,M) and by dashed lines in sections (D,G,H,K,L,O,P). Dashed arrows denote expansion of IM markers into the somite compartment (C,G,K) or reduction in somite size and gene expression (O). nt, neural tube; som, somite.

Fig. 2.

Vg1 alone is sufficient to expand the IM. Electroporation was used to introduce a Vg1-expressing plasmid (A-D) or a control empty plasmid (E-H) into stage 3 chick embryos, targeted to 20% streak length. Diagram at upper left indicates location targeted by electroporation. Embryos were grown for 48 hours and stained with DAPI (A,E), and with antibodies to Pax2 (B,F) and GFP (C,G). Merged channels of B,C (D) and F,G (H) are shown. In Vg1-electroporated embryos, Pax2 expression reached the midline at the expense of the posterior somites and notochord, which did not develop properly (B,D). The control embryo electroporated with pMES alone did not induce ectopic Pax2 expression (F,H). n, notochord; nt, neural tube; som, somite.

Fig. 3.

Activation of IM markers by the Alk4 receptor is non-cell-autonomous. Electroporation was used to introduce a constitutively active Alk4-expressing plasmid (A-E) or a control empty plasmid (F-J) into stage 3 chick embryos, as in Fig. 2. Embryos were grown for 48 hours and stained with DAPI to mark nuclei (B,G), with antibodies to Pax2 (C,H) and with antibodies to GFP (D,I). Merged channels of C,D (E) and H,I (J) are shown. The red lines crossing the embryos (A,F) indicate approximate planes of section shown in B-E and G-J, respectively. The differences in GFP distribution in the caAlk4-electroporated embryo (A,D) and the control pMES electroporated embryo (F,I) suggested that somites were not capable of expressing high levels of caAlk4, resulting in groups of GFP-expressing cells being found mostly outside the somites, both medially and laterally (A, green arrows in D,E). In the caAlk4-electroporated embryo, ectopic expression of Pax2 was found in areas outside the IM, including the somites and groups of cells adjacent to the neural tube (red arrows in C,E). Cells expressing high levels of GFP did not express Pax2, but Pax2 was expressed in neighboring cells (red arrows in E), indicating non-cell-autonomous activation of Pax2 by caAlk4. Control embryos electroporated with pMES alone did not induce ectopic Pax2 expression (H,J). d, nephric duct; n, notochord; nt, neural tube; som, somite.

Fig. 4.

Inhibition of Nodal-like signaling prevents IM gene expression. (A-F) Lefty2 reduces expression of IM markers. COS cells transfected with Lefty2 (A,C,E) or control pMES (B,D,F) were implanted adjacent to the primitive streak of stage 4 chick embryos at 30-40% streak length (prospective IM). Embryos were grown for 48 hours and analyzed by in situ hybridization for Pax2. The dashed lines crossing the embryos in A and B indicate approximate planes of section shown in C,E and D,F, respectively. Locations of implants are indicated by arrows (A-D). COS-Lefty2 cells produced diminished expression of Pax2 on the transplanted side. (G-J) Inhibition of IM gene expression by cerberus-short. Stage 4 embryos were implanted with COS cells expressing cerberus-short (G,I) or control COS cells (H,J) and analyzed after 48 hours for expression of Pax2. Cerberus-short cells induced significant reduction of Pax2 expression on the implanted side (G,I). The red arrows in G and H indicate pellet locations at the time of fixation. The dashed white lines in G and H indicate the plane of section in I and J, respectively.

Fig. 5.

Osr1 expression requires Alk4/5/7 signaling. (A-D) Explants from the mid-primitive streak of stage 5 embryos were cultured with (C,D) or without (A,B) the Alk4/5/7 inhibitor SB431542 (50 ng/ml) and evaluated by in situ hybridization for expression of the IM marker Osr1 (B,D) or the PM marker Paraxis (A,C). Treatment with SB431542 resulted in a strong inhibition of Osr1 expression (D), whereas Paraxis expression was relatively unaffected (C). Note that the ‘IM’ explants express Paraxis because at the primitive streak stage the IM and PM domains are not well delineated, so explants of the mid-streak region contain precursors to both IM and PM.

Fig. 6.

Effects of addition and inhibition of the Nodal/Vg1 co-receptor Cripto on IM gene expression. (A-F) Increasing Cripto levels causes expansion of early IM markers. Electroporation was used to introduce Cripto, a Nodal/Vg1 co-receptor (A-C), or a control empty plasmid (D-F) into stage 3 chick embryos, targeted to the PM and IM regions (25-30% streak length). Embryos were grown for 48 hours and analyzed by in situ hybridization for Pax2 (B,C,E,F). GFP expression in A,D indicates electroporated areas. The red lines crossing the embryos in B and E indicate approximate planes of section shown in C and F, respectively. Pax2 staining was significantly increased in Cripto-electroporated embryos (B,C). Asterisk marks a expanded IM with strong Pax2 expression (C) compared with the control (F). (G-L) Inhibition of Cripto reduces expression of early IM markers. Electroporation was used to introduce tomoregulin-1, an inhibitor of the co-receptor Cripto (G-I), or a control empty plasmid (J-L) into the prospective IM region (30-40% streak length) of stage 3 embryos. Embryos were grown for 48 hours and analyzed by in situ hybridization for Osr1 (H,I,K,L). GFP expression in G and J indicates electroporated areas. The red lines crossing the embryos in H and K indicate approximate planes of section shown in I and L, respectively. Weak and patchy staining was observed in tomoregulin-1-electroporated embryos (H,I) compared with the control embryo (K,L). nt, neural tube; som, somite.

Fig. 7.

Separate time windows for the IM-inducing and left-right patterning effects of Nodal-like signaling. COS cells expressing Vg1 (B) but not control COS cells (A) induce right-sided Nodal-expression if placed on the right side of the embryo at stage 7. COS-Vg1 implanted at stage 7 on the right side in a similar location as in B did not affect the expression of Pax2 (C), nor did COS cells expressing cerberus-short implanted on the left side (D). The insets in C and D show the location of the COS cell pellets (white arrows) at the start of the experiment, and the red arrows indicate pellet locations at the end of the experiment.

Fig. 8.

Interactions of Nodal-like and BMP signaling in the regulation of IM gene expression. (A-F) Induction of Osr1 by activin requires BMP signaling. Explants of stage 4 anterior-primitive streak (A) were grown in culture in control medium (B) or in medium supplemented with 10 ng/ml activin (C), 10 ng/ml activin and 1 μg/ml Noggin (D), 50 ng/ml BMP2 (E) or 50 ng/ml BMP2 and 50 ng/ml SB431542 (F) and stained by in situ hybridization for Osr1. Note strong inhibition of the Osr1-inducing effects of activin by Noggin (D), and partial inhibition of the Osr1-inducing effects of BMP2 by SB431542 (F). (G-J) Vg1 induces Smad1/5/8 phosphorylation. COS cells expressing Vg1 (G,H) or control COS cells (I,J) were placed in early stage 5 embryos, grown overnight, and analyzed at stage 9 by immunofluorescence for presence of phosphorylated Smad1/5/8. COS-Vg1 cells (G,H) but not control cells (I,J) induced a rim of pSmad1/5/8 adjacent to the cell pellet (arrows). G and I are wide-field microscope views, and H and J are confocal images.

Fig. 9.

Model of the effects of Nodal/Vg1 signaling on IM patterning. (A) Summary of the expression patterns of Nodal-like signals, inhibitors and co-receptors in chick stage 4 embryos. At stage 4 (mid-gastrula), Nodal and Vg1 are expressed throughout the primitive streak with highest concentrations in the mid-streak (pink); the co-receptor Cripto is expressed primarily in the anterior part of the streak (diagonal purple lines); and the Nodal-like inhibitor Lefty is expressed in a gradient with strongest expression at the anterior end of the streak (black, with width of the black domain indicative of expression level). PM, IM and LP indicate the region of the primitive streak containing prospective paraxial mesoderm, intermediate mesoderm, and lateral plate mesoderm, respectively. Whereas both the prospective PM and IM express Nodal-like signals and Cripto co-receptor, the Lefty1 inhibitor is expressed at higher levels in the prospective PM, consistent with a model in which effective Nodal-like signaling is highest in the prospective IM region. (B) Comparison of the effects of BMP and Nodal-like signaling on mesodermal patterning. The left-hand diagram depicts a stage 10 chick embryo. The right-hand diagrams summarize the effects of addition of a pellet of control cells (top), cells secreting BMP (middle) or cells secreting Vg1 (bottom). The dashed box in the left-hand figure shows the area that is expanded in the diagrams on the right. Excess BMP produced a medial shift and contraction of both PM and IM, with a corresponding expansion of the LP. Excess Vg1 produced an expansion of the IM at the expense of the PM, but the LP was unaffected. See text for further discussion. (C) Illustration of how the observation that Nodal-like signaling induces pSmad1 could explain the results summarized in panel B. The y-axis denotes levels of BMP signaling, with ‘a’ denoting the threshold of BMP signaling that distinguishes PM from IM, and ‘b’ denoting the threshold of BMP signaling that distinguishes IM from LP. Addition of a BMP source (middle) results in a broad elevation of BMP signaling centered around the source, with a resulting shifting of the PM-IM and the IM-LP borders medially and reduction in the size of the IM and PM. Addition of a Vg1 source (bottom) results in a localized rise in BMP signaling, which moves the PM-IM border medially but does not affect the IM-LP border, resulting in a broadened IM. Black dotted vertical lines in the middle and bottom panels indicate the borders between PM, IM and LP in control embryos (as taken from the top panel).