Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2009 Mar 1.
Published in final edited form as: Dev Biol. 2008 Jan 3;315(1):203–216. doi: 10.1016/j.ydbio.2007.12.031

TGF-β signaling is required for multiple processes during Xenopus tail regeneration

Diana M Ho 1, Malcolm Whitman 1
PMCID: PMC2292344  NIHMSID: NIHMS42557  PMID: 18234181

Abstract

Xenopus tadpoles can fully regenerate all major tissue types following tail amputation. TGF-β signaling plays essential roles in growth, repair, specification, and differentiation of tissues throughout development and adulthood. We examined the localization of key components of the TGF-β signaling pathway during regeneration and characterized the effects of loss of TGF-β signaling on multiple regenerative events. Phosphorylated Smad2 (p-Smad2) is initially restricted to the p63+ basal layer of the regenerative epithelium shortly after amputation, and is later found in multiple tissue types in the regeneration bud. TGF-β ligands are also upregulated throughout regeneration. Treatment of amputated tails with SB-431542, a specific and reversible inhibitor of TGF-β signaling, blocks tail regeneration at multiple points. Inhibition of TGF-β signaling immediately following tail amputation reversibly prevents formation of a wound epithelium over the future regeneration bud. Even brief inhibition immediately following amputation is sufficient, however, to irreversibly block the establishment of structures and cell types that characterize regenerating tissue and to prevent the proper activation of BMP and ERK signaling pathways. Inhibition of TGF-β signaling after regeneration has already commenced blocks cell proliferation in the regeneration bud. These data reveal several spatially and temporally distinct roles for TGF-β signaling during regeneration: 1) wound epithelium formation, 2) establishment of regeneration bud structures and signaling cascades, and 3) regulation of cell proliferation.

INTRODUCTION

The process of epimorphic regeneration involves the replacement of damaged, injured, or amputated tissues or structures with new and functionally equivalent tissues or structures. The frog Xenopus laevis can at tadpole stages regenerate the posterior half of its tail following experimental amputation; all of the complex structures of the tail, including neural tissue, notochord, vasculature, muscle, connective tissue, and skin can regenerate completely (Slack et al., 2004). Xenopus tadpoles provide an excellent model system for regeneration studies because they develop rapidly (~3 days after fertilization), can be amputated in large numbers with high and reproducible rates of regeneration, and can be kept in small and non-circulating volumes, making chemical perturbations feasible.

The tadpole tail regenerates completely over a period of about 1–2 weeks. Within about 24–48 hours, regenerative structures can already be clearly observed. Following wound epithelium formation, a regeneration bud is formed, which contains regenerative neural and notochord tissues as well as a blastema of undifferentiated mesenchymal cells including at least one stem cell type, muscle satellite cells (Chen et al., 2006; Slack et al., 2004). Later, cells in the regeneration bud undergo cell proliferation and differentiation to generate new tissues; for example, satellite cells in the blastema differentiate into mature muscle fibers (Chen et al., 2006; Gargioli and Slack, 2004). A common theme in Xenopus tail regeneration is the re-expression of genes and re-activation of signaling pathways that are active in the embryonic tailbud, which acts as a molecular organizer for posterior structures during development (Beck et al., 2003; Sugiura et al., 2004). Several signaling cascades, such as the FGF and BMP pathways, have been implicated in both tailbud patterning and tail regeneration (Beck et al., 2006; Beck et al., 2003). While these pathways are clearly necessary for regeneration, the specific events they regulate during the regenerative process have not been identified. A significant limitation to defining the role of signaling pathways in regeneration has been the temporal resolution with which pathway inhibition can be achieved; traditional genetic or transgenic approaches are difficult to regulate on a time scale fine enough to distinguish early steps in the regenerative process.

TGF-β signaling is essential for numerous processes of growth, repair, specification, and differentiation. Canonical TGF-β as well as activin/nodal-like ligands bind to two serine-threonine kinase receptors, designated Type I and Type II, resulting in the phosphorylation of the Type I receptor by the Type II receptor. This leads to phosphorylation of the signal transducer Smad2/3, which then translocates to the nucleus and interacts with transcription factors to regulate downstream gene expression (Shi and Massague, 2003; Whitman, 1998). The Type I receptors which are responsible for mediating TGF-β signaling, as distinct from BMP signaling, are Alk4, Alk5, and Alk7 (Carcamo et al., 1994; Ryden et al., 1996).

A chemical inhibitor of the TGF-β signaling pathway, SB-431542, rapidly, specifically, and potently inhibits Alk4/5/7 but not other related receptors such as BMP receptors (Inman et al., 2002). We have previously reported that SB-431542 functions specifically and reversibly in Xenopus and zebrafish embryos to inhibit Alk4/5/7 in vivo (Ho et al., 2006). Upon treatment with SB-431542, both developmental phenotype and Smad2 phosphorylation are severely disrupted, phenocopying mutations in TGF-β signaling components; these defects can be fully rescued by introduction of inhibitor-insensitive Alk4, demonstrating that SB-431542 acts specifically to block the TGF-β signaling pathway during embryogenesis (Ho et al., 2006). The use of this highly specific and well-characterized TGF-β inhibitor provides a number of advantages over more traditional genetic manipulations: a) it is easily applied to large numbers of tadpoles with highly reproducible effects, b) it can be added and removed at specific time points, allowing for the individual dissection of multiple TGF-β-dependent events within complex processes such as regeneration, and c) it sidesteps the problem of early developmental defects or lethality.

Since TGF-β signals are active during normal tail development (Kondaiah et al., 2000); D.M.H. and M.W., unpublished), we examined the localization of components of the TGF-β signaling pathway during tail regeneration in Xenopus tadpoles and found that phosphorylated Smad2 (p-Smad2) as well as TGF-β ligands were present throughout regeneration. At a macroscopic level, inhibition of TGF-β signaling causes failure of tail regeneration at several timepoints after amputation. On a cellular level, we have determined that TGF-β signaling has distinct roles in wound epidermis formation, establishment of regenerative structures, and cell proliferation. We also found that TGF-β signaling is required for the proper activation of BMP and ERK signaling pathways in the regeneration bud. These results reveal that the activation of the TGF-β signaling pathway is essential at multiple timepoints for establishing and maintaining regenerative competence.

MATERIALS AND METHODS

Regeneration experiments

All regeneration experiments were performed on Xenopus tadpoles at stages 40–43 (Nieuwkoop and Faber, 1967). Tadpoles were anesthetized in Tricaine (Sigma) and the posterior third to half of the tail was cut off using a micro-feather scalpel. All regenerates were grown at 20° C in 0.1X MMR with 50 units/ml penicillin/streptomycin (Invitrogen). SB-431542 (Tocris) was used at a final concentration of 100 μM (50 mM stock in DMSO), taking into account tadpole volume; SB-431542 was always premixed with the media to ensure solubilization. SB-505124 (Sigma) and LY-364947 (Sigma) were used at 25 μM and 100 μM respectively from stocks in DMSO. Nocodazole (Sigma) and hydroxyurea (Sigma) were used at 10 μg/ml and 30 mM respectively. For experiments in which SB-431542 or nocodazole was washed out, tadpoles were thoroughly rinsed off at least twice in large volumes of clean media. 0.2% DMSO was used throughout as a control.

For biotinylation experiments, live tadpoles were treated with 1 mg/ml sulfo-NHS-LC-biotin (Pierce) for 5′ at room temperature, followed by two brief rinses and two 5′ washes in 100 mM Tris, pH 7.6 to quench unbound biotin.

For BrdU incorporation assay, regenerates were treated with inhibitor for 6 hours prior to BrdU addition. BrdU (Roche) was then added to a final concentration of 0.2 mM.

Antibodies

Anti-p-Smad2 polyclonal antibody (Faure et al., 2000) was used at 1:10 –1:100. Monoclonal antibody to p63 (clone 4A4) was used at 1:50 (Yang et al., 1998). Monoclonal antibodies to Xen1 and Pax7 were used at 1:25. Tor70 monoclonal antibody was used at 1:20. Anti-p-Smad1 polyclonal antibody (Cell Signaling Technology) was used at 1:100, anti-phospho-Histone H3 polyclonal (Upstate) was used at 1:400, and anti-pERK1/2 monoclonal (Clone MAPK-YT, Sigma) was used at 1:200. HRP-conjugated secondary antibodies (Jackson ImmunoResearch) were used at 1:750. Secondary antibodies conjugated to AlexaFluor 488 or AlexaFluor 546 (Molecular Probes) were used at 1:500.

Staining and in situ hybridization

Samples for whole-mount immunostaining and in situ hybridization were fixed in MEMFA. Fixative for anti-p-Smad2 samples was supplemented with 5 mM iodoacetamide (Sigma). Standard protocols were used for immunostaining (Faure et al., 2000). HRP activity was detected using TrueBlue substrate (KLP). Biotinylation was detected using streptavidin-AlexaFluor 546 (1:500, Molecular Probes). Total DNA was labeled using TO-PRO-3 (Molecular Probes) or Hoechst (Molecular Probes).

A standard protocol for whole-mount in situ hybridization was used (Harland, 1991), with modifications as suggested by M. Levin. Antisense probes for xTGFβ2 and xTGFβ5 were obtained from I. Dawid (Kondaiah et al., 2000; Rebbert et al., 1990).

Samples for hematoxylin and eosin (H&E) staining were fixed in MEMFA, dehydrated in ethanol, embedded in Paraplast, and sectioned to 10 μm thickness. H&E staining was performed using a basic protocol by Roza Vazin (http://tropicalis.berkeley.edu/home/gene_expression/sections/H&E-protocol.html).

BrdU detection was performed using a modified protocol and reagents from the BrdU Labeling and Detection kit II (Roche).

Microscopy

All confocal microscopy was performed on a Nikon TE2000 microscope with C1 confocal. Light microscopy was performed using Leica MZFLIII and Leica DMIRB microscopes.

RESULTS

Phospho-Smad2 is upregulated throughout regeneration with changing localization

We used an antibody specific for phosphorylated Smad2 (p-Smad2) (Faure et al., 2000; Lee et al., 2001) for whole-mount immunostaining on regenerating tadpole tails to examine the activation of the TGF-β signaling pathway during regeneration. We find that p-Smad2 is upregulated throughout regeneration and is localized differentially at early and late timepoints. A freshly amputated tail (0 hpa) has no detectable p-Smad2 staining; however, nuclear p-Smad2 is present within 1 hour post-amputation (1 hpa) at the edges of the amputation site, where it continues to be expressed from 2–8 hpa (Figure 1A). This early p-Smad2 appears restricted to a single layer near the edge of the amputation site (Figure 1A). At 24 hpa, Smad2 phosphorylation is more widespread, and it extends throughout the entire regeneration bud region by 48–72 hpa (Figure 1A). A closer examination of very early timepoints reveals weak activation of p-Smad2 near the outer edges of the amputated fin as early as 15 minutes after amputation; this signal has extended medially by 30 minutes (Figure 1C).

FIGURE 1. Timecourse of phospho-Smad2 activation during tail regeneration.

FIGURE 1

Open in a new tab

(A) p-Smad2 staining of regenerates harvested at 0, 1, 2, 4, 8, and 24 hpa, and at 2 and 3 dpa (B) 4 hpa and 2 dpa regenerates treated with 100 μM SB-431542 at 0 hpa and 24 hpa respectively show no p-Smad2 staining. Phenotypic differences in SB-431542-treated regenerates are discussed later in this work (see Figure 4). Arrowheads: amputation plane. (C) Regenerates stained for p-Smad2 at 0′, 15′, 30′, and 1 h post-amputation. A higher-magnification view of the boxed region of each image is shown below it. Arrows point to representative p-Smad2-positive nuclei.

Regenerates treated with the TGF-β inhibitor SB-431542 do not show any staining, indicating that the nuclear signal we observe with the anti-p-Smad2 antibody is indeed specific (Figure 1B). We often observe staining around the edges of the somite; this signal is non-specific, as it is non-nuclear and insensitive to SB-431542 (Figure 1B and data not shown).

We also detect p-Smad3, an additional SB-431542-sensitive transducer of Alk4/5/7 signaling downstream of TGF-β, by Western blot (data not shown). Since p-Smad2 and p-Smad3 are coordinately regulated by Alk4/5/7 signaling, and available antibodies cannot reliably detect p-Smad3 by immunohistochemistry, we have focused on p-Smad2 as a marker of TGF-β signal activation during regeneration.

To identify the specific cell types in which p-Smad2 is present, we performed double immunofluorescence staining and confocal imaging on early (4 hpa), intermediate (24 hpa), and late (3 dpa) regenerates. At 4 hpa, all nuclei positive for p-Smad2 in the regeneration bud are also positive for the transcription factor p63 (white arrow, Figure 2A). p63 is a marker of the basal layer of the Xenopus epidermis (Lu et al., 2001). P-Smad2 is restricted to p63+ nuclei; notably, the apical layer of the skin (pink arrow, Figure 2A) expresses neither p-Smad2 nor p63. P-Smad2 is only expressed in the newly-formed wound epidermis of the regeneration bud; p63+ cells elsewhere in the tail do not express p-Smad2 (Figure 2B, white arrowhead). As expected, regenerates treated with SB-431542 do not express p-Smad2 even in p63+ cells (Figure 2A). By 24 hpa, p-Smad2 is no longer restricted to cells expressing p63 (Figure 2A).

FIGURE 2. Localization of phospho-Smad2 staining.

FIGURE 2

Open in a new tab

(A) Regenerates treated with 0.2% DMSO or 100 μM SB-431542 at 0 hpa and fixed at 4 or 24 hpa were double stained with anti-p-Smad2 (green) and anti-p63 (red). DNA is stained in blue. A high-power confocal section through the regeneration bud area is shown. Note total overlap between p-Smad2 and p63 at 4 hpa but not at 24 hpa. White arrow marks a representative nucleus in the p-Smad2+/p63+ layer. Pink arrow marks an apical epidermal cell that does not express p-Smad2. (B) Low power confocal section of the DMSO-treated regenerate in (A), showing that p63+ cells in the fin (white arrowhead) do not express p-Smad2. (C) 3 dpa regenerates, treated with 0.2% DMSO or 100 μM SB-431542 at 2 dpa, were stained for p-Smad2 (green), phospho-Tyrosine (pY, red), and DNA (blue). arrow: apical epithelial cell; arrowhead: basal epithelial cell; bl: blastema; nt: neural tube; not: notochord. Scale bar: 50 μm.

At 3 dpa, p-Smad2 is detected in all cells in the regeneration bud. Double staining for p-Smad2 and phospho-Tyrosine (pTyr, which labels cell boundaries) reveals that p-Smad2-positive nuclei are found in multiple cell types, including the blastema, neural ampulla, regenerating notochord, and both layers of the epidermis (Figure 2C).

TGF-β ligands are expressed in regenerating tails

RT-PCR analysis reveals that both xTGF-β2 and xTGF-β5, the two known TGF-β ligands in Xenopus, are present in uncut tails and upregulated during regeneration (Figure S1). Whole-mount in situ hybridization shows that xTGF-β2 message is first detected very weakly at 4 hpa, in the center of the regeneration bud, and expression continues there, growing stronger as regeneration proceeds (Figure 3A). This transcript appears to be localized to the mesenchyme in the regeneration bud region and is excluded from the overlying epidermal layer (Figure 3B). In contrast, xTGF-β5 expression is seen as early as 2 hpa, where it is located along the ventral edge of the regeneration bud (Figure 3C). By 24 hpa, it is found along the dorsal edge as well; expression continues along both edges through at least 72 hpa (Figure 3C). Like xTGF-β2, xTGF-β5 is in the regeneration bud proper, beneath the skin (Figure 3D).

FIGURE 3. Expression pattern of TGF-β ligands during tail regeneration.

FIGURE 3

Open in a new tab

(A) Whole-mount in situ hybridization with xTGF-β2 antisense probe on regenerating tails at 0 hpa, 4 hpa, 8 hpa, 24 hpa, 2 dpa, and 3 dpa. Red arrows point to xTGF-β2 expression (blue staining) in the regeneration bud. (B) Phase-contrast image at higher magnification of the 24 hpa tail from (A), showing xTGF-β2 expression (red arrow) in the regeneration bud proper and not in the overlying epithelium (black arrow). (C) Whole-mount in situ hybridization with xTGF-β5 antisense probe on regenerating tails at 0 hpa, 2 hpa, 4 hpa, 24 hpa, 2 dpa, and 3 dpa, showing expression of xTGF-β5 (red arrows) as early as 2 hpa. (D) Phase contrast image at higher magnification of the 2 dpa sample from (C), showing xTGF-β5 expression (red arrows) in the regeneration bud proper but not in the overlying epithelium (black arrow). Arrowheads: amputation plane.

RT-PCR also shows that the TGF-β family members xActivinβA and xGDF11 are both upregulated at early (4 hpa) and late (48 hpa) timepoints (Figure S1). The Alk5 homolog xTrR-I was present in both cut and uncut tails (data not shown).

Phenotypic effect of inhibition of TGF-β signaling during regeneration

To further examine the role of TGF-β signaling during regeneration, we used SB-431542 to block p-Smad2 during specific time periods post-amputation. Treatment with SB-431542 immediately following amputation resulted in a failure of the wound to heal properly. 100% of DMSO-treated controls had a clearly defined wound epidermis and protruding regeneration bud by 8 hpa (n=107; Table 1 and Figure 4A). In contrast, all amputated tadpoles treated with SB-431542 at 0 hpa failed to form a wound epidermis; accordingly, no regeneration bud was formed and cellular material began to leak from the wound in 100% of treated tadpoles (n=123; Table 1 and Figure 4A). This leakage was not caused by failure to form a blood clot, as both control and SB-treated tadpoles rapidly stop bleeding (Figure 4A and data not shown).

Table 1.

Wound healing phenotype in 8 hpa regenerates treated at 0 hpa.

healed unhealed # of experiments
100 μM SB-431542 @ 0 hpa 0/123 (0%) 123/123 (100%) 3
0.2% DMSO @ 0 hpa 107/107 (100%) 0/107 (0%) 3
Open in a new tab

FIGURE 4. Inhibition of TGF-β signaling perturbs tail regeneration.

FIGURE 4

Open in a new tab

(A) Amputated tadpoles were treated with 100 μM SB-431542 (bottom row) or 0.2% DMSO (top row) at 0 hpa, and photographed at 0 and 8 hpa. Note that an epithelial layer has formed in the DMSO control (arrow), but not in the SB-treated tail, where cellular material leaks from the wound (arrowhead). Tadpoles were stained in vivo with Nile Blue to improve resolution. (B) Regenerates were treated with 100 μM SB-431542 at 3 dpa and photographed at 6 dpa. Arrowheads mark the plane of amputation. (C) Regenerates were treated with 100μM SB-431542 from 0–8 hpa and photographed at 6 dpa. Compare to the DMSO control in (B). Arrow: healed epithelium.

Treatment with SB-431542 at later timepoints post-amputation halts regeneration at the time of application. 86% of control tadpoles show full regeneration by 6 dpa, with new muscle, vasculature, neural tube, notochord, and fin in evidence (n=36; Table 2 and Figure 4B, left panel). In contrast, only 4% of tadpoles treated with SB-431542 at 3 dpa had an equivalent amount of regenerated tissue; the other 96% display a reduced amount (approximately one-half to two-thirds) of regenerated tissue at 6 dpa, comparable to that seen in a 3 dpa regenerate (n=45; Table 2 and Figure 4B, right panel). Treatment with a different TGF-β inhibitor, SB-505124 (DaCosta Byfield et al., 2004), generated a similar result (0% full length, 100% short length, n=21; Table 2). A third inhibitor, LY-364947(Sawyer et al., 2003), also yielded the same phenotype (data not shown).

Table 2.

Phenotype of 6 dpa regenerates treated from 3–6 dpa.

Full regen. Short regen.* # of experiments
100 μM SB-431542 @ 3–6 dpa 2/45 (4%) 43/45 (96%) 3
25 μM SB-505124 @ 3–6 dpa 0/21 (0%) 21/21 (100%) 2
0.2% DMSO @ 3–6 dpa 31/36 (86%) 5/36 (14%) 3
Open in a new tab

Only those tadpoles showing good regeneration at 3 dpa were used for this experiment.

*

Short regenerates are between one-half to two-thirds the length of full regenerates, or approximately the length of a 3 dpa regenerate.

Finally, we examined the effect of a short window of inhibition of TGF-β signaling on the regeneration phenotype. Treatment of amputated tadpoles from 0–8 hpa was sufficient to permanently block regeneration: very little to no regeneration was observed at 6 dpa (0% regenerated fully, 79% unregenerated, n=97; Table 3 and Figure 4C) in contrast to the DMSO control (87% regenerated fully, 1% unregenerated, n=75; Table 3 and Figure 4B). SB-505124-treatment over the same window resulted in an identical phenotype (0% regenerated fully, 75% unregenerated, n=69; Table 3), as did LY-364947 treatment (data not shown). In a minority of inhibitor-treated cases (19% for SB-431542, 23% for SB-505124; Table 3), a very small spike containing no recognizable structures was present; this spike was never seen in controls (Table 3). Interestingly, these tadpoles do have a complete epithelium over the amputation site (arrow in Figure 4C), suggesting that the SB-431542-induced block to epithelial healing is reversible. P-Smad2 signaling recovers within 4 hours of SB-431542 removal, confirming inhibitor reversibility (data not shown).

Table 3.

Phenotype of 6 dpa regenerates treated from 0–8 hpa

Full regen. Partial regen.* Spike** No regen. # of experiments
100 μM SB-431542 @ 0–8 hpa 0/97 (0%) 2/97 (2%) 18/97 (19%) 77/97 (79%) 3
25 μM SB-505124 @ 0–8 hpa 0/69 (0%) 1/69 (1%) 16/69 (23%) 52/69 (75%) 4
0.2% DMSO @ 0–8 hpa 65/75 (87%) 9/75 (12%) 0/75 (0%) 1/75 (1%) 4
Open in a new tab
*

Partial regenerates contain some but not all recognizable structures and are considerably smaller than full regenerates.

**

Spikes do not contain any recognizable structures and are much smaller than regenerates.

Inhibition of TGF-β signaling reversibly blocks wound epithelium formation

We used confocal microscopy to examine bilayered wound epidermis formation following amputation. p63 was used to mark the basal epidermal layer (green); the apical layer was labeled with non-cell permeant Sulfo-NHS-LC-biotin (red). In the control, a full bilayered epithelium forms within 8 hours after amputation (Figure 5A, top row). Interestingly, the p63+ basal layer forms first, within 2 hpa; the apical layer does not completely form until 8 hpa (Figure 5A, top row). Epithelial healing appears to result from cell migration rather than proliferation, as very few mitotic nuclei are observed and healing is nocodazole-insensitive (data not shown). In contrast, amputated tadpoles treated at 0 hpa with SB-431542 do not form either the basal or the apical epithelial layer over the future regeneration bud, although the edges of the fins do heal (Figure 5A, bottom row). Tadpoles pre-treated with SB-431542 prior to amputation still displayed the same pattern of healing (data not shown), indicating that the differential effects on regeneration bud and fin healing are not caused by failure of the inhibitor to act quickly enough.

FIGURE 5. Inhibition of p-Smad2 signaling reversibly prevents formation of an epithelial layer over the amputation site.

FIGURE 5

Open in a new tab

The apical layer of the tadpole skin was labeled with 1 mg/ml sulfo-NHS-LC-biotin before amputation. (A) Regenerates were treated with 100μM SB-431542 or 0.2% DMSO at 0 hpa, fixed at 0, 1, 2, 4, and 8 hpa, and stained with streptavidin (red, apical epithelial layer) and anti-p63 (green, basal epithelial layer). Arrow: medial area showing no epithelial healing; arrowhead: fin showing normal epithelial healing. (B) Inhibition of epithelial healing is reversible. Regenerates were treated with 100 μM SB-431542 or 0.2% DMSO from 0–8 hpa, harvested at 2 dpa, and stained as above. Arrow: thickened epithelium seen at 2 dpa in the control but not the SB-431542-treated tadpole. Scale bar: 100 μm.

The effect of SB-431542 on wound epidermis formation is reversible. Amputated tadpoles treated with SB-431542 from 0–8 hpa and then recovered in inhibitor-free media display a bilayered wound epidermis by 24 hpa and later (Figure 5B and data not shown). The area below the new epidermis, however, appears flatter compared to controls at 2 dpa (Figure 5B), suggesting that a proper regeneration bud has not formed. Furthermore, the epithelium in control but not SB-431542-treated regenerates tends to become thicker by 2 dpa (arrow), likely representing regeneration of fin tissue (Figure 5B).

A brief, early inhibition of TGF-β signaling prevents establishment of regenerative structures

Although the effect of a short, early inhibition of TGF-β on wound epidermis formation is reversible (Figure 5B), these tails subsequently fail to regenerate (Figure 4C). We therefore examined regeneration bud structures such as the neural ampulla, notochord tip, and blastema in 2 dpa amputated tadpoles that had been treated from 0–8 hpa with SB-431542. Hematoxylin and eosin (H&E) staining of a sagittal section through the regeneration bud region reveals that the neural tube in SB-431542-treated tadpoles fails to form a hollow, rounded terminal ampulla (Figure 6A). This result is confirmed by staining with the neural antibody Xen1, which also reveals that the cut end of the neural tube in the SB-431542-treated tail is distorted and disorganized (Figure 6B). Treatment with SB-505124 and LY-364947 also resulted in similar phenotype of the neural tube (data not shown).

FIGURE 6. Early SB-431542 treatment blocks formation of regenerative structures at 48 hpa.

FIGURE 6

Open in a new tab

Regenerates were treated from 0–8 hpa with 100 μM SB-431542 or 0.2% DMSO and harvested at 2 dpa. (A) H&E staining of sagittal sections (B) Xen-1 (neural tissue) staining (C) Tor70 (regenerating notochord) staining (D) double staining for Pax7 (red, muscle satellite cells) and DNA (blue). nt: neural tube, not: notochord, bl: blastema, sc: satellite cells. Arrowheads: amputation plane. Scale bar: 100 μm.

Similarly, the densely packed mass of cells normally found at the caudal end of the cut notochord is not observed by H&E staining in an amputated tadpole treated with SB-431542 from 0–8 hpa (Figure 6A). Staining with the antibody Tor70, which recognizes immature (regenerating) but not mature notochord cells, shows a severe reduction in size of the regenerating notochord as well as a failure to form the characteristic bullet-shaped notochord tip (arrows, Figure 6C).

The number of mesenchymal blastema cells observed by H&E staining at 2 dpa in the SB-431542-treated (0–8 h) regeneration bud appears severely diminished compared to the control (Figure 6A). We therefore stained these samples for Pax7, a marker of satellite cells, which are resident in the blastema at this time. No Pax7+ cells are present in the regeneration bud of the SB-431542-treated tail (Figure 6D), indicating failure of satellite cells to populate the regeneration bud.

The SB-431542 treatment did not cause increased apoptosis, as assessed by activated caspase-3 immunostaining, either immediately following treatment (8 hpa) or at 2 dpa (Supplementary Figure S2). Additionally, nocodazole-treated embryos still developed a regeneration bud, indicating that this early SB-431542 phenotype does not result from a block to proliferation (data not shown).

Taken together with the phenotype analysis, these data show that a short, early inhibition (from 0–8 hpa) of p-Smad2 signaling is sufficient to block the later appearance of regenerative structures. Even though certain effects of TGF-β inhibition (wound epidermis formation) are reversible, regenerative capacity, as assessed by the presence of neural, notochord, and muscle precursors, is permanently lost.

Early TGF-β signal inhibition prevents activation of BMP and ERK signaling pathways in the regeneration bud

Previous studies suggest that BMP and ERK (Extracellular Signal-related Kinase) signaling pathways are essential for establishment and/or maintenance of regeneration bud structures in Xenopus (Beck et al., 2006; Suzuki et al., 2007). Therefore, we asked whether early SB-431542 treatment would affect the activation of BMP and ERK in regenerating tails.

Amputated tadpoles treated with SB-431542 or DMSO control from 0–8 hpa were examined at 2 dpa for the presence of BMP pathway components. Immunostaining for phospho-Smad1 (p-Smad1), the direct intracellular activator of BMP signaling, showed that control regenerates showed high levels of p-Smad1 in the regeneration bud, especially in the blastema, the regenerating fin, and the dorsal side of the neural ampulla (Figure 7A). In contrast, SB-431542-treated tadpoles did not display any p-Smad1 in the regeneration bud region (Figure 7A). This reduction in p-Smad1 staining in the regeneration bud is not caused by direct, non-specific inhibition of BMP receptors by SB-431542, since p-Smad1 in stage-matched, unamputated tadpole tails was unaffected by this treatment (Figure 7A).

FIGURE 7. Early inhibition of TGF-β signaling prevents subsequent activation of the BMP signaling pathway in the regeneration bud.

FIGURE 7

Open in a new tab

Regenerates treated with 100 μM SB-431542 or 0.2% DMSO from 0–8 hpa were harvested at 2 dpa and processed for the following markers: (A) immunofluorescence against phospho-Smad1 (p-Smad1, green), showing that p-Smad1 in the regeneration bud proper is eliminated by SB-431542 treatment. p-Smad1 (blue) in unamputated stage-matched tadpole tails was unaffected by SB-431542 (bottom row); (B) in situ hybridization with xMsx1 antisense probe (purple) –black arrow: xMsx1 expression in the dorsal blastema; (C) immunofluorescence for the mitotic marker phospho-Histone H3 (pH3, green). Total DNA in (A) and (C) is shown in blue. Arrowheads: amputation plane. Scale bar: 100 μM.

We also examined expression of the transcription factor Msx1, a known target of BMP signaling that is upregulated in virtually all regenerating systems (Stoick-Cooper et al., 2007). In situ hybridization for xMsx1 shows that Msx1 expression in the dorsal blastema of 2 dpa regenerates is abolished in tadpoles treated with SB-431542 from 0–8 hpa (Figure 7B), consistent with the lack of p-Smad1 in these tadpoles.

Beck et al (2007) have shown that regenerates transgenic for the BMP inhibitor noggin demonstrate reduced proliferation in the regeneration bud. Therefore, we examined cellular proliferation by staining for the mitotic marker phospho-Histone H3 (pH3). We find that the increased proliferation normally seen in the regeneration bud by 2 dpa is absent in SB-431542-treated tails (Figure 7C). Taken together, these data indicate that TGF-β signaling very early in the regenerative process is required upstream of BMP signaling and its dependent processes in the regeneration bud.

We have also examined the role of ERK signaling downstream of TGFβ in tail regeneration, as it is implicated in both wound healing and regeneration and is downstream of FGFs (Christen and Slack, 1999; Poss et al., 2000; Suzuki et al., 2007). Staining for diphosphorylated ERK1/2 (pERK1/2) in regenerating tails shows that pERK1/2 activation at 8 hpa is primarily in the wound epithelium, consistent with a role in wound healing, but shifts to a largely blastemal pattern by 2 dpa (Figure 8). Interestingly, this blastemal pERK1/2 component is never seen in amputated tadpoles treated from 0–8 hpa with SB-431542 (Figure 8B). Instead, these treated tadpoles often display a continued expression of pERK1/2 in all or part of the wound epithelium, although in some cases no expression at all is seen (Figure 8B and data not shown).

FIGURE 8. Early inhibition of TGF-β signaling results in missing or aberrant pERK1/2 activation in the regeneration bud.

FIGURE 8

Open in a new tab

Immunofluorescence staining for phospho-ERK1/2 (red) and total DNA (blue). (A) 8 hpa untreated regenerates, showing pERK1/2 staining primarily in the wound epithelium (orange arrow). (B) 2 dpa regenerates, treated with 100 μM SB-431542 or 0.2% DMSO from 0–8 hpa. White arrow: blastemal p-ERK1/2 in controls. Green arrow: aberrant pERK1/2 in the epithelium of SB-431542-treated tadpoles. Arrowheads: amputation plane. Scale bar: 100 μM.

TGF-β-dependent cell proliferation specifically in the regeneration bud

Tadpoles in which TGF-β signaling was blocked only after the establishment of a regeneration bud (at 3 dpa) still stop regenerating (Figure 4B). High levels of proliferation are normally observed in the regeneration bud starting at about 2 dpa ((Beck et al., 2006); D.M.H and M.W., unpublished). We therefore examined the effect of late TGF-β inhibition on cellular proliferation by immunostaining for pH3, which revealed that regenerates treated with SB-431542 at 2 dpa have greatly reduced proliferation in the regeneration bud at 3 dpa (Figure 9A). We quantitated the number of pH3+ cells in a 2000 μm2 area of the regeneration bud, fin, and somite anterior to the amputation site. This analysis showed that the number of mitotic nuclei was significantly increased (approximately four-fold) in the regeneration bud compared to surrounding tissues, and that this increase was completely abolished with SB-431542 treatment (p=0.00005, Figure 9B). SB-431542 treatment had no effect on numbers of mitotic nuclei in the fin or somite (Figure 9B). In order to more clearly determine which cells in the regeneration bud were affected, nocodazole was used to “trap” dividing cells, effectively allowing us to visualize by pH3 staining all of the cells that have divided in a 3 hour time window. This analysis showed that proliferation in the neural tube, notochord, and blastema at 3 dpa were all reduced by SB-431542 treatment (Figure 9C).

FIGURE 9. Effect of late TGF-β inhibition on proliferation in the regeneration bud.

FIGURE 9

Open in a new tab

(A) Phospho-Histone H3 (pH3) expression. Regenerates were treated with 100 μM SB-431542 or 0.2% DMSO at 2 dpa and harvested at 3 dpa for immunofluorescence for anti-pH3 (green). Total DNA is shown in blue. (B) Quantitation of pH3+ nuclei in 3 dpa regenerates treated as in (A). Numbers of pH3+ nuclei in a 2000 μm2 area of a confocal section were counted in the regeneration bud, fin, and somite. For each sample, a similarly located area was selected. N=7 for DMSO and 6 for SB, except for fin samples, where N=5 for both DMSO and SB. Samples were drawn from 3 separate experiments, except for fins (2 experiments). Error bars: ±1SD. *: p=0.00005. (C) Regenerates were treated with 100 μM SB-431542 or 0.2% DMSO for 9 hours total, including final 3 hours with nocodazole (NOC) arrest, and stained at 3 dpa for pH3 (green) and DNA (blue). nt:neural tube, not:notochord, bl:blastema. (D) BrdU incorporation assay. Regenerates were treated with 100 μM SB-431542 or 0.2% DMSO at 2 dpa and harvested at 4 dpa. BrdU was added to the media 6 hours after initiation of SB-431542 treatment. BrdU+ nuclei are labeled in red and total DNA in blue. BrdU incorporation in the control is increased at the regeneration bud (white arrow) but not rostral to the level of the amputation plane (orange arrow); in contrast, the SB-431542-treated tail does not show any obvious rostro-caudal differences in BrdU incorporation. Similarly, proliferation in the fin (green arrow) is not affected. White arrow: regeneration bud; orange arrow: tail rostral to amputation plane; green arrow: fin in regeneration bud. Scale bar: 100 μm. Arrowheads: amputation plane.

Similarly, a BrdU incorporation assay shows that, over a longer time window of 2–4 dpa, very few cells in the regeneration bud of SB-treated tadpoles have divided, whereas almost all of the cells in the control have incorporated BrdU (white arrows, Figure 9D). Levels of BrdU incorporation rostral to the plane of amputation remain at a low baseline level in both treated and untreated cases (orange arrows in Figure 9D, top row). BrdU incorporation in the fin and epidermis surrounding the regeneration bud also does not seem to be affected by SB-431542 (green arrows in Figure 9D, lower panels). Taken together, these data indicate that TGF-β signaling stimulates cell division only in the regeneration bud proper but not in outlying tissues.

Treatment of regenerates at 3 dpa with the cell cycle inhibitors hydroxyurea or nocodazole results in a very similar phenotype to SB-431542 treatment (Figure S3). Direct proliferation inhibition results in a reversible regeneration block, as regenerates treated with nocodazole from 3–4 dpa can recover and begin to regenerate once again after nocodazole removal (Figure S3). In contrast, regenerates treated during the same time window with SB-431542 are unable to maintain a regenerative state and regeneration does not recommence (Figure S3), despite recovery of p-Smad2 signaling shortly after inhibitor removal (data not shown).

DISCUSSION

Rapid and sustained activation of the TGF-β signaling pathway in regeneration

We have combined spatial analysis using a p-Smad2-specific antibody with temporal analysis using the chemical inhibitor SB-431542 to elucidate multiple distinct roles for TGFβ signaling during regeneration. The TGF-β signaling pathway is rapidly activated and remains on throughout regeneration; p-Smad2 is observed as early as 15 minutes post-amputation (Figure 1). To the best of our knowledge, this is earliest reported molecular event in regeneration.

Activation of αvβ6 integrin by mechanical stimulation at wounds and other sites of epithelial remodeling is a rapid, highly-localized mechanism for the activation of ECM-bound latent TGF-β protein (Breuss et al., 1995; Munger et al., 1999; Sheppard, 2005). This mechanism provides the most plausible explanation for the rapidly appearing Smad2 phosphorylation in the basal epidermal layer. Additionally, in mammalian studies, platelets and macrophages can migrate to wound sites and release TGF-β (Assoian et al., 1987; Assoian and Sporn, 1986; Blakytny et al., 2004). Although the role of platelets and macrophages have not been studied in Xenopus, blood clotting occurs within minutes following amputation; this blood clot and its associated molecules may have a significant early role in the regenerative process. Finally, although the literature suggests that activation of latent TGF-β is the most likely candidate for an early, localized signal, we cannot rule out the involvement of other family members, such as activin.

Inhibition of TGF-β signaling blocks the following events: (1) formation of a wound epithelium over the amputation site, (2) establishment of regenerative structures, (3) proper activation and localization of BMP and ERK signaling pathways, and (4) cell proliferation in the regeneration bud. Interestingly, inhibition of BMP signaling does not affect early events such as epithelial healing or neural ampulla formation, but only blocks later events such as proliferation (Beck et al., 2006). In contrast, our data show that a sustained level of p-Smad2 is required for both establishment and maintenance of regeneration and that TGF-β signaling is upstream of BMP and ERK signaling pathways, suggesting that TGF-β activation may be one of the first crucial and defining events of the regenerative pathway. Notably, recent work in zebrafish has shown that ActivinβA is important for cell migration and proliferation in regenerating zebrafish tail fins (Jazwinska et al., 2007), indicating that a crucial role for TGF-β signals is conserved between multiple regenerative systems.

Spatial regulation of p-Smad2 activation

P-Smad2 expression is initially restricted to a single cell type (the basal layer of the wound epidermis) in the regeneration bud, and only later does its expression become widespread. The mechanism for this differential regulation of signal activation is unclear. One hypothesis is that only the p63+ basal epidermal layer is exposed to active TGF-β ligand. This possibility is consistent with the activation of TGF-β by αvβ6 integrin, which requires direct cell-to-cell contact and results in extremely localized signaling (Munger et al., 1999; Sheppard, 2005). Another possibility is that only p63+ cells have the capacity to respond to TGF-β during the early stages of regeneration. Examination of positive and negative regulators of TGF-β signaling at the amputation site will be an interesting area for future study.

In contrast to the highly localized activation of p-Smad2 immediately post-amputation, the later broad distribution of p-Smad2 suggests a freely diffusible ligand that activates signaling in an unrestricted, concentration-dependent manner. Thus, the early restriction must be lifted from 24 hpa onward, perhaps by widespread activation of the necessary signaling components and/or downregulation of repressors in previously unresponsive cells. The change from early, regulated p-Smad2 to later, widespread signaling may represent a molecular switch between two broad stages of regeneration: an early stage characterized by cellular migration and establishment of regenerative structures and a late stage characterized by cellular proliferation and differentiation of new tissues.

Epithelial-mesenchymal interactions in the regeneration bud

TGF-β signaling stimulates the formation of a wound epidermis following amputation, which is required, but not in itself sufficient, for tail regeneration. A second signaling event between the epidermis and the underlying mesenchyme appears to be necessary to initiate the regeneration program and establish a regeneration bud. TGF-β may be needed simply for formation of the epithelium, which can then participate in an additional, TGF-β independent inductive event. We do not favor this hypothesis because p-Smad2 signaling remains high even after epithelial healing is complete. Alternatively, TGF-β may be directly or indirectly involved in an inductive event. During the period of regenerative competency, p-Smad2 is present specifically in the basal layer of the wound epithelium (Figure 2), and TGF-β2 and TGF-β5 are both expressed in the underlying tissue (Figure 3). One possible scenario is that ligand produced in the putative regeneration bud region would stimulate the TGF-β signaling pathway in the basal epithelial layer, which in turn would send a second signal back to the mesenchyme, thus enabling the activation of a regenerative pathway. In the absence of this type of signaling mechanism following amputation and epithelial healing, a simple “wound healing” pathway would prevail, resulting in loss of regenerative competence.

FGFs may provide this epithelial-to-mesenchymal signal. FGF-20 has been implicated in blastema formation in zebrafish fin regeneration (Whitehead et al., 2005); FGF-8 is produced by the apical ectodermal ridge (AER), an epithelial structure in the developing limb which signals to and patterns the underlying mesenchyme, and is expressed in regenerating Xenopus tails and limbs (Beck et al., 2006; Mahmood et al., 1995). Furthermore, FGFs signal through the ERK pathway, which is activated in the regeneration bud in a TGF-β-dependent manner (Figure 8).

TGF-β signaling and cellular proliferation during regeneration

In this work, we report that TGF-β signaling drives a localized proliferation response in the regeneration bud proper, but does not affect cell division in adjacent tissues (Figure 9). Although TGF-β is generally considered anti-proliferative, it can stimulate proliferation in certain cell types or under certain conditions (Ruscetti et al., 2005; Seoane, 2006). Jazwinska et al (2007) have also recently shown that activin can promote proliferation in zebrafish fin blastema, indicating that TGF-β signals can induce proliferation in multiple regenerative systems. TGFβ may promote proliferation through its effects on BMP and/or ERK signaling, as inhibition of either of these pathways has been shown to reduce proliferation in regenerative structures (Beck et al., 2006; Poss et al., 2000).

p63 and regeneration

The tight localization of p-Smad2 to the p63+ layer of the epidermis during early regeneration is particularly striking. p63 is expressed at high levels in the basal, progenitor layers of many types of epithelia, and may play a role in the maintenance or commitment of an epithelial stem cell population (Yang et al., 1999). p63 null mice fail to form an AER, resulting in severely diminished FGF-8 production and ultimately truncated limbs missing various structures (Mills et al., 1999; Yang et al., 1999). By analogy, the p63+ layer over the amputation site may act as an organizing, AER-like structure that directs the early tissue interactions and epithelial-mesenchymal signaling events required to establish a regeneration bud. It will be interesting in the future to examine the effect of loss of p63 function on regenerative phenotypes in Xenopus tadpoles and other model systems.

In our experiments, the p63+ layer always forms first over the wound site, while migration of the second, outer layer is somewhat delayed. We have observed the same phenomenon in non-regenerating situations, such as in the rostral cut surface of an explanted tail, indicating that the p63+ cell layer may be the first structure formed during healing of wounds, both regenerative and non-regenerative, in Xenopus tadpoles. p63 has been identified as a key regulator of epithelial adhesion in mammals (Carroll et al., 2006), and may confer distinctive adhesive properties that allow rapid wound covering by the basal epidermal layer in Xenopus tadpoles. Xenopus skin, with its relatively simple bilayered organization, may therefore serve as a useful model system for the study of the role of p63 during wound healing as well as regeneration.

A time window of regenerative competence

The molecular basis for the decision between regeneration and simple wound healing is central to the problem of understanding epimorphic regeneration. The ability to establish a regeneration program is lost after as little as eight hours in SB-431542, even though restoration of p-Smad2 signaling and formation of a morphologically normal bilaminar wound epidermis occur rapidly following SB-431542 removal. This observation indicates that there is a limited time window following amputation within which regeneration can commence, and after this window closes only simple healing is possible. Whether this represents a change in the wound epidermis itself (e.g. a time-dependent loss of capacity to produce a regenerative signal), a change in underlying mesenchymal cells (e.g. a time-dependent loss of capacity to respond to a signal from the wound epidermis), or some other mechanism will be an interesting area for further investigation.

Signals regulating wound healing and regeneration

Our work demonstrates that TGF-β signaling is essential both for initial closure of a wound, and subsequently for the establishment and maintenance of the regeneration program. Early events in regeneration are likely to be similar to those in non-regenerative wound healing; however, downstream events must then diverge.

What regulates the distinction between wound healing and regeneration? Our examination of p-ERK1/2 signals during regeneration suggests the intriguing possibility that a shift in p-ERK activation from an early, primarily epithelial localization to a later, primarily blastemal localization (Figure 8) may underlie the shift from the initial wound healing phase to the regenerative phase. Furthermore, the shift in p-ERK activation patterns does not occur when TGF-β signaling is inhibited early; rather, p-ERK often remains high in the wound epithelium long after it normally subsides (Figure 8), suggesting that in the absence of TGF-β signaling, a sustained wound healing response prevails and regeneration is inhibited. Interestingly, experiments in younger Xenopus embryos showed that developmentally regulated p-ERK signals were FGF-dependent, whereas p-ERK induced by wounding was not (Christen and Slack, 1999). In light of this observation, it will be interesting to examine the FGF dependence of the two phases of ERK signal activation during regeneration.

In in vivo and in vitro wounding studies, TGF-β ligands have been implicated both in productive processes such as re-epithelialization and contractility and in non-productive ones such as fibrosis (Leask and Abraham, 2004; Werner et al., 2007). For example, both the rate of wound re-epithelialization and the degree of scarring were enhanced in mice overexpressing activin and reduced in those overexpressing follistatin (Munz et al., 1999; Wankell et al., 2001). Smad3-null mice, intriguingly, display either enhanced or inhibited wound closure depending on the type of wound (Arany et al., 2006; Ashcroft et al., 1999). These and other studies underscore the importance of proper signal regulation in the choice between a productive, regenerative response and a non-productive, fibrotic one.

This distinction is likely to be critical for understanding the molecular pathways that distinguish regeneration from simple wound healing. Different cell types and different transcriptional response programs are likely to lie downstream of signals common to both wound healing and regeneration. The combination of specific pharmacological inhibitors such as SB-431542, techniques that permit labeling of specific cell populations in vivo (Gargioli and Slack, 2004), and tissue-specific rescue from inhibition (Ho et al., 2006) will allow further spatiotemporal resolution of the role of signaling pathways in specific events during the regenerative process.

Supplementary Material

01

Figure S1: Temporal expression profile of TGF-β ligands during regeneration. RT-PCR analysis of xTGF-β2, xTGF-β5, xActivinβA, xGDF11, and xEF1α expression in regenerating tail tips at 0, 1, 4, and 48 hpa. Control samples were also taken from an equivalent area of the tail of unamputated tadpoles at 48 hpa.

02

Figure S2: Treatment with SB-431542 from 0–8 hpa does not affect apoptosis in the regeneration bud. Amputated embryos treated with 100 μM SB-431542 or 0.2% DMSO from 0–8 hpa were stained with antibody against activated Caspase-3 at 8 hpa and 2 dpa.

03

Figure S3: Effect of cell cycle inhibitors on regeneration. 3 dpa regenerates were treated with 100 μM SB-431542, 30 mM hydroxyurea (HU), 10 μg/ml nocodazole (NOC), or 0.2% DMSO. Some SB-431542 and NOC regenerates were subsequently washed out of inhibitor at 4 dpa as indicated. All regenerates were photographed at 6 dpa.

Acknowledgments

We thank Dr. Frank McKeon for 4A4 antibody, Dr. Richard Harland for Tor70 antibody, and Dr. Igor Dawid for xTGF-β2 and xTGF-β5 in situ probes. Monoclonal antibodies from Atsushi Kawakami (Pax7) and Ariel Ruiz i Atalba (Xen1) were obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by The University of Iowa, Department of Biological Sciences, Iowa City, IA 52242. We also thank Dr. Jennifer Waters and Lara Petrak at the Nikon Imaging Facility at Harvard Medical School for assistance with confocal microscopy. We would like to thank Dr. Michael Levin for help with optimizing in situ hybridization in regenerating tails. Finally, we thank the members of the Whitman Lab, past and present, specifically Dr. Stefan Wawersik, who initiated the study of tail regeneration in our lab. This work was supported by NIH grant HD29468 to M.W.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  1. Arany PR, Flanders KC, Kobayashi T, Kuo CK, Stuelten C, Desai KV, Tuan R, Rennard SI, Roberts AB. Smad3 deficiency alters key structural elements of the extracellular matrix and mechanotransduction of wound closure. Proc Natl Acad Sci U S A. 2006;103:9250–5. doi: 10.1073/pnas.0602473103. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Ashcroft GS, Yang X, Glick AB, Weinstein M, Letterio JL, Mizel DE, Anzano M, Greenwell-Wild T, Wahl SM, Deng C, et al. Mice lacking Smad3 show accelerated wound healing and an impaired local inflammatory response. Nat Cell Biol. 1999;1:260–6. doi: 10.1038/12971. [DOI] [PubMed] [Google Scholar]
  3. Assoian RK, Fleurdelys BE, Stevenson HC, Miller PJ, Madtes DK, Raines EW, Ross R, Sporn MB. Expression and secretion of type beta transforming growth factor by activated human macrophages. Proc Natl Acad Sci U S A. 1987;84:6020–4. doi: 10.1073/pnas.84.17.6020. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Assoian RK, Sporn MB. Type beta transforming growth factor in human platelets: release during platelet degranulation and action on vascular smooth muscle cells. J Cell Biol. 1986;102:1217–23. doi: 10.1083/jcb.102.4.1217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Beck CW, Christen B, Barker D, Slack JM. Temporal requirement for bone morphogenetic proteins in regeneration of the tail and limb of Xenopus tadpoles. Mech Dev. 2006;123:674–88. doi: 10.1016/j.mod.2006.07.001. [DOI] [PubMed] [Google Scholar]
  6. Beck CW, Christen B, Slack JM. Molecular pathways needed for regeneration of spinal cord and muscle in a vertebrate. Dev Cell. 2003;5:429–39. doi: 10.1016/s1534-5807(03)00233-8. [DOI] [PubMed] [Google Scholar]
  7. Blakytny R, Ludlow A, Martin GE, Ireland G, Lund LR, Ferguson MW, Brunner G. Latent TGF-beta1 activation by platelets. J Cell Physiol. 2004;199:67–76. doi: 10.1002/jcp.10454. [DOI] [PubMed] [Google Scholar]
  8. Breuss JM, Gallo J, DeLisser HM, Klimanskaya IV, Folkesson HG, Pittet JF, Nishimura SL, Aldape K, Landers DV, Carpenter W, et al. Expression of the beta 6 integrin subunit in development, neoplasia and tissue repair suggests a role in epithelial remodeling. J Cell Sci. 1995;108(Pt 6):2241–51. doi: 10.1242/jcs.108.6.2241. [DOI] [PubMed] [Google Scholar]
  9. Carcamo J, Weis FM, Ventura F, Wieser R, Wrana JL, Attisano L, Massague J. Type I receptors specify growth-inhibitory and transcriptional responses to transforming growth factor beta and activin. Mol Cell Biol. 1994;14:3810–21. doi: 10.1128/mcb.14.6.3810. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Carroll DK, Carroll JS, Leong CO, Cheng F, Brown M, Mills AA, Brugge JS, Ellisen LW. p63 regulates an adhesion programme and cell survival in epithelial cells. Nat Cell Biol. 2006;8:551–61. doi: 10.1038/ncb1420. [DOI] [PubMed] [Google Scholar]
  11. Chen Y, Lin G, Slack JM. Control of muscle regeneration in the Xenopus tadpole tail by Pax7. Development. 2006;133:2303–13. doi: 10.1242/dev.02397. [DOI] [PubMed] [Google Scholar]
  12. Christen B, Slack JM. Spatial response to fibroblast growth factor signalling in Xenopus embryos. Development. 1999;126:119–25. doi: 10.1242/dev.126.1.119. [DOI] [PubMed] [Google Scholar]
  13. DaCosta Byfield S, Major C, Laping NJ, Roberts AB. SB-505124 is a selective inhibitor of transforming growth factor-beta type I receptors ALK4, ALK5, and ALK7. Mol Pharmacol. 2004;65:744–52. doi: 10.1124/mol.65.3.744. [DOI] [PubMed] [Google Scholar]
  14. Faure S, Lee MA, Keller T, ten Dijke P, Whitman M. Endogenous patterns of TGFbeta superfamily signaling during early Xenopus development. Development. 2000;127:2917–31. doi: 10.1242/dev.127.13.2917. [DOI] [PubMed] [Google Scholar]
  15. Gargioli C, Slack JM. Cell lineage tracing during Xenopus tail regeneration. Development. 2004;131:2669–79. doi: 10.1242/dev.01155. [DOI] [PubMed] [Google Scholar]
  16. Harland RM. In situ hybridization: an improved whole-mount method for Xenopus embryos. Methods Cell Biol. 1991;36:685–95. doi: 10.1016/s0091-679x(08)60307-6. [DOI] [PubMed] [Google Scholar]
  17. Ho DM, Chan J, Bayliss P, Whitman M. Inhibitor-resistant type I receptors reveal specific requirements for TGF-beta signaling in vivo. Dev Biol. 2006;295:730–42. doi: 10.1016/j.ydbio.2006.03.050. [DOI] [PubMed] [Google Scholar]
  18. Inman GJ, Nicolas FJ, Callahan JF, Harling JD, Gaster LM, Reith AD, Laping NJ, Hill CS. SB-431542 is a potent and specific inhibitor of transforming growth factor-beta superfamily type I activin receptor-like kinase (ALK) receptors ALK4, ALK5, and ALK7. Mol Pharmacol. 2002;62:65–74. doi: 10.1124/mol.62.1.65. [DOI] [PubMed] [Google Scholar]
  19. Jazwinska A, Badakov R, Keating MT. Activin-betaA signaling is required for zebrafish fin regeneration. Curr Biol. 2007;17:1390–5. doi: 10.1016/j.cub.2007.07.019. [DOI] [PubMed] [Google Scholar]
  20. Kondaiah P, Taira M, Vempati UD, Dawid IB. Transforming growth factor-beta5 expression during early development of Xenopus laevis. Mech Dev. 2000;95:207–9. doi: 10.1016/s0925-4773(00)00326-9. [DOI] [PubMed] [Google Scholar]
  21. Leask A, Abraham DJ. TGF-beta signaling and the fibrotic response. Faseb J. 2004;18:816–27. doi: 10.1096/fj.03-1273rev. [DOI] [PubMed] [Google Scholar]
  22. Lee MA, Heasman J, Whitman M. Timing of endogenous activin-like signals and regional specification of the Xenopus embryo. Development. 2001;128:2939–52. doi: 10.1242/dev.128.15.2939. [DOI] [PubMed] [Google Scholar]
  23. Lu P, Barad M, Vize PD. Xenopus p63 expression in early ectoderm and neurectoderm. Mech Dev. 2001;102:275–8. doi: 10.1016/s0925-4773(01)00315-x. [DOI] [PubMed] [Google Scholar]
  24. Mahmood R, Bresnick J, Hornbruch A, Mahony C, Morton N, Colquhoun K, Martin P, Lumsden A, Dickson C, Mason I. A role for FGF-8 in the initiation and maintenance of vertebrate limb bud outgrowth. Curr Biol. 1995;5:797–806. doi: 10.1016/s0960-9822(95)00157-6. [DOI] [PubMed] [Google Scholar]
  25. Mills AA, Zheng B, Wang XJ, Vogel H, Roop DR, Bradley A. p63 is a p53 homologue required for limb and epidermal morphogenesis. Nature. 1999;398:708–13. doi: 10.1038/19531. [DOI] [PubMed] [Google Scholar]
  26. Munger JS, Huang X, Kawakatsu H, Griffiths MJ, Dalton SL, Wu J, Pittet JF, Kaminski N, Garat C, Matthay MA, et al. The integrin alpha v beta 6 binds and activates latent TGF beta 1: a mechanism for regulating pulmonary inflammation and fibrosis. Cell. 1999;96:319–28. doi: 10.1016/s0092-8674(00)80545-0. [DOI] [PubMed] [Google Scholar]
  27. Munz B, Smola H, Engelhardt F, Bleuel K, Brauchle M, Lein I, Evans LW, Huylebroeck D, Balling R, Werner S. Overexpression of activin A in the skin of transgenic mice reveals new activities of activin in epidermal morphogenesis, dermal fibrosis and wound repair. Embo J. 1999;18:5205–15. doi: 10.1093/emboj/18.19.5205. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Nieuwkoop PD, Faber J. Normal Table of Xenopus laevis (Daudin) Amsterdam: North Holland Publishing Company; 1967. [Google Scholar]
  29. Poss KD, Shen J, Nechiporuk A, McMahon G, Thisse B, Thisse C, Keating MT. Roles for Fgf signaling during zebrafish fin regeneration. Dev Biol. 2000;222:347–58. doi: 10.1006/dbio.2000.9722. [DOI] [PubMed] [Google Scholar]
  30. Rebbert ML, Bhatia-Dey N, Dawid IB. The sequence of TGF-beta 2 from Xenopus laevis. Nucleic Acids Res. 1990;18:2185. doi: 10.1093/nar/18.8.2185. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Ruscetti FW, Akel S, Bartelmez SH. Autocrine transforming growth factor-beta regulation of hematopoiesis: many outcomes that depend on the context. Oncogene. 2005;24:5751–63. doi: 10.1038/sj.onc.1208921. [DOI] [PubMed] [Google Scholar]
  32. Ryden M, Imamura T, Jornvall H, Belluardo N, Neveu I, Trupp M, Okadome T, ten Dijke P, Ibanez CF. A novel type I receptor serine-threonine kinase predominantly expressed in the adult central nervous system. J Biol Chem. 1996;271:30603–9. doi: 10.1074/jbc.271.48.30603. [DOI] [PubMed] [Google Scholar]
  33. Sawyer JS, Anderson BD, Beight DW, Campbell RM, Jones ML, Herron DK, Lampe JW, McCowan JR, McMillen WT, Mort N, et al. Synthesis and activity of new aryl- and heteroaryl-substituted pyrazole inhibitors of the transforming growth factor-beta type I receptor kinase domain. J Med Chem. 2003;46:3953–6. doi: 10.1021/jm0205705. [DOI] [PubMed] [Google Scholar]
  34. Seoane J. Escaping from the TGFbeta anti-proliferative control. Carcinogenesis. 2006;27:2148–56. doi: 10.1093/carcin/bgl068. [DOI] [PubMed] [Google Scholar]
  35. Sheppard D. Integrin-mediated activation of latent transforming growth factor beta. Cancer Metastasis Rev. 2005;24:395–402. doi: 10.1007/s10555-005-5131-6. [DOI] [PubMed] [Google Scholar]
  36. Shi Y, Massague J. Mechanisms of TGF-beta signaling from cell membrane to the nucleus. Cell. 2003;113:685–700. doi: 10.1016/s0092-8674(03)00432-x. [DOI] [PubMed] [Google Scholar]
  37. Slack JM, Beck CW, Gargioli C, Christen B. Cellular and molecular mechanisms of regeneration in Xenopus. Philos Trans R Soc Lond B Biol Sci. 2004;359:745–51. doi: 10.1098/rstb.2004.1463. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Stoick-Cooper CL, Weidinger G, Riehle KJ, Hubbert C, Major MB, Fausto N, Moon RT. Distinct Wnt signaling pathways have opposing roles in appendage regeneration. Development. 2007;134:479–89. doi: 10.1242/dev.001123. [DOI] [PubMed] [Google Scholar]
  39. Sugiura T, Taniguchi Y, Tazaki A, Ueno N, Watanabe K, Mochii M. Differential gene expression between the embryonic tail bud and regenerating larval tail in Xenopus laevis. Dev Growth Differ. 2004;46:97–105. doi: 10.1111/j.1440-169x.2004.00727.x. [DOI] [PubMed] [Google Scholar]
  40. Suzuki M, Satoh A, Ide H, Tamura K. Transgenic Xenopus with prx1 limb enhancer reveals crucial contribution of MEK/ERK and PI3K/AKT pathways in blastema formation during limb regeneration. Dev Biol. 2007;304:675–86. doi: 10.1016/j.ydbio.2007.01.019. [DOI] [PubMed] [Google Scholar]
  41. Wankell M, Munz B, Hubner G, Hans W, Wolf E, Goppelt A, Werner S. Impaired wound healing in transgenic mice overexpressing the activin antagonist follistatin in the epidermis. Embo J. 2001;20:5361–72. doi: 10.1093/emboj/20.19.5361. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Werner S, Krieg T, Smola H. Keratinocyte-fibroblast interactions in wound healing. J Invest Dermatol. 2007;127:998–1008. doi: 10.1038/sj.jid.5700786. [DOI] [PubMed] [Google Scholar]
  43. Whitehead GG, Makino S, Lien CL, Keating MT. fgf20 is essential for initiating zebrafish fin regeneration. Science. 2005;310:1957–60. doi: 10.1126/science.1117637. [DOI] [PubMed] [Google Scholar]
  44. Whitman M. Smads and early developmental signaling by the TGFbeta superfamily. Genes Dev. 1998;12:2445–62. doi: 10.1101/gad.12.16.2445. [DOI] [PubMed] [Google Scholar]
  45. Yang A, Kaghad M, Wang Y, Gillett E, Fleming MD, Dotsch V, Andrews NC, Caput D, McKeon F. p63, a p53 homolog at 3q27-29, encodes multiple products with transactivating, death-inducing, and dominant-negative activities. Mol Cell. 1998;2:305–16. doi: 10.1016/s1097-2765(00)80275-0. [DOI] [PubMed] [Google Scholar]
  46. Yang A, Schweitzer R, Sun D, Kaghad M, Walker N, Bronson RT, Tabin C, Sharpe A, Caput D, Crum C, et al. p63 is essential for regenerative proliferation in limb, craniofacial and epithelial development. Nature. 1999;398:714–8. doi: 10.1038/19539. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

01

Figure S1: Temporal expression profile of TGF-β ligands during regeneration. RT-PCR analysis of xTGF-β2, xTGF-β5, xActivinβA, xGDF11, and xEF1α expression in regenerating tail tips at 0, 1, 4, and 48 hpa. Control samples were also taken from an equivalent area of the tail of unamputated tadpoles at 48 hpa.

NIHMS42557-supplement-01.tif (1.6MB, tif)
02

Figure S2: Treatment with SB-431542 from 0–8 hpa does not affect apoptosis in the regeneration bud. Amputated embryos treated with 100 μM SB-431542 or 0.2% DMSO from 0–8 hpa were stained with antibody against activated Caspase-3 at 8 hpa and 2 dpa.

NIHMS42557-supplement-02.tif (2.9MB, tif)
03

Figure S3: Effect of cell cycle inhibitors on regeneration. 3 dpa regenerates were treated with 100 μM SB-431542, 30 mM hydroxyurea (HU), 10 μg/ml nocodazole (NOC), or 0.2% DMSO. Some SB-431542 and NOC regenerates were subsequently washed out of inhibitor at 4 dpa as indicated. All regenerates were photographed at 6 dpa.

NIHMS42557-supplement-03.tif (9.1MB, tif)

RESOURCES