Present address: Max Planck Institute of Molecular Cell Biology and Genetics, 01307 Dresden, Germany.
Spinal cord is required for proper regeneration of the tail in Xenopus tadpoles
Article first published online: 19 JAN 2008
© 2008 The Authors
Development, Growth & Differentiation
Volume 50, Issue 2, pages 109–120, February 2008
How to Cite
Taniguchi, Y., Sugiura, T., Tazaki, A., Watanabe, K. and Mochii, M. (2008), Spinal cord is required for proper regeneration of the tail in Xenopus tadpoles. Development, Growth & Differentiation, 50: 109–120. doi: 10.1111/j.1440-169X.2007.00981.x
- Issue published online: 19 JAN 2008
- Article first published online: 19 JAN 2008
- Received 7 May 2007; revised 8 November 2007; accepted 15 November 2007.
- fibroblast growth factors;
- spinal cord;
Tail regeneration in urodeles is dependent on the spinal cord (SC), but it is believed that anuran larvae regenerate normal tails without the SC. To evaluate the precise role of the SC in anuran tail regeneration, we developed a simple operation method to ablate the SC completely and minimize the damage to the tadpole using Xenopus laevis. The SC-ablated tadpole regenerated a twisted and smaller tail. These morphological abnormalities were attributed to defects in the notochord (NC), as the regenerated NC in the SC-ablated tail was short, slim and twisted. The SC ablation never affected the early steps of the regeneration, including closure of the amputated surface with epidermis and accumulation of the NC precursor cells. The proliferation rate of the NC precursor cells, however, was reduced, and NC cell maturation was retarded in the SC-ablated tail. These results show that the SC has an essential role in the normal tail regeneration of Xenopus larvae, especially in the proliferation and differentiation of the NC cells. Gene expression analysis and implantation of a bead soaked with growth factor showed that fibroblast growth factor-2 and -10 were involved in the signaling molecules, which were expressed in the SC and stimulated growth of the NC cells.
Nerve-dependent regeneration has been well documented in the limbs and tails of urodeles (reviewed in Goldfarb 1909; Goss 1969; Ferretti & Geraudie 1998; Mescher 2001). Denervation of the three spinal nerves from the forelimb by cutting them at the trunk region results in no regeneration of the limb (reviewed in Singer 1952). Ablation of the spinal cord (SC) from the tail inhibits the tail regeneration completely (Holtzer et al. 1955). The initial response induced by the amputation is not affected by the denervation. The amputation plane is covered with wound epidermis, under which mesenchymal cells accumulate whether a nerve is present or not, but the following step in which the mesenchymal cells proliferate to form the blastema is inhibited by the denervation.
Singer showed that extracts from the nerve and brain exhibited soluble activity that promoted DNA and protein synthesis in blastema cells (Singer et al. 1976). The nerve-derived soluble factors, or neurotrophic factors, have been extensively studied since the publication of Singer's report. Neuregulin and fibroblast growth factor (FGF)-2 have been shown to stimulate cell proliferation and induce limb regeneration from the denervated stump (Brockes & Kintner 1986, Mescher & Gospodarowicz 1979; Mullen et al. 1996; Wang et al. 2000). Expression of fgf-2 in the SC is increased after tail amputation, and blastema growth is promoted by FGF-2 (Zhang et al. 2000; Ferretti et al. 2001).
Patterning and differentiation-inducing activities from the SC have also been reported in urodeles. If the SC is reversed along the dorsal to ventral (DV) direction before the tail amputation, the regenerated tail has a reversed DV orientation. Implantation of the SC into the tail induces an ectopic muscle and cartilage (Holtzer 1956). A major signaling molecule of such activity is the sonic hedgehog (shh), which is exclusively expressed in the ventral SC of the regenerating axolotl tail and is essential for the proliferation of blastema cells, cartilage differentiation and DV pattering (Schnapp et al. 2005).
The nerve dependency of tail regeneration is believed to be quite different between urodeles and anurans. It has been reported that even if the SC of an anuran tadpole is injured or removed, the tail regeneration is almost normal (Morgan & Davis 1902; Roguski 1957; reviewed in Goss 1969). In past reports the SC was removed surgically by making a large incision or was destroyed mechanically using a needle. However, this surgical procedure can damage tadpoles so seriously that their viability is reduced, while the mechanical procedure can leave a small piece of the SC after operation. A more reliable experimental system should be developed to clarify the dependency of the SC in the regeneration of the larval anuran tail. If SC-dependent regeneration in the Xenopus tail could be described clearly, we could gain a more detailed understanding of the molecular basis for the phenomenon, since tail regeneration in the Xenopus tadpole is the advanced model for genetic studies of organ regeneration (reviewed in Slack et al. 2004; Mochii et al. 2007). To determine whether the SC is required in the normal tail regeneration, we developed a simple method to remove the SC completely from the amputated larval tail with minimal damage, and revealed that the SC ablation impaired proliferation and retarded the differentiation of the notochord (NC) precursor cells. Our approach suggests that the SC of the larval anuran tail plays a more significant role in tail regeneration than previously reported. Furthermore, we showed that application of FGFs partially rescued the defect caused by the SC ablation.
Materials and methods
Animals and surgery
Xenopus laevis larvae were maintained at 22°C in 0.25% NaCl. The developmental stages were determined according to Nieuwkoop & Faber (1956). Tail amputations and manipulations were carried out on tadpoles anesthetized in 0.03% MS222 (Sigma, St Louis, MO, USA). The st.50 tadpole was mounted on wet paper and the tail was amputated vertically at its midpoint using a razor (FA-10; Feather, Osaka, Japan). After the amputation, the remaining tail was incised at the most proximal region from the dorsal side to expose the spinal cord (SC). The SC in the tail between the incision and the amputation plane was then removed from the incision using fine forceps. Control tadpoles were subjected to tail amputation and incision in the same way, but the spinal cord was left in the remaining tail. The operated tadpoles were maintained in 0.25% NaCl containing penicillin (50 U/mL) and streptomycin (50 µg/mL) without feeding for 3 days and with feeding thereafter.
Histological and immunological analyses
For histological analysis, the operated tails were fixed in Bouin's fixative at 4°C overnight, washed, dehydrated through a graded ethanol series and embedded in paraffin. Sections of 6–8 µm were stained with hematoxylin and eosin according to a standard procedure.
For whole-mount immunohistochemical staining, the operated tails were fixed in 4% paraformaldehyde in phosphate buffered saline (PBS) at 4°C overnight and treated with 7.5% H2O2 in ethanol for bleaching. The subsequent procedure of antibody treatment and detection was described previously (Sugiura et al. 2004). Monoclonal antibodies 4d (anti-N-CAM; Watanabe et al. 1986), Xen1 (Ruizi Altaba 1992) and 12/101 (Kintner & Brockes 1984) were obtained from the Developmental Studies Hybridoma Bank at the University of Iowa (Iowa City, IA, USA).
For BrdU labeling, the operated tadpoles were incubated in BrdU solution (1 mg/mL; Sigma) at room temperature for 12 h, fixed in Bouin's fixative and embedded in paraffin. Frontal sections of 6 µm were treated with PBS containing 0.1% triton X-100 (PBST), blocked in PBST containing 10% goat serum for 1 h at room temperature and incubated with anti-BrdU antibody (1 : 200 in the blocking solution; Becton Dickinson, Franklin Lakes, NJ, USA) at 4°C overnight. The sections were then washed in PBST and incubated with Cy3-conjugated antimouse IgG antibody (1 : 500 in PBST; Molecular Probes/Invitrogen, Carlsbad, CA, USA) for 2 h at room temperature. After washing in PBS, nuclei were stained with Hoechst33342.
Section in situ hybridization
Section in situ hybridization was carried out as previously described using BM purple reagent (Roche, Basel, Switzerland) for color detection (Sugiura et al. 2004). Plasmids for Xbra3 (Sugiura et al. 2004), Xfgf-8 (Yokoyama et al. 1998), and Xfgf-10a (Yokoyama et al. 2000) were described previously. Bluescript plasmids encoding Xenopus FGF-2 (xl091e15) and FGF receptor (FGFR)-1 (xl003g02) and -2 (xl039g18) were characterized using the NIBB/NIG/NBRP Xenopus laevis expressed sequence tag (EST) database (XDB3; http://xenopus.nibb.ac.jp/). Complementary DNAs for Xfgfr-3 and -4 (Hongo et al. 1999) were amplified from cDNA of the regenerating tail and cloned into the BamHI/Xho1 site of pCS2 plasmid (Turner & Weintraub 1994). The primers used for the cloning were as follows: Xfgfr-3 (US: 5′-TGCGACTAGAGGATGTCTAAGG-3′; DS: 5′-GGTCATTGACTTCTTGGAAGGG-3′) and Xfgfr-4 (US: 5′-GTGTTTTGCAGGAGACCGTGGA-3′; DS: 5′-GCACAGCTTGAGGAGAGACGAG-3′).
Fibroblast growth factor-2 (Sigma), -8b (R & D Systems, Minneapolis, MN, USA) and -10 (R & D Systems) proteins were dissolved in PBS containing 1% bovine serum albumin (BSA) at a concentration of 1 mg/mL. Affi-Gel blue beads (Bio-Rad, Hercules, CA, USA) were incubated in a drop of the FGF solution for 2 h on ice and then used for the experiment. The amputated and SC-ablated tadpoles were implanted with the FGF-soaked bead by inserting them from the distal cut end of the tail into the meningeal cavity where the SC had been located. A bead incubated with PBS containing 1% BSA were used as the control.
Reverse transcription–polymerase chain reaction
The regenerating tail was cut off vertically at a site about 0.5 mm proximal to the amputation plane. Total RNA was extracted from the sample with TRIzol reagent according to the manufacturer's manual (Invitrogen, Carlsbad, CA, USA). Complementary DNA was synthesized from total RNA (5 µg) using Ready-To-Go You-Prime First-Strand Beads (GE Healthcare, Buckinghamshire, UK). The primers, annealing temperature and cycle number for each polymerase chain reaction (PCR) were as described in Table 1. The reaction conditions were empirically determined in each case within the linear range of amplification.
|Xenopus genes||Sequence||Annealing temperature (°C)||Cycle no.||Reference|
|fgf-2||U: 5′-TCGTTCTAATGTGGTCGATGGAAAT-3′ D: 5′-GACGCCATGATTGTTATGGGTAGTC-3′||60||35||Present study|
|fgf-3||U: 5′-AGATTAACGGGACTCTGGAGAAAAA-3′ D: 5′-TTGTATGTTTCCGAGGCGTAAAGT-3′||60||35||Present study|
|efgf/fgf-4||U: 5′-TCTGCCCAACAACTATAACGCTTACGA-3′ D: 5′-ATGGGTCAGTGTCATTGTCGGAGAA-3′||65||35||Present study|
|fgf-8||U: 5′-TTGGAAGCAGAGTTCGCATTAAAG-3′ D: 5′-ACAATTTCCGAGAAGACGCAGTCT-3′||60||35||Present study|
|fgf-9||U: 5′-TCCCTGGGAAGAGTGGAGACAGTAG-3′ D: 5′-ACCGAAATAGTTGCCAACTTCACCC-3′||65||35||Present study|
|fgf-10||U: 5′-CGGAGCAGGAAACTCTTCTCTTACAC-3′ D: 5′-TGTTAATGGCTTTCACTGCCACAA-3′||60||35||Present study|
|fgf-20||U: 5′-AGAGAGAGCTTTGAGTCACTGTCTGACA-3′ D: 5′-CAGTTTCGGCATGTAGTGACCTTCT-3′||60||35||Present study|
|fgfr-1||U: 5′-TTGCAGTGCAGGCTTCGAGAA-3′ D: 5′-TAACACAGGCATACAGCCCATTGTC-3′||60||35||Present study|
|fgfr-2||U: 5′-GCATCGGATAGGTGGATACAAGGTT-3′ D: 5′-ACGTTCAATGACATCCAAGTGGTAAGT-3′||60||35||Present study|
|fgfr-3||U: 5′-TTGATGTCCTTGAGAGGTCCTCTCA-3′ D: 5′-ACTTCCACGTGTTTAAGCCACTGAA-3′||60||35||Present study|
|fgfr-4||U: 5′-AGTATACAGCGATGCTCAGCCACAT-3′ D: 5′-ATTCCGTAGGTGAAGCACCTCTACCT-3||60||35||Present study|
|Xsox2||U: 5′-TGATGTCAGTGCCCTGCAATATAACTCCAT-3′ D: 5′-GGAGCTGGATTCCGACTTGACTACCGAGCC-3′||55||32||Present study|
|XmyoD||U: 5′-AACTGCTCCGATGGCATGATGGATTA-3′ D: 5′-GATGCTGGGAGAAGGGATGGTGATTA-3′||55||32||Xenopus molecular marker resource|
|Xshh||U: 5′-CTTGGAGGAGTCGCTACATTATGAGGGGAG-3′ D: 5′-CGAACAGTGAATATGAGCTTTGGACTCGTA-3′||55||32||Present study|
|Xbra3||U: 5′-CACCCTCATTATAGGACAGAAACTTGTTTC-3′ D: 5′-GGATCTACTGGTGGACAAAACATTTTCTGT-3′||50||30||Present study|
|Xmsx1||U: 5′-ATGGATCGCACTCCCCTACTGTAACTT-3′ D: 5′-TGCATCCTATTCAAGGGACGTTCTTC-3′||65||30||Present study|
|EF-1α||U: 5′-CAGGCCAGATTGGTGCTGGATATGC-3′ D: 5′-GCTCTCCACGCACATTGGCTTTCCT-3′||60||25||Tazaki et al. (2005)|
The length, diameter and proliferation rate of the regenerating notochord were expressed as the means±standard deviation. The statistical difference was determined by Student's t-test. Values of P ≤ 0.05 were considered statistically significant.
SC ablation of the Xenopus larval tail
We developed a simple operation method using Xenopus laevis to remove the SC completely and to minimize damage to the tadpole (Fig. 1A). At first the distal half of the tadpole tail was cut off as in an ordinary amputation (Sugiura et al. 2004). Then the remaining tail was incised at the most proximal region from the dorsal side to expose the SC. The SC in the remaining tail was finally removed from the incision by pulling it out using fine forceps.
To evaluate ablation of the SC, whole-mount tails were stained with a monoclonal antibody against N-CAM to visualize the larval SC. No fragment of the SC was detected in any operated tail, while a clear staining for the SC was found in a mock-operated control larva (Fig. 1B,C). To confirm complete ablation of the SC in detail, the operated tail was serially sectioned transversally and stained with hematoxylin and eosin (Fig. 1D,E). No SC was found in the operated tail, but several cells, which were thought to be sensory ganglion cells based on their morphology, were found (Fig. 1D). Sections of the SC isolated from the tail showed that an entire SC and most of the associated sensory ganglia were removed by the operation (inset in Fig. 1D). No abnormality was observed in other tissues, including the notochord (NC), muscle and fins. All the SC-ablated and control larvae were alive 15 days after the operation. The morphological and immunological analyses revealed that the SC was completely removed from the larval tail by the operation with no serious damage. The remaining SC in the intact region sometimes regenerated and grew caudally into the SC-ablated region but never reached the amputation plane, even 80 days after the operation (data not shown), showing that the influence of the regenerated SC was minimal for our experimental period.
Regeneration of the SC-ablated tail
To clarify the effect of the SC ablation on the tail regeneration in Xenopus tadpoles, we compared the operated tails with the mock-operated ones using a stereoscopic microscope (Fig. 2A,B). No difference was observed between them until the second day after the operation. Epidermis covered the amputated plane within 24 h and formed multiple layers at the distal portion of both the operated and mock-operated tails on day 2, as in the simple amputation described previously (Fig. 3; Sugiura et al. 2004). On day 3, the regenerated tails formed from the SC-ablated stumps were smaller than those of the controls. The size difference became increasingly obvious at later stages. On the 20th day the regenerated tails of the SC-ablated larvae were about half the length of those of the controls. After that the SC-ablated tails seemed to stop growing, while the control tails continued to grow at least until the 80th day (cf. Fig. 5E,F). In addition to the size difference the SC-ablated regenerated tails were morphologically abnormal. They showed twisted morphology in which the axis of the tail was bent several times, while the control tails were straight (Fig. 2A,B). These observations showed that the SC ablation caused obvious growth and morphological defects in tail regeneration in Xenopus laevis larvae.
Whole-mount observation using Nomarski differential interference contrast (DIC) optics revealed a clear difference in regenerated NC between the SC-ablated and control larvae (Fig. 2C–F). The length of the regenerated NC in the SC-ablated tail was less than half that of the control on day 20 (Fig. 2E). The tail's diameter, measured at the midpoint of the regenerated tail, was also less than half that of the control (Fig. 2F). The diameter of the NC in the SC-ablated tail was nearly equal along the regenerated region at day 4, while the control NC gradually increased in diameter toward the amputation plane (Fig. 2C,D). The axis of the NC was bent several times in the SC-ablated tail in correspondence with the bend in the tail. These observations suggest that the morphological defects in the SC-ablated regenerated tail were mainly caused by the abnormal structure of the regenerated NC.
SC ablation impairs proliferation and retards differentiation of NC precursor cells
Notochord precursor cells migrated caudally and accumulated at the edge of the NC sheath at day 2 and made a cone-shaped cell mass at day 3 (Fig. 3A). Mitotic cells were observed in the cone-shaped NC region after day 2 (data not shown). The reduced length and diameter of the regenerated NC in the SC-ablated tail may have been caused by the reduced number and/or the lowered proliferation rate of the NC precursor cells. The number of NC precursor cells at day 2 was not significantly different between the SC-ablated and control tails (Fig. 3A,B; data not shown), while the cone-shaped cell mass in the SC-ablated tail was smaller than that in the control at day 3. These observations suggested that the proliferation rate of the NC precursor cells was reduced by the SC ablation after day 2. We tested this possibility by labeling the proliferating cells with BrdU at day 2.5 and counting cells after sectioning (Fig. 4). The BrdU-positive cells comprised 81% of the NC precursor cells in the control but 59% in the SC-ablated tail, showing that the SC ablation impaired proliferation of the NC precursor cells during tail regeneration.
Proliferating NC precursor cells elongate transversally and are aligned in the rostral to caudal direction in the cell mass. Cells in the proximal region start to vacuolate and finally differentiate into mature NC cells with enormous cell volumes (Sugiura et al. 2004). NC precursor cells in the SC-ablated tail became elongated and were aligned like the control cells but did not vacuolate on day 5 after amputation when most of the proximal cells in the controls had vacuolated (Fig. 3). Mature vacuolated cells were, however, found in the SC-ablated tail on day 15 (data not shown), showing that the SC ablation did not inhibit the NC cell differentiation completely but did retard its timing.
The monoclonal antibody Xen-1 recognizes most of the nerve cells in both the central and peripheral nervous systems in the Xenopus larval tail (Fig. 5B). A Xen-1 signal was observed in the dorsal and ventral regions of the SC-ablated tail on day 3 after amputation, showing that some peripheral nerves survived without the SC. The dorsal and ventral nerves grew into regenerated tails on day 10 in both the SC-ablated and control larvae with no remarkable difference between the two (Fig. 5A,B).
Myogenic cells derived from muscle satellite cells migrate into the amputated region, proliferate and differentiate to myofibers in the regenerated tail (Gargioli & Slack 2004; Chen et al. 2006). Newly formed myofibers were detected on day 4 with a 12/101 monoclonal antibody in both the SC-ablated and control tails, and their number increased as the regenerated tails grew (Fig. 5C,D). We did not observe a significant difference in the number of myofibers between SC-ablated and control tails at this stage. Regenerated myofibers in the SC-ablated tail were, however, slightly shorter than those in the controls. Some myofibers appeared to degenerate after day 10 in the SC-ablated larvae, and most of them had degenerated by day 80, while control larvae showed regular patterns of myofibers (Fig. 5E,F). Myofiber degeneration was sometimes observed in the muscle region proximal to the amputation plane in the SC-ablated larvae (Fig. 5F′).
The morphological analyses showed that the major defects found in the first 10 days were reduced proliferation and retarded maturation of the regenerating NC precursor cells in the SC-ablated tail. The late defects involved the muscle degeneration, which might have affected the tail morphology indirectly. We therefore focused on early events of the tail regeneration in the following study to eliminate any secondary effects.
Reduced expression of FGFs in the SC-ablated tail
To compare gene expression between the SC-ablated and control tails, the expression of several genes that have been shown to be upregulated during tail regeneration was analyzed by semiquantitative reverse transcription (RT)–PCR (Fig. 6A). Xsox2, which was abundantly expressed in regenerating SC, was not upregulated in the SC-ablated regenerating tail as expected. But no significant difference in expression between the SC-ablated and control tails was observed in other marker genes, including XmyoD for muscle tissue, Xbra3 for notochord and Xmsx1 for blastema (Beck et al. 2003; Sugiura et al. 2004).
It is possible that the impaired proliferation and differentiation of the NC precursor cells is caused by the absence or reduced signaling molecules that are secreted from the SC in the normal tail. Sonic hedgehog (shh) is an essential signaling molecule for tail regeneration in axolotl larvae and is exclusively expressed in the ventral SC (Schnapp et al. 2005). Xenopus shh is abundantly expressed in the regenerating NC but not in the SC (Sugiura et al. 2004). In fact, the expression level of Xshh was not affected by the SC ablation (Fig. 6A). Xshh is therefore not the SC-derived signaling molecule required for proliferation and differentiation of the NC cells, whether it is essential for the tail regeneration or not.
Fibroblast growth factor-2 is another signaling molecule that is expressed in the SC and stimulates cell growth in urodele tail regeneration (Zhang et al. 2000). In this study we focused on members of the FGF family. To identify FGFs abundantly expressed in the SC in the Xenopus larval tail, we compared the expression of FGF genes in regenerating tails between the SC-ablated and control tails (Fig. 6B). Expression of Xfgf-2, -4 (efgf), -8, -9, -10 and -20 was somewhat lower in the SC-ablated tail than in the control tail at day 1 after amputation. RT–PCR using the isolated SC confirmed that these genes were indeed expressed in the SC (Fig. 6C). Spatial expression of Xfgf-2, -8, and -10 was analyzed by in situ hybridization on sections of regenerating tail (Fig. 6D–F). Xfgf-2 was expressed in both proximal and distal regions of the regenerating SC, while Xfgf-8 and -10 were expressed in the proximal SC. All of the fgfs examined were also expressed in the regenerating NC and epidermis. Signals for the fgfs in the SC were lost in the SC-ablated regenerating tail as expected (Fig. 6H–J). These data showed that the regenerating SC expressed FGF genes but was not the only source of these growth factors in the regenerating tail.
The expression of FGF receptor genes was analyzed by RT–PCR and section in situ hybridization. All the fgfrs analyzed were expressed in the regenerating tail, as shown by RT–PCR (Fig. 6B). In situ hybridization revealed that the regenerating NC as well as the SC expressed Xfgfr-1 and Xfgfr-2, the major receptors for FGF-2/4 and FGF-3/10, respectively (Fig. 6K,L). Xfgfr-3, the major receptor for FGF-8/9/20, was abundantly expressed in the regenerating NC but not in the SC (Fig. 6M). Xfgfr-4, the minor receptor for many FGFs, showed no clear signal in the regenerating tail (Fig. 6N). These analyses showed that the regenerating SC expressed several fgfs and that the NC expressed fgfrs. One or more of the FGFs expressed in the SC may play a significant role in normal proliferation and/or differentiation of the regenerating NC cells.
FGFs stimulate NC growth
To determine whether FGFs stimulate proliferation and/or differentiation of the NC precursor cells in the regenerating tail, we implanted a bead soaked with FGF-2, -8b or -10 into amputated stumps of SC-ablated tails (Fig. 7A,B). Measuring the length of the regenerated NC on day 3 and 4 revealed that the NC treated with FGF-10 was longer than the untreated NC (Table 2; data not shown), but FGF-2 and -8b had no effect. The difference in the NC length became unclear after 4 days. Using the BrdU-labeling method described above, we measured the proliferation rate of the NC precursor cells at day 3 and found that proliferation was stimulated by FGF-2 and FGF-10 but not by FGF-8b (Fig. 7C; Table 3). No effect on differentiation of the NC cells was observed for this period. The bead implanting experiment therefore showed that FGF-10 and probably FGF-2 stimulate proliferation of the NC precursor cells and partially rescued the defect caused by the SC ablation.
Spinal cord is essential for proper tail regeneration
The SC-ablated tail caused reduced growth and abnormal morphology of the regenerated tail in our experiments. Roguski showed that destruction of the SC impaired late-stage but not early-stage tail regeneration in Xenopus larvae (Roguski 1954). Our conclusion is that the SC is essential for the proper regeneration of the Xenopus tadpole even in the early stage. We think that the difference in results between our study and Roguski's study can be attributed to the different surgeries used. Roguski tried to eliminate the influence of the SC by repeated insertion of a glass probe from the amputation plane, but this operation can leave small fragments of the SC (Roguski 1954). We ablated the SC from the Xenopus larval tail by a very simple operation, which caused complete removal of the SC and retained high tadpole viability.
The SC dependency in tail regeneration is still quite different between anuran larvae and urodeles, as the SC ablation has been shown to cause no blastema formation in urodeles (Holtzer et al. 1955), while in the present study it caused only reduced growth and morphological defects in Xenopus larvae. Peripheral nerves that remained in the SC-ablated tail and grew into the regenerating tail might have stimulated cell growth. The most likely basis for this difference lies, however, in the varying regeneration strategies between anuran larvae and urodeles. The axolotl larva regenerates cartilage rather than the NC, and the SC is the only source of Shh, which is one of the essential signaling molecules for growth and differentiation of cartilage and myogenic cells (Schnapp et al. 2005). Shh may be an essential factor for tail regeneration in the Xenopus tadpole as well. If so, ablation of the NC rather than the SC should cause a more severe defect in the regenerated tail, because Xenopus shh is expressed in the regenerating NC but not in the SC (Sugiura et al. 2004). This possibility will be explored in future analysis to reveal the roles of the NC and Shh in Xenopus tail regeneration.
FGFs stimulate NC growth
We showed that the signal from the SC included FGF. Many FGF genes are expressed in the SC, and FGF-2 and -10 partially restored the impaired growth of the regenerating NC until day 3 but not after day 4. The reason for the inability of the FGFs to rescue the defect at the late stage might be a technical problem. We inserted FGF-soaked beads into the meningeal cavities where the removed SCs had been located. The beads remained in the proximal region of the regenerated tail and were far away from the actively proliferating region after day 4. Fixing the bead in the distal part of the regenerated tissue might stimulate regenerative growth at a later stage, but such an operation would be very difficult using our experimental system.
Reduced FGFs from the SC might be compensated for by the upregulation of the genes in other tissues. In fact, expression of fgfs in the SC-ablated tail was lower than in the control tail on day 1 but became comparable after day 2 (see Fig. 6B). In situ hybridization revealed that the SC was not the only source of FGFs and that some fgfs were abundantly expressed in the regenerating NC at day 3 (see Fig. 6). These observations suggest that FGFs synthesized by the regenerating NC cell stimulate the NC growth in an autocrine manner. FGFs synthesized by the SC may have a major role in stimulating the NC growth at a very early stage of the regeneration.
Late effect of SC-ablation
The effect of SC ablation became increasingly clear at the later stage of regeneration, as reported previously (Roguski 1954). Most myofibers were degraded in the SC-ablated tail by day 80 (see Fig. 5F), suggesting that maintenance of myofibers is dependent on innervations from the SC. The impaired growth of the NC at the later stage might be an indirect effect of the SC ablation. If growth of the NC cells depends on signals from other tissues which are themselves dependent on innervation, such as muscle tissue, then NC defects should be evident at a later stage. Another possibility is that the SC may be required for the maintenance of NC stem cells. The presence of lineage-restricted stem cells for the NC has been suggested by the lineage tracing of NC cells during tail regeneration (Gargioli & Slack 2004; Mochii et al. 2007) and by the continuous growth of the NC during the overall larval period. If the NC stem cells are lost by SC ablation, the NC should still grow depending on proliferating precursor cells at the early stage but should stop growing after all of the precursor cells finally differentiate. These possibilities should be elucidated by further analyses of the cellular and molecular mechanisms underlying NC growth.
Twisted tail morphology
The impaired proliferation and delayed differentiation of the NC cells did cause growth retardation of the SC-ablated tail, but we have no information about the relation of these defects to the twisted tail, another defect caused by the SC ablation. We could not determine whether the twisted tail was repaired by FGF, because the defect was observed later than day 7 and the FGF-soaked bead could influence only the first several days of growth. Hauser reported that interruption of the SC connection in the Xenopus tadpole resulted in a similar morphological defect of the regenerating tail (Hauser 1969). He then disrupted various parts of the brain and suggested that the normal morphology of the tail required an unidentified factor derived from a part of the diencephalon. At this time, we cannot determine whether the critical signal for the normal tail morphology is secreted from the diencephalon or the SC itself.
We showed in this work that regenerative growth of the NC was partially dependent on the SC, and that the FGF signal was involved in this mechanism. However, FGF was responsible for only a part of the regenerative growth in the Xenopus tadpole tail. Further analysis of FGFs and other signaling molecules will be needed to understand the precise mechanism of tail regeneration.
This work was supported in part by a Grant-in-Aid for Scientific Research from JSPS to MM and by a grant from the Inoue Foundation for Science to YT. We are grateful to Dr K. Tamura (Tohoku University) for provision of the Xfgf-8 and -10 plasmids. We thank Dr Orii and other members of our laboratory for their insightful advice and their help in maintaining the animals.
- 2003. Molecular pathways needed for regeneration of spinal cord and muscle in a vertebrate. Dev. Cell 5, 429–439. , &
- 1986. Glial growth factor and nerve-dependent proliferation in the regeneration blastema of Urodele amphibians. Cell 45, 301–306. &
- 2006. Control of muscle regeneration in the Xenopus tadpole tail by Pax7. Development 133, 2303–2313. , &
- 1998. Cellular and Molecular Basis of Regeneration: from Invertebrates to Humans. John Wiley & Sons, Chichester. &
- 2001. FGF signalling and blastema growth during amphibian tail regeneration. Int. J. Dev. Biol. 45, S127–S128. , , &
- 2004. Cell lineage tracing during Xenopus tail regeneration. Development 131, 2669–2679. &
- 1909. The influence of the nervous system in regeneration. J. Exp. Zool. 7, 643–722.
- 1969. Principles of Regeneration. Academic Press, New York.
- 1969. [Dependence of normal tail regeneration in xenopus larvae upon a diencephalic factor in the central canal.] Wilhelm Roux′ Archiv 163, 221–247 (In German with English abstract).
- 1956. The inductive activity of the spinal cord in urodele tail regeneration. J. Morph. 99, 1–39.
- 1955. An experimental analysis of the development of the spinal column. IV. Morphogenesis of tail vertebrae during regeneration. J. Morph. 96, 145–168. , &
- 1999. FGF signaling and the anterior neural induction in Xenopus. Dev. Biol. 216, 561–581. , &
- 1984. Monoclonal antibodies identify blastemal cells derived from dedifferentiating limb regeneration. Nature 308, 67–69. &
- 2001. Regeneration: nerve-dependent and nerve-independent. In: Encyclopedia of Life Sciences. John Wiley & Sons, Ltd., Chichester. http://www.els.net[doi: 10.1038/npg.els.0001103].
- 1979. Mitogenic effect of a growth factor derived from myelin on denervated regenerates of newt forelimbs. J. Exp. Zool. 207, 497–503. &
- 2007. Tail regeneration in the Xenopus tadpole. Dev. Growth Differ. 49, 155–161. , &
- 1902. The internal factors in the regeneration of the tail of the tadpole. Arch. Entwickelungsmechanik. Organ. 15, 314–318. &
- 1996. Nerve dependency of regeneration: the role of Distal-less and FGF signaling in amphibian limb regeneration. Development 122, 3487–3497. , , , &
- 1956. Normal Table of Xenopus Laevis (Daudin): a Systematical and Chronological Survey of the Development from the Fertilized Egg Till the End of Metamorphosis. Garland, New York. &
- 1954. [Effect of the spinal cord on regeneration of the tail in tadpole Xenopus laevis.] Folia Biol. (Krakow) 2, 189–200 (In Polish with English abstract).
- 1957. [Influence on the spinal cord on the regeneration of the tail in urodele and anuran larvae.] Folia Biol. (Krakow) 5, 249–266 (In Polish with English abstract).
- 1992. Planar and vertical signals in the induction and patterning of the Xenopus nervous system. Development 116, 67–80.
- 2005. Hedgehog signaling controls dorsoventral patterning, blastema cell proliferation and cartilage induction during axolotl tail regeneration. Development 132, 3243–3253. , , &
- 1952. The influence of the nerve in regeneration of the amphibian extremity. Q. Rev. Biol. 27, 169–200.
- 1976. Neurotrophic activity of brain extracts in forelimb regeneration of the urodele, Triturus. J. Exp. Zool. 196, 131–150. , &
- 2004. Cellular and molecular mechanisms of regeneration in Xenopus. Philos. Trans. R. Soc. Lond. B Biol. Sci. 359, 745–751. , , &
- 2004. Differential gene expression between the embryonic tail bud and regenerating larval tail in Xenopus laevis. Dev. Growth Differ. 46, 97–105. , , , , &
- 2005. Macroarray-based analysis of tail regeneration in Xenopus laevis larvae. Dev. Dyn. 233, 1394–1404. , , , , &
- 1994. Expression of achaete-scute homolog 3 in Xenopus embryos converts ectodermal cells to a neural fate. Genes Dev. 15, 1434–1447. &
- 2000. Cloning and neuronal expression of a type III newt neuregulin and rescue of denervated, nerve-dependent newt limb blastemas by rhGGF2. J. Neurobiol. 43, 150–158. , &
- 1986. Topography of N-CAM structural and functional determinants. I. Classification of monoclonal antibody epitopes. J. Cell Biol. 103, 1721–1727. , &
- 1998. Multiple digit formation in Xenopus limb bud recombinants. Dev. Biol. 196, 1–10. , , , &
- 2000. Mesenchyme with fgf-10 expression is responsible for regenerative capacity in Xenopus limb buds. Dev. Biol. 219, 18–29. , , , , &
- 2000. FGF-2 up-regulation and proliferation of neural progenitors in the regenerating amphibian spinal cord in vivo. Dev. Biol. 225, 381–391. , &