Introduction

Organ regenerative ability varies among animal species. Among vertebrates, fish and amphibians possess high regenerative ability, whereas mammals have restricted regenerative ability1,2,3. Xenopus laevis tadpoles have high regenerative ability, and can regenerate functional whole tails within a week of amputation3,4,5. In X. laevis tadpole tail regeneration, the spinal cord and notochord regenerate from the same tissue type in the stump, and myofibers arise from satellite cells and not from the pre-existing myofibers6,7. A recent single-cell resolution analysis of tail regeneration revealed no evidence for the emergence of multipotent progenitors or intermediate cell states suggesting transdifferentiation8. These findings suggest that lineage-restricted tissue stem cells or precursors are major contributors to tail regeneration in X. laevis tadpoles6,7,8,9. After tail amputation, these precursors are activated to generate a mass of undifferentiated proliferating cells at the wound stump to form the regeneration bud, and then differentiate to the tissues from which they are derived6,7. Induction of undifferentiated proliferating cells is an early and indispensable event for organ regeneration10. We previously reported a factor involved in this process; interleukin-11 (il11), which is highly expressed in the regeneration bud, but not in the intact tail or developmental tail bud11, is essential for inducing undifferentiated sensory neurons, notochord, and muscle during tail regeneration12. Immediate early expression of il11 occurs within 2 h post amputation (hpa), and undifferentiated cells of different cell lineages are induced only by il11, suggesting that it initiates regeneration12.

Il11 is a member of the interleukin-6 (Il6) type cytokine family, whose signaling cascade is well described in mammals13,14,15,16,17,18,19. Il11 is received by the Il11 receptor complex comprising Il11 receptor subunit alpha (Il11ra) and glycoprotein 130 (Gp130; also known as interleukin-6 signal transducer, Il6st), a common co-receptor for the Il6 type cytokine family13,14,15,16. Binding of Il11 to the Il11 receptor complex activates Janus kinase (Jak), which is associated with Gp130 and phosphorylates Stat3 (signal transducer and activator of transcription 3), and then phosphorylated Stat3 (P-Stat3) translocates to the nucleus to activate downstream transcription17,18,19.

Il11 activity in tissue regeneration is thought to be elicited by cells receiving the secreted Il11, and thus we hypothesized that Il11 receptor-expressing cells play pivotal roles in regeneration. Although X. laevis has a gene annotated as interleukin-11 receptor subunit alpha on chromosome 1L (il11ra.L), its function as an Il11 receptor has not been investigated. We performed a functional analysis of il1ra.L in culture cells and in vivo, and an expression analysis of intact tadpoles and regenerating tails. We offer a possible model of il11ra.L-expressing cell function in tadpole tail regeneration.

Results and discussion

X. laevis Il11ra.L functions as an Il11 receptor component in cultured cells

First, we investigated whether X. laevis Il11ra.L functions as an Il11 receptor component. Il11 elicits Stat3 phosphorylation and its nuclear translocation in mammals13,14,15,16,17,18,19. We therefore examined whether P-Stat3 nuclear translocation occurs when X. laevis Il11 is applied to X. laevis cells expressing il11ra.L, and whether P-Stat3 nuclear translocation is abolished by il11ra.L knockout (KO).

We utilized the XTC-YF cultured cell line20, derived from X. laevis tadpole. In mammals, signal transduction to Stat3 phosphorylation in response to Il11 occurs via the Il11ra and Gp130 (also known as IL6ST) complex, JAK and STAT313,14,15,16,17,18. The corresponding genes in X. laevis are il11ra.L, il6st.L, il6st.S, jak1.L, jak1.S, jak2.L, jak2.S, stat3.S, and stat3.L, and reverse transcription-polymerase chain reaction (PCR) confirmed that these genes are expressed in XTC-YF (Fig. 1A, S1A). Next, we established 4 lines of il11ra.L KO XTC-YF by transfection of cas9 mRNA and 2 types of guide RNAs (#1 or #2) having different il11ra.L target sites (Fig. S1B, C). We also synthesized recombinant X. laevis Il11 protein by injecting X. laevis il11 mRNA into X. laevis oocytes. For a negative Il11 treatment control, we synthesized recombinant glutathione S-transferase (GST). The mRNAs have a signal peptide sequence and a 3 × flag tag sequence (Fig. S2). Oocytes injected with these mRNAs were cultured, and culture supernatants containing the secreted proteins were collected. Protein syntheses were confirmed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) followed by Coomassie Brilliant Blue (CBB) staining and Western blotting with anti-FLAG antibody (Fig. 1B–D). Both CBB staining and Western blotting detected 3 bands with molecular weights of approximately 23, 26 and 28 kDa in the culture supernatants of oocytes expressing recombinant Il11 (Fig. 1B, C), which were considered to be differently modified forms of Il11.

Figure 1
figure 1

X. laevis Il11ra.L functions as an Il11 receptor component. (A) Gene expression involved in Il11 signaling in the XTC-YF cell line. (B–D) Oocyte culture supernatants were subjected to SDS-PAGE, followed by (B) CBB staining and (C) Western blotting using anti-FLAG antibody and (D) mouse IgG2b isotype control. Size of the synthesized proteins estimated by their amino acid sequences was: Il11, 22.2 kDa; Lif, 25.5 kDa; and GST, 28.4 kDa. Arrowheads in (B) indicate the synthesized proteins. The 66.4-kDa bands (asterisk) in (B) the CBB staining represent BSA contained in the culture medium. (E–Q) Detection of XTC-YF P-Stat3 nuclear translocation by immunostaining. (Top) P-Stat3 signal is shown in red. (Middle) Nuclei were stained with DAPI. (Bottom) Merged images of P-Stat3 and DAPI staining. Data shown are representative of 3 independent experiments (except the experimental condition of Il11 50 ng/mL; n = 2). Scale bars: 50 µm. Raw images of the gel and blots are shown in Fig. S8.

Full size image

We added the oocyte culture supernatants to the medium of the XTC-YF wild-type (WT) and XTC-YF il11ra.L KO cell lines, and performed immunostaining using an anti-P-Stat3 antibody. Nuclear translocation of P-Stat3 was detected when Il11 was added to the WT cells (Fig. 1E–G). The signal intensity increased depending on the concentration of recombinant Il11 added to the WT cell culture medium (Fig. 1E–G), whereas no significant signal was detected when recombinant GST was added (Fig. 1H), strongly suggesting that Il11 treatment induced the nuclear translocation of P-Stat3. On the other hand, when recombinant Il11 was added to the KO cell culture medium, no significant nuclear translocation of P-Stat3 was detected (Fig. 1I–L). To confirm that the KO cells retain responsivity to another ligand that elicits nuclear P-Stat3 accumulation, we performed the same experiments using leukemia inhibitory factor (Lif). Lif is also a member of the Il6 cytokine family, and binds to the Lif receptor complex comprising the Lif receptor subunit (Lifr) and Gp130, which elicits Stat3 phosphorylation and nuclear translocation21. X. laevis has 2 lifr genes, lifr.L and lifr.S, and we confirmed their expression on XTC-YF (Fig. 1A, S1A). We synthesized recombinant X. laevis Lif.L that was Flag-tagged at the C-terminal by injecting the appropriate mRNAs into oocytes (Fig. S2). Protein synthesis was confirmed by CBB staining and Western blotting (Fig. 1B–D). The synthesized Lif-Flag protein had 224 amino acid residues whose predicted molecular size was 25.5 kDa. We detected a smeared band at 35–40 kDa (Fig. 1B, C), probably due to glycosylation because murine Lif is expressed in highly and variably glycosylated forms22,23. We administered the recombinant Lif to the WT and KO cells, and detected nuclear translocation of P-Stat3 in all of them (Fig. 1M–Q), indicating that the KO cells retain the functional signaling machinery required for nuclear translocation of P-Stat3. These results strongly suggest that X. laevis Il11ra.L functions as a Il11 receptor component in XTC-YF.

X. laevis il11ra.L is required for tadpole tail regeneration

We next investigated whether X. laevis Il11ra.L also acts as an Il11 receptor component in tadpole tail regeneration. Knockdown (KD) of il11 impairs tail regeneration in tadpoles12. Thus, we examined whether KD of il11ra.L also impairs tadpole tail regeneration.

We set up 3 experimental groups; (1) embryos injected with cas9 mRNA and il11ra.L guide RNA #1 (#1 KD); (2) embryos injected with cas9 mRNA, and both guide RNAs #1 and #2 to improve KD efficiency (#1&#2 KD); and (3) a control group in which the embryos were injected with only cas9 mRNA. We selected normally developed tadpoles 4 days post fertilization (4 dpf; stage 41) and amputated their tails. A heteroduplex mobility assay24,25 using amputated tails confirmed that the gene editing was successful (Fig. S3). Tail regeneration was evaluated in the tadpoles at 7 days post amputation (dpa). Tail regenerative abilities were significantly reduced in both the #1 KD and #1&#2 KD groups compared with the control group (Fig. 2A, B). The #1&#2 KD group showed or tend to show a significant decrease in tail regenerative ability compared with the #1 KD group. To assess the regeneration defects in il11ra.L KD tadpoles, we measured several parameters of the regenerated tails of cas9 and #1&#2 KD tadpoles (Fig. 2C–J). The regenerated tails in the KD group were significantly shorter than those in the cas9 group (Fig. 2E). In contrast to the axis angle of the regenerated tails in the cas9 group, which ranged from -20 to 20 degrees, the axis angle in the KD group ranged from -40 to 60 degrees, with the dispersion of the axis angle of the KD group being significantly larger than that in the cas9 group (Fig. 2F). Thus, the regenerated tails in the KD group were significantly shorter and had a more bent axis compared with those in the cas9 group (Fig. 2G). We also assessed tissue regeneration by measuring the side view area of the whole regenerated tail, the muscles and the notochord in the regenerated tails, and found that the regenerated tails in the KD group were significantly smaller, and had smaller muscles and notochord (Fig. 2H–J), indicating that the il11ra.L KD affected muscle and notochord regeneration. These results strongly suggested that il11ra.L is necessary for tadpole tail regeneration, and also suggesting that X. laevis Il11ra.L also functions as an Il11 receptor component in tadpole tail regeneration. The ratios of normally developed tadpoles, and of surviving tadpoles after tail amputation did not differ significantly among the cas9, #1 KD, and #1&#2 KD groups (Table S1), suggesting that while il11ra.L is dispensable for development and survival after tail amputation, it has a crucial role in tail regeneration, like il1112.

Figure 2
figure 2

il11ra.L is required for X. laevis tadpole tail regeneration. (A) Typical examples of regenerated tails at 7 dpa of tadpoles injected with either (left) cas9 mRNA only, (center) cas9 mRNA and guide RNA #1, or (right) cas9 mRNA, and guide RNA #1 and #2. Dashed lines indicate the amputation site. Scale bars: 500 µm. (B) Regeneration outcomes of il11ra.L KD tadpoles. Tadpoles at 7 dpa were classified into 4 groups depending on their tail regeneration: Excellent, completely regenerated with all tissues; Good, with all tissues but short or bent tail; Partial, with tissue(s) defect; None, no regeneration. Numbers shown in the bar chart are numbers of tadpoles classified into each group. Statistical significance was assessed by Fisher’s exact test with the Benjamini–Hochberg method, *P < 0.05, **P < 0.01, ***P < 0.001. (C–J) Measurement of regenerated tails. (C) Regeneration outcomes of the KD experiment used for measurement. The same classification criteria as in (B) were used. cas9 group, n = 25; #1&#2 KD group, n = 32. ***P < 0.001 by Fisher’s exact test. (D) Schematic figure of the measurement. (E) Length of the regenerated tails. Bars indicate means. *P < 1 × 10–10 by Welch's t-test. (F) Axis angle of the regenerated tails. The dispersion in the 2 groups was compared by F-test, †P < 1 × 10–10. (G) Combined plot of length and axis angle of the regenerated tails. (H–J) Measured area of (H) whole regenerated tail, (I) muscle and (J) notochord in the regenerated tail. Bars indicate means. *P < 1 × 10–9 by Welch's t-test.

Full size image

To assess whether il11ra.L KD affects other early signaling events of regeneration, we compared gene expression profile changes after tail amputation between the cas9 and il11ra.L KD groups. Tail stumps of tadpoles at 0 and 24 h post amputation (hpa) in the cas9 and il11ra.L KD groups were sampled, and RNA-sequencing (RNA-seq) was performed. We screened those genes that were significantly more upregulated in the 24 hpa cas9 sample compared with the others; i.e. genes for which 1) expression did not differ significantly between the cas9 0 hpa samples and the KD 0 hpa samples, suggesting that their developmental expression was not affected by il11ra.L KD, 2) expression was higher in cas9 24 hpa samples than in cas9 0 hpa samples, suggesting that their expression was upregulated by tail amputation, and 3) expression did not differ significantly between 0 and 24 hpa of the KD samples, suggesting that the upregulation by tail amputation was abolished in the il11ra.L KD tadpoles. The screening identified 6 genes exhibiting the described expression pattern (familywise error rate < 0.05; Table S2, Fig. S4). Notably, wnt5a.L was among the 6 genes. In X. laevis tadpole tail regeneration, wnt5a expression is upregulated at tail stumps after amputation26,27,28, and administration of Wnt5a to the tail trunk induces an ectopic tail28, suggesting function of induce tail regeneration. The RNA-seq suggested that induction of wnt5a expression after tail amputation requires il11ra.L-mediated signaling, emphasizing the important role of Il11 signaling in the early events of regeneration12.

il11ra.L is widely and constitutively expressed in intact tadpoles and regenerating tails

We next performed in situ hybridization studies to investigate the localization of il11ra.L-expressing cells in tissue sections of intact tadpoles and regenerating tails. We analyzed WT intact 4 dpf tadpoles, and WT 1, 3, 5 dpa tadpoles that were amputated at 4 dpf.

In the whole bodies of intact tadpoles, il11ra.L expression was detected in many, but not all cells of many tissues, including the spinal cord, notochord, skeletal muscle, epidermis, and mesenchyme (Fig. 3A), indicating that il11ra.L is expressed constitutively in 4 dpf tadpoles. In the tails of 1 dpa tadpoles, il11ra.L expression was observed in the uninjured site tissues, as well as throughout the wound epidermis (Fig. 3B). In tails of 3 dpa and 5 dpa tadpoles, expression was observed in the uninjured tissue, and the expression pattern in regenerated tissues at the anterior side of the regenerating tail was similar to that in the tail tissues anterior to the amputation (Fig. 3C, D). In addition, stronger signal was observed throughout the regenerating notochord and spinal cord at the tip of the regenerating tail (Fig. 3C, D). These observations correspond to those of the previous quantitative analysis of il11ra.L expression during tail regeneration in which il11ra.L expression was detected at 0 hpa and modestly increased thereafter12, as well as to the result of single cell RNA-seq of intact/regenerating tails8 in which il11ra.L expression was detected in neurons, spinal cord progenitors, floor plate, notochord, myotome, epidermis, and mesenchyme (Fig.S5).

Figure 3
figure 3

il11ra.L is widely and constitutively expressed in the intact tadpole and regenerating tail. (A) In situ hybridization of il11ra.L on a section of intact tadpole (upper) with an antisense probe, and (lower) with a sense probe. Anterior is left and dorsal is up. Scale bars: 500 µm. n = 4. (B–D) In situ hybridization of il11ra.L in a section of a regenerating tail. Anterior is left and dorsal is up. Arrowheads indicate notable signals. Scale bars: 500 µm. High magnification images of boxed areas in (C) and (D) are shown below. Scale bars: 100 µm. 1 dpa, n = 5; 3 dpa, n = 4; 5 dpa, n = 5.

Full size image

il11ra.L KO precursors can also contribute to form regenerated tails

We previously reported that il11 is necessary for tail regeneration, and forced expression of il11 induces the expression of undifferentiated precursor marker genes, indicating that il11 plays a key role in inducting and maintaining undifferentiated precursors for tail regeneration12. This function of il11 is likely mediated via the Il11 receptor that receives Il11 secreted in response to tail amputation. Therefore, we next investigated whether Il11 is received by tissue stem cells or precursors to generate undifferentiated cells (“Direct model”; Il11 directly activates stem cells or precursors), or if it is received by some other cells that subsequently activate tissue stem cells or precursors to generate undifferentiated cells (“Indirect model”; Il11 triggers downstream events that activate stem cells or precursors).

We performed the following experiments (Fig. S6). We generated il11ra.L KD tadpoles by injecting mRNA and guide RNA #1 into 1-cell stage embryos (same as in Fig. 2). The il11ra.L KD embryos were assumed to have developed into mosaics comprising 3 cell types; WT, heterozygously KO (“HT”), and homozygously KO (referred to as “KO”) cells. il11ra.L KD tadpoles had a normal morphology and survival rate (Table S1), and the tail cells were mosaics of the above 3 cell types. The tissue stem cells or progenitors that contribute to form regenerated tail tissues were also assumed to be mosaics. Tail amputation triggers Il11 expression and induces undifferentiated cells. If Il11 directly activates stem cells or progenitors upon tail amputation (“Direct model”), il11ra.L KO progenitors are not activated, and these KO cells do not contribute to form the regenerated tail; thus, regenerated tail tissues mainly comprise il11ra.L WT or HT cells, but not KO cells. On the other hand, if Il11 indirectly activates progenitors via some Il11 receptor-expressing cells (“Indirect model”), il11ra.L KO progenitors are activated like WT or HT cells, and these KO cells contribute to form the regenerated tail. To test this, we generated il11ra.L KD tadpoles and amputated their tails at 4 dpf. Amputated tails were sampled as “developmental tails”. Amputated tadpoles were maintained for 7 days and regenerated tails were sampled as “regenerated tails”. We extracted genomic DNA from these tail samples, and compared the mutation (insertion/deletion) ratio at the guide RNA #1 target site of the il11ra.L genomic locus of developmental tails with that of regenerated tails in the same individual. If il11ra.L KO cells do not contribute to form regenerated tails, the mutation ratio of the regenerated tail would be lower than that of the developmental tail, which contains KO cells.

The mutation ratio of the developmental tails was not significantly different from that of the regenerated tails (Fig. 4), suggesting that some il11ra.L KO precursors contribute to form regenerated tails and that these precursors do not require cell autonomous il11ra.L expression for tail regeneration. Although the mode of Il11 function might differ according to the tissue type, we detected no significant decrease in the mutation ratios, supporting the idea that at least the tissues comprising a large portion of the regenerated tails are formed in a way depicted by the “Indirect model”.

Figure 4
figure 4

il11ra.L KO cells can contribute to form the regenerated tail. Estimated mutation ratio at the guide RNA #1 target site of the developmental and regenerated tail of each individual is shown. Linked points represent the same individual. n = 24. P = 0.44, paired t-test. Representative result of 2 experiments is shown. The result of the second experiment is shown in Fig. S7.

Full size image

Discussion

The present findings demonstrated that il11ra.L, whose function as an Il11 receptor has not been verified experimentally in X. laevis, is required for the nuclear localization of P-Stat3 in response to Il11 in cultured X. laevis cells (Fig. 1), strongly suggesting that il11ra.L encodes an Il11 receptor complex component. This finding also suggests that the receptor complex containing Il11ra.L is the only or at least the major Il11 receptor mediating Il11 signals to phosphorylate Stat3 in cultured X. laevis cells. In the assay, we used recombinant proteins at a concentration of 1800 ng/mL, which is thought to be well above the endogenous levels of most ligands. The P-Stat3 nuclear localization signals were weaker with 300 ng/mL Il11, and hardly detected with 50 ng/mL Il11 (Fig. 1E–G). For example, human umbilical vein endothelial cells29 and rat microglia30 respond to 50 ng/mL Il11. Possible reasons for XTC-YF requiring a higher concentration of Il11 to respond in our experimental conditions are as follows: 1) XTC-YF is less sensitive to Il11, which might be due to the relatively low expression of il11ra.L (Fig. S1A), and/or 2) not all forms of the recombinant Il11 that we detected (Fig. 1B, C) were functional. Even in this situation, we observed clear differences in the response to Il11 between WT and il11ra.L KO cells, strongly suggesting that il11ra.L is essential for Il11 signaling.

il11ra.L KD did not affect the ratios of normally developed tadpoles and surviving tadpoles after tail amputation (Table S1), suggesting that il11ra.L is dispensable for normal development and wound healing after tail amputation. On the other hand, il11ra.L KD significantly impaired tail regeneration (Fig. 2), indicating a specific role of il11ra.L in tail regeneration. These phenotypes are similar to il11 KD12, supporting the idea that il11ra.L is the only or at least a major mediator of Il11 signaling in tadpole tail regeneration. We also observed that il11ra.L KD significantly affected the expression of 6 genes that are induced by tail amputation, including wnt5a (Table S2, Fig. S4). There are several reports of wnt5a function in development, particularly elongation/outgrowth of body parts in several species31,32,33,34, suggesting that one of the wnt5a functions in tail regeneration is elicited during regenerative outgrowth26,27,28. The results suggested that wnt5a induction after tail amputation requires Il11 signaling. It is plausible that a pivotal function of Il11 signaling in regeneration is controlling wnt5a expression and thus regenerative outgrowth, because KD of il1112 or il11ra.L (Fig. 2E) shortened the regenerated tails. Although there are no reports on the function of the other 5 upregulated genes in tail regeneration of X. laevis, it is expected that these genes also have pivotal roles in regeneration. Our results indicated that Il11 signaling mediated by Il11ra.L has a specific and indispensable role in tail regeneration.

Il11 signaling is mediated by Il11ra.L, indicating that il11ra.L-expressing cells play pivotal roles in tail regeneration. We observed constitutive and widespread expression of il11ra.L in most tissues in intact tadpoles and regenerating tails (Fig. 3). Tail amputation triggers il11 expression near the wound stump11,12, suggesting its specific and local role in regeneration. The constitutive and widespread expression of il11ra.L might be beneficial for the response to accidental (i.e., temporally and spatially unpredictable) wounding; il11ra.L-expressing cells throughout the body are ready to respond to Il11 secreted from wounds in any body part. It is unclear, however, whether all or only some il11ra.L-expressing cells have the potential to respond for regeneration.

Although il11ra.L is required for tail regeneration (Fig. 2), we detected no significant differences in the mutation rates of developmental and regenerated tails of il11ra.L KD tadpoles (Fig. 4). These results support the existence of a tissue(s) in which Il11 functions via il11ra.L expressing cells in a non-cell autonomous manner; the functions of Il11 are elicited by factors expressed in cells receiving Il11. Some il11ra.L-expressing cells probably control tail regeneration depending on Il11 secretion via downstream genes of il11ra.L. It remains unclear, however, whether the il11ra.L KO cell contribution is common in all tail tissues or differs according to the tissue type. To address this question, a transgenic frog line in which all il11ra.L KO cells are labeled must be established; at least 1 generation of this long-generation (more than 18 months) animal is required.

In X. laevis, both il1112 and il11ra.L (Fig. 2) are required for tail regeneration. il11 is assumed to be required for heart regeneration in fish, because il11 expression is observed in the regenerating heart, and Stat3 inhibition impairs heart regeneration35. In mouse, on the other hand, il11 enhances fibrosis in the heart, which causes mechanical dysfunction, and genetic deletion of Il11ra1 reduces the fibrosis36. Mouse heart has full regenerative capacity at postnatal days 1–6 (P1-6), but this capacity is lost by P7 with the development of fibrosis after apical resection37. Whereas induction of il11 expression in organ regeneration or wound healing is conserved across animal species, it is possible that the downstream events of Il11, which are elicited via il11ra expressing cells, differ between regeneration-capable or incapable species/developmental stages/tissues. Our study highlights the pivotal role of il11ra in organ regeneration and provides a basis for future studies of the role of Il11 in organ regenerative ability.

Materials and methods

Animals

Xenopus laevis adults were purchased from domestic breeders (Hamamatsu Seibutsu Kyozai, Shizuoka, Japan; Watanabe Zoshoku, Hyogo, Japan) and maintained in fresh water at 20 °C. Animal experiments were carried out in accordance with the Guidelines for Proper Conduct of Animal Experiments of Science Council of Japan. The protocols of experiments using living modified organisms in this study were approved by the Committee on living modified organisms experiments of the Graduate School of Science at the University of Tokyo (DNA Exp. 17–3). Experiments and data analysis procedures were performed in compliance with the ARRIVE guidelines (https://arriveguidelines.org).

Cultured cells

The XTC-YF cell line derived from X. laevis tadpoles20 (RIKEN BioResource Research Center, Ibaraki, Japan) was maintained in 0.5 × L-15 medium for cultured cells (0.5 × Leibovitz's L-15 medium (Merck, Darmstadt, Germany), 10% fetal bovine serum [FBS; BioWest, Nuaillé, France], 2 mM L-glutamine [Wako, Osaka, Japan], and 1% penicillin–streptomycin mixture [Nacalai Tesque, Kyoto, Japan]) at 25 °C.

Reverse transcription-PCR (RT-PCR) and quantitative RT-PCR (qRT-PCR) of XTC-YF

For RT-PCR, total RNA was extracted from XTC-YF using TRIzol Reagent (Thermo Fisher Scientific, Waltham, MA). Reverse transcription was performed using a PrimeScript RT reagent Kit with gDNA Eraser (Takara, Shiga, Japan). A group without reverse transcription (RT-) was also prepared as a control. Sequences of primers used are shown in Table S3.

For qRT-PCR, total RNA was extracted from 3 lots of XTC-YF using RNeasy Mini Kit (QIAGEN, Hilden, Germany). Reverse transcription was performed using a PrimeScript RT reagent Kit with gDNA Eraser. A group without reverse transcription (RT-) was also prepared as a control. Sequences of primers used are shown in Table S4. Realtime-PCR was performed with TB Green Premix Ex Taq II (Tli RNaseH Plus) (Takara) and LightCycler 480 (Roche, Basel, Switzerland). The threshold cycle was calculated using the 2nd derivative maximum method.

Establishment of XTC-YF il11ra.L knockout cell lines

The CRISPR/cas9 system was used. The il11ra.L guide RNAs were designed with CRISPRdirect38. The target sequences were as follows: #1, GGTGAGAAGCACAATCACC[CGG]; and #2, GGGAGCAGCGTCTCAACAC[TGG] (sequences in brackets indicate the protospacer adjacent motif). The target sequences were inserted into a DR274 plasmid39 and guide RNAs were synthesized using an AmpliScribe T7-Flash Transcription Kit (Lucigen, Middleton, WI) before purification with an RNeasy Mini Kit. For cas9 mRNA synthesis, pXT7-hcas940 (China Zebrafish Resource Center, Wuhan, China) was digested with Xba I (Takara) and then mRNA was synthesized using an mMESSAGE mMACHINE T7 ULTRA Kit (Thermo Fisher Scientific, Waltham, MA) and purified with an RNeasy Mini Kit (QIAGEN).

il11ra.L guide RNA #1 or #2 and cas9 mRNA were transfected into XTC-YF cells using Lipofectamine LTX Reagent (Thermo Fisher Scientific). After cloning by limiting dilution, cells of cloned cultures were suspended in 50 mM NaOH at 98 °C for 10 min, and neutralized with 1/10 vo1ume of 1 M Tris–HCl pH 7.5 to obtain genomic DNA extract. Regions including the guide RNA target site were amplified by PCR with primers (sequences are shown in Table S5), and Sanger sequencing with the forward primers was performed. The sequence data were analyzed with CRISP-ID41. Colonies with only 1 or 2 types of base insertions/deletions with frameshifts were determined to be derived from 1 cell with knockout of both alleles of il11ra.L. We established 4 lines of il11ra.L KO XTC-YF (2 lines [KO1 and KO2] with guide RNA #1, and 2 lines [KO3 and KO4] with guide RNA #2). The obtained sequences are shown in Fig. S1B, C.

Synthesis of recombinant Xenopus laevis Il11

Recombinant protein was synthesized by mRNA injection into X. laevis oocytes. The construction procedures are shown in Fig. S2. The il11.L sequence (AB933563) was amplified and the 3 × flag tag sequence was inserted between the il11.L signal peptide and mature peptide sequences (Fig. S2). The resultant N-terminal 3 × flag-tagged il11.L sequence was inserted into a pXT7 plasmid. mRNA was synthesized as described above. For a negative Il11 treatment control, we synthesized recombinant GST, which is hydrophilic and has a molecular weight close to that of Il11. For mRNA of GST, the il11.L signal peptide sequence and 3 × flag tag sequence were added, and the mRNA was synthesized (Fig. S2). We also synthesized recombinant Lif; for lif mRNA, the lif.L sequence was obtained from Xenbase (http://www.xenbase.org/, RRID:SCR_003280) and amplified, the flag sequence was added to the C-terminus, and the mRNA was synthesized (Fig. S2).

Oocyte injection was performed essentially as described previously42. A part of the ovary was excised from an adult X. laevis. Oocyte follicles were removed manually with forceps in 1 × MBS (88 mM NaCl, 1 mM KCl, 0.7 mM CaCl2, 1 mM MgSO4, 5 mM HEPES [pH 7.8], and 2.5 mM NaHCO3), or enzymatically by treating oocytes in 0.5 × L-15 medium for oocytes (0.5 × Leibovitz's L-15 medium [Merck], 0.5 mg/mL bovine serum albumin [BSA; Wako], 2 mM L-glutamine [Wako], and 1% penicillin–streptomycin mixture [Nacalai Tesque]) supplemented with 25 mM HEPES [pH 7.6], 10 mM CaCl2, and 10 mg/ml collagenase (Nacalai Tesque) for 1 h. Oocytes were maintained overnight in 0.5 × L-15 medium for oocytes at 20 °C. We injected 50.6 nl of 1 ng/nl mRNA into oocytes. For the control group, an equal volume of RNase-free water was injected. Injected oocytes were maintained overnight and then the medium was changed. After 3 or 4 days, the oocyte culture supernatants containing the secreted protein were collected.

The supernatants were subjected to glycine-SDS-PAGE and concentrations of synthesized proteins were estimated on the basis of the intensity of CBB stained bands that were detected near the expected molecular size, and bands of a series of known amounts of BSA, using ImageJ. For Western blotting, mouse anti-DYKDDDDK (FLAG) antibody (Wako; clone 1E6) or mouse IgG2b isotype control (BioLegend, San Diego, CA) was used as the primary antibody, and Clean Blot™ IP Detection Reagent (HRP) (Thermo Fisher Scientific) was used as the secondary antibody. The signal was detected by chemiluminescence using ECL select (Cytiva, Tokyo, Japan), and ImageQuant LAS 4000 (Cytiva).

Il11 administration to XTC-YF and immunohistochemistry of phosphorylated STAT3

XTC-YF wild-type and il11ra.L knock-out cell lines were seeded on 96-well plates. After serum starvation in a serum-free medium (0.5 × L-15 medium for cultured cells without FBS) for 4 h, the oocyte culture supernatant containing recombinant proteins was added to the medium. The final concentrations were 1800 ng/ml, 300 ng/ml, and 50 ng/ml for Il11 and GST; and 1800 ng/ml for Lif. After 20 min, cells were fixed with 4% paraformaldehyde/PBS and then permeabilized with cooled methanol. Anti-phospho-STAT3 (Tyr705) antibody (Cell Signaling Technology, Beverly, MA, #9138S) and Alexa Fluor 555-conjugated anti-mouse IgG goat antibody (Invitrogen, Carlsbad, CA, A-21424) were used to stain phospho-STAT3. Nuclei were stained with 1 µg/ml DAPI. Cells were observed using a fluorescence microscope BZ-X800 (KEYENCE, Osaka, Japan).

Generation of il11ra.L KD tadpoles and evaluation of tail regenerative ability

Guide RNAs and cas9 mRNA prepared as described above were used. Injection was performed essentially as described previously12,43,44. Fertilized eggs were dejellied with 3% cysteine and divided into 3 groups; 18.4 nl of (1) cas9 mRNA (700 ng/µl), (2) cas9 mRNA (700 ng/µl) and il11ra.L guide RNA #1 (40 ng/µl), (3) cas9 mRNA (700 ng/µl), il11ra.L guide RNA #1 and #2 (40 ng/µl, respectively) were injected into the animal hemisphere of the egg. Injected embryos were maintained at 12 °C for 6 h and reared at 20 °C. Tadpoles of 4 dpf (stage 41) were anesthetized with 0.02% MS-222 (Sigma, St. Louis, MO) at 0.1 × Steinberg solution (5.8 mM NaCl, 67 µM KCl, 34 µM Ca(NO3)2, 83 µM MgSO4, 300 µM HEPES, pH7.4), and the tails were amputated. Tadpoles with amputated tails were kept in 0.1 × Steinberg solution 20 °C for 7 days, and then classified into 4 groups depending on tail regeneration (Excellent, Good, Partial, and None). Statistical significance was assessed by Fisher’s exact test with the Benjamini–Hochberg method45. For measuring the regenerated tails, we took photos of the regenerated tails of cas9 and #1&#2 KD tadpoles at 7 dpa and the parameters shown in Fig. 2D were measured using Fiji46.

Heteroduplex mobility assay (HMA)

A schematic of HMA is shown in Fig. S3A. Genomic DNA was extracted as described above. The regions including the guide RNA target site were amplified by PCR with primers (sequences are shown in Table S5). PCR products were subjected to electrophoresis using 3% agarose gel containing 45 mM Tris, 1 mM EDTA, and 45 mM boric acid.

RNA-sequencing (RNA-seq) of tail stumps of il11ra.L KD tadpoles

il11ra.L KD (using gRNA #1 and #2) tadpoles and control tadpoles (no gRNAs) were generated as described above. The tails of 4 dpf tadpoles were amputated, and the amputated tail stumps (tail tips were removed) were collected as 0 hpa samples. Tadpoles with amputated tails were kept for 24 h and the tail stumps were collected as 24 hpa samples. We collected 3 lots of 17–24 tail stumps, and total RNA was extracted using the RNeasy Mini Kit. RNA-seq was performed at Genome-Lead Co., Ltd. (Takamatsu, Japan). The mRNA was isolated using KAPA mRNA Capture Kit (NIPPON Genetics, Tokyo, Japan), followed by production of cDNA libraries using MGIEasy RNA Directional Library Prep Set (MGI, Shenzhen, China). RNA-seq was performed using DNBSEQ-G400 RS (MGI) with the DNBSEQ G400RS High Throughput Sequencing Set, to generate approximately 4 × 107 paired-end reads (150 bp × 2) from each cDNA library.

Obtained reads were mapped to X. laevis genome version 9.1 sequence (Xla.v91.repeatMasked.fa) using HISAT2 v2.1.047, and the number of reads mapped to each gene was counted using FeatureCounts v2.0.148 with a gene model (XL_9.1_v1.8.3.2.primaryTranscripts.gff3). The genome sequence and gene model were downloaded from Xenbase. Differential expression analysis was performed using baySeq package v2.26.049.

Estimation of mutation ratios of KD embryo tails

We generated il11ra.L KD tadpoles with guide RNA #1 as described above. Tails of 4 dpf tadpoles were amputated, and amputated tails were sampled as “developmental tails”. Amputated tadpoles were kept separately in 24 well plates (1 tadpole per well) for 7 days, and then regenerated tails were sampled as “regenerated tails”. Genomic DNA was extracted as described above. PCR with primers (sequences are shown in Table S6) was performed, followed by Sanger sequencing with the forward primer. Sequence data were subjected to ICE analysis50 (https://ice.synthego.com/) to estimate the mutation ratio.

In situ hybridization of tissue sections

For RNA probe synthesis, a part of the il11ra.L coding sequence was amplified by PCR with primers (sequences are shown in Table S7) and cDNA from the thymi of X. laevis J strain tadpoles, and subcloned into a pGEM-T easy vector (Promega, Madison, WI). Digoxigenin-labeled RNA probes were synthesized with T7 or SP6 RNA polymerase (Roche) and DIG RNA Labeling Mix (Roche).

Tadpoles were obtained by artificial fertilization and maintained in 0.2% sea salt water at 20 °C. In situ hybridization was performed as described previously12,51 with several modifications. Wild-type intact 4 dpf tadpoles, and wild-type 1, 3, 5 dpa tadpoles which were amputated at 4 dpf, were fixed in MEMFA fixative (0.1 M MOPS (pH7.4), 2 mM EGTA (pH8.0), 1 mM MgSO4, and 10% formalin) at room temperature for 48 h. They were dehydrated with an ethanol and Clear Plus (Falma, Tokyo, Japan) series and embedded in Paraplast plus (Merck), and 7-µm thick sagittal sections were prepared. Sections were rehydrated, permeabilized with 20 µg/ml proteinase K (Roche) and 1% Triton X-100, followed by acetylation with 0.25% acetic anhydride. The digoxigenin-labeled probe were hybridized at 58 °C for 20 h, and washed with 0.3 × SSC (4.5 mM trisodium citrate, 45 mM NaCl, pH7.0). The slides were incubated with alkaline phosphatase conjugated anti-digoxigenin antibody (Roche). Signals were detected with NBT/BCIP stock solution (Roche) for 24 h. To bleach the endogenous melanin pigments, slides were placed in 0.5 × SSC containing 0.3% H2O2 under fluorescent light overnight. The slides were observed using a differential interference microscope.