Xenopus laevis il11ra.L is an experimentally proven interleukin-11 receptor component that is required for tadpole tail regeneration

Xenopus laevis tadpoles possess high regenerative ability and can regenerate functional tails after amputation. An early event in regeneration is the induction of undifferentiated cells that form the regenerated tail. We previously reported that interleukin-11 (il11) is upregulated immediately after tail amputation to induce undifferentiated cells of different cell lineages, indicating a key role of il11 in initiating tail regeneration. As Il11 is a secretory factor, Il11 receptor-expressing cells are thought to mediate its function. X. laevis has a gene annotated as interleukin 11 receptor subunit alpha on chromosome 1L (il11ra.L), a putative subunit of the Il11 receptor complex, but its function has not been investigated. Here, we show that nuclear localization of phosphorylated Stat3 induced by Il11 is abolished in il11ra.L knocked-out culture cells, strongly suggesting that il11ra.L encodes an Il11 receptor component. Moreover, knockdown of il11ra.L impaired tadpole tail regeneration, suggesting its indispensable role in tail regeneration. We also provide a model showing that Il11 functions via il11ra.L-expressing cells in a non-cell autonomous manner. These results highlight the importance of il11ra.L-expressing cells in tail regeneration.

Organ regenerative ability varies among animal species. Among vertebrates, fish and amphibians possess high regenerative ability, whereas mammals have restricted regenerative ability [1][2][3] . Xenopus laevis tadpoles have high regenerative ability, and can regenerate functional whole tails within a week of amputation [3][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 myofibers 6,7 . A recent single-cell resolution analysis of tail regeneration revealed no evidence for the emergence of multipotent progenitors or intermediate cell states suggesting transdifferentiation 8 . These findings suggest that lineage-restricted tissue stem cells or precursors are major contributors to tail regeneration in X. laevis tadpoles [6][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 derived 6,7 . Induction of undifferentiated proliferating cells is an early and indispensable event for organ regeneration 10 . 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 bud 11 , is essential for inducing undifferentiated sensory neurons, notochord, and muscle during tail regeneration 12 . 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 regeneration 12 .
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 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 tadpoles 12 . 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 assay 24,25 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   www.nature.com/scientificreports/ 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 il11 12 .
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 (RNAseq) 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 amputation [26][27][28] , and administration of Wnt5a to the tail trunk induces an ectopic tail 28 , 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 regeneration 12 . 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 thereafter 12 , as well as to the result of single cell RNA-seq of intact/regenerating tails 8 in which il11ra.L expression was detected in neurons, spinal cord progenitors, floor plate, notochord, myotome, epidermis, and mesenchyme (Fig.S5).
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 regeneration 12 . 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".

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 cells 29 and rat microglia 30 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 www.nature.com/scientificreports/ (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 KD 12 , 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 species [31][32][33][34] , suggesting that one of the wnt5a functions in tail regeneration is elicited during regenerative outgrowth [26][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 il11 12 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 stump 11,12 , 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 il11 12 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 regeneration 35 . In mouse, on the other hand, il11 enhances fibrosis in the heart, which causes mechanical dysfunction, and genetic deletion of Il11ra1 reduces the fibrosis 36 . 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 resection 37 . 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

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. 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-ID 41 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. xenba se. org/, RRID:SCR_003280) and amplified, the flag sequence was added to the C-terminus, and the mRNA was synthesized (Fig. S2).

Establishment of XTC-YF
Oocyte injection was performed essentially as described previously 42  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. www.nature.com/scientificreports/ 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 serumfree 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 previously 12,43,44 . 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 45 . 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 Fiji 46 . 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.

Heteroduplex mobility assay (HMA). A schematic of HMA is shown in
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 × 10 7 paired-end reads (150 bp × 2) from each cDNA library.
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 analysis 50 (https:// ice. synth ego. 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).