Salamander-like tail regeneration in the West African lungfish

Abstract

Salamanders, frog tadpoles and diverse lizards have the remarkable ability to regenerate tails. Palaeontological data suggest that this capacity is plesiomorphic, yet when the developmental and genetic architecture of tail regeneration arose is poorly understood. Here, we show morphological and molecular hallmarks of tetrapod tail regeneration in the West African lungfish Protopterus annectens, a living representative of the sister group of tetrapods. As in salamanders, lungfish tail regeneration occurs via the formation of a proliferative blastema and restores original structures, including muscle, skeleton and spinal cord. In contrast with lizards and similar to salamanders and frogs, lungfish regenerate spinal cord neurons and reconstitute dorsoventral patterning of the tail. Similar to salamander and frog tadpoles, Shh is required for lungfish tail regeneration. Through RNA-seq analysis of uninjured and regenerating tail blastema, we show that the genetic programme deployed during lungfish tail regeneration maintains extensive overlap with that of tetrapods, with the upregulation of genes and signalling pathways previously implicated in amphibian and lizard tail regeneration. Furthermore, the lungfish tail blastema showed marked upregulation of genes encoding post-transcriptional RNA processing components and transposon-derived genes. Our results show that the developmental processes and genetic programme of tetrapod tail regeneration were present at least near the base of the sarcopterygian clade and establish the lungfish as a valuable research system for regenerative biology.

Keywords: evolution; lungfish; regeneration; tail; tetrapod.

Figure 1.

Morphological characterization of tail regeneration in the West African lungfish. (a) Progression of lungfish tail regeneration and the extent of growth up to 56 dpa. Vertical bars in the graph represent standard deviation. (b) Histological sections of regenerating lungfish tail. (c) The regeneration of skeletal elements of the tail at 60 dpa. (d) BrdU staining of proliferative cells during tail regeneration. We, wound epithelium; aec, apical epithelial cap; bl, blastema; et, ependymal tube; ptc.c, postcaudal cartilage; ns, neural spine; na, neural arch; hs, haemal spine; ha, haemal arch. Scale bars of 1 cm (a), 1 mm (b,d), 0.5 cm (c). (Online version in colour.)

Figure 2.

Establishment of dorsoventral organization and the requirement for Shh signalling during lungfish tail regeneration. (a) Histological transversal sections of uninjured and 21 dpa regenerating tail. (b) Immunostaining of DAPI and phalloidin in the uninjured tail, and DAPI and MHC in the 28 dpa regenerating tail. (c) Immunostaining of DAPI and βIII-tubulin in uninjured and 28 dpa regenerating spinal cord. (d) The effect of DMSO and cyclopamine treatment in tail regeneration. m, muscle; ptc.c, postcaudal cartilage; et, ependymal tube; bv, blood vessel. Scale bars of 1 mm (a and b, panoramic views), 0.5 mm (c, enlarged view) and 1 cm (d). Arrowheads indicate point of amputation and bars in graph represent standard deviation (d). (Online version in colour.)

Figure 3.

Upregulated genes and over-represented pathways in lungfish tail blastema relative to uninjured tail. (a) Volcano plot showing differentially expressed genes in lungfish uninjured tail tissue and 14 dpa tail blastema (FDR < 0.05, FC > 2), selected lungfish genes up or downregulated in the blastema are noted as black dots. (b) Pathways over-represented in the tail blastema. (c) Heatmap denoting subset of upregulated genes. (d) In situ hybridization of genes upregulated in the blastema. (e) Area-proportional Venn diagram showing commonly upregulated genes in lungfish tail and pectoral fin blastema datasets, enriched pathways in the shared tail and pectoral fin blastema dataset, and pathways enriched exclusively on tail blastema. (f) Transposon-derived genes upregulated in the tail blastema. (g) Genes coding for serine-arginine (SR)-rich proteins upregulated in the tail blastema. Scale bars of 1 mm (panoramic views) and 0.25 mm (enlarged view). In (c), ‘max’ and ‘min’ represent maximum and minimum expression levels of each gene. (Online version in colour.)

Figure 4.

Hypothesis for the evolution of tail regeneration in sarcopterygians. Regeneration-incompetent lineages are shown in black, lineages with one or more regeneration-competent species are shown green, orange denotes de novo appearance of tail regeneration in Lepidosauria; green arrowhead indicates earliest occurrence of tail regeneration, black arrowhead indicates earliest loss and orange arrowhead, reemergence. Cross signifies extinct taxon. Phylogeny from Fröbisch et al., 2015 [1] and Amemiya et al., 2013 [55]. (Online version in colour.)