The Phylogenetic Distribution of Regeneration
The ability to regenerate structures lost to damage varies widely across and within Metazoan phyla (Sanchez Alvarado 2000; Brockes and Kumar 2008; Bely and Nyberg 2010). This variation is far from random, and adequate knowledge about the distribution patterns of regeneration in animals can help understand both the proximal causes (i.e., the cellular and molecular mechanisms of regeneration) and the ultimate causes (i.e., the evolutionary forces acting on regeneration) of this variation.
An overview of regenerative abilities across metazoans suggests that the earliest animals were good at regenerating (Figure 10.1). Phyla radiating near the base of the Metazoan tree, like Porifera (sponges), Cnidaria (jellyfishes, corals, sea anemones,
FIGURE 10.1 Phylogenetic distribution of regenerative abilities across metazoans. Ability to regenerate appendages (whenever appendages are part of a body plan) is shown by upward- pointing triangles to the left. Ability to regenerate parts of the main body axis are shown as four increasing levels: (1) only terminal structures regenerate, (2) limited extent of subterminal structures regenerate, (3) large extent of structures regenerate, and (4) whole bodies can regenerate from small fragments. The extent of known regenerative capabilities for each phylum is shown based on the highest level known for species within the phylum and may not reflect the typical abilities for the phylum or its ancestral condition. Phylogenetic relationships after Laumer et al. (2019). Modified after Zattara (2012).
box jellies, and many others), and Ctenophora (comb jellies), are all characterized by extensive regenerative abilities (Bely and Nyberg 2010; Morgan 1901; Slack 2017). Systematic surveys are still lacking, but high regenerative abilities are likely part of the ancestral abilities of these phyla, suggesting they were also present in the common ancestors of all animals. The same goes for groups originating near the base of the Bilateria: species of Placozoa and Xenacoelomorpha all show excellent regeneration (Thiemann and Ruthmann 1991; Martinelli and Spring 2004; Haszprunar 2016), supporting that the developmental toolkit of the Urbilateria, the last common ancestors of all bilaterians, included ample regenerative abilities.
Among the Deuterostomia, extensive regenerative abilities are reported for many species of Echinodermata (sea urchins, starfishes, sea cucumbers, and sea lilies), Hemichordata (acorn worms) and Urochordata (sea squirts and salps), and more modest abilities—appendage and terminal axial regeneration—are known from Cephalochordata (lancelets) and Craniata (the vertebrates) (Tsonis 2000; Bely and Nyberg 2010). This phylogenetic pattern suggests that regenerative abilities inherited from Urbilateria were retained in the deuterostome ancestor.
In contrast, all phyla within Ecdysozoa (which includes nematodes, tardigrades, onychophorans, arthropods, and other groups characterized by a periodically molted external cuticle) show poor axial regeneration (Bely 2010), suggesting this ability was lost during the evolution of the last common ecdysozoan ancestor. Interestingly, poor axial regeneration in ecdysozoans does not imply a lack of regenerative mechanisms: for example, imaginal discs of holometabolous insects can readily regenerate lost parts (Worley, Setiawan, and Hariharan 2012). Furthermore, many arthropods are capable of regenerating their articulated limbs (Maruzzo and Bortolin 2013); since arthropod limbs represent a novel structure relative to the ecdysozoan stem, it is likely that arthropod appendage regeneration represents an evolutionary gain rather than a plesiomorphy inherited from the Urbilateria.
The third main clade of animals, the Spiralia, comprises about 13 phyla including rotifers, flatworms, mollusks, nemerteans, brachiopods, phoronids, bryozoans, and annelids, among others (Edgecombe et al. 2011). Recent works on spiralian phylog- eny propose that this clade is comprised of Gnathifera (Rotifera, Acanthocephala, and Gnathostomulida), Rouphozoa (Gastrotricha and Platyhelminthes) and Lophotrochozoa (Cycliophora, Ectoprocta, Mollusca, Nemertea, Annelida Brachiopoda, Phoronida, and Bryozoa) (Laumer et al. 2019). Spiralian regenerative abilities vary widely, from the almost total absence of regeneration in Rotifera to the whole-body regeneration seen in some species of flatworms, annelids, and nemerteans (Bely, Zattara, and Sikes 2014). Within Gnathifera, some rotifers and acanthocephalans can regenerate appendages, but there are no reports of axial regeneration; there is no information about regenerative capabilities of Gnathostomulida. Among Rouphozoa, both axial and appendage regeneration has been reported in species of Gastrotricha (Manylov 2010), but most of the attention has been given to the outstanding regenerative abilities of many species of Platyhelminthes. Several groups of flatworms can regenerate whole animals from small fragments, an ability most intensely studied in species from Tricladida, but also reported from other groups, including the basally branching catenulids. Lophotrochozoan phyla show substantial within-group variation in their regenerative ability, but overall display good to outstanding regeneration. Mollusks lack whole-body regeneration, but several species have been reported to regenerate various axial and appendicular structures, including the head and limbs (Bely, Zattara, and Sikes 2014). Nemerteans show widespread ability for posterior axial regeneration, but the ability to regenerate the head is rare and seems unlikely to represent the ancestral condition for the phylum (Zattara et al. 2019). Brachiopods can regenerate their lophophore, shell, and pedicle (Chuang 1994), but given their body plan, it is difficult to determine if reconstruction of any of these organs represent axial regeneration. In contrast, pho- ronids and ectoprocts (bryozoans) have several species capable of whole-body regeneration (Emig 1973; Nielsen 1994). Annelids also show great axial and appendage regenerative ability, although many groups seem to have lost the ability to regenerate their anterior ends and even the posterior ends (Bely 2006; Zattara and Bely 2016). Overall, spiralians show abundant examples of good structure regeneration, suggesting that their last common ancestor had good to great regenerative ability. However, systematic surveys and formal phylogenetic analysis and ancestral trait estimation of regeneration have only been conducted in two phyla, Annelida and Nemertea (Zattara and Bely 2016; Zattara et al. 2019); these studies have shown that species showing outstanding regenerative ability might not be representative of all other species of the phylum, or of their last common ancestor.
Of all metazoan phyla mentioned earlier, only three are organized in true segments: the deuterostome Craniata, the ecdysozoan Panarthropoda, and the spira- lian Annelids. Segmental organization along the main, anteroposterior body axis in all three groups is achieved by generation of iterated modules at the posterior end during embryonic development (see Chapters 2 to 5) and in many cases during post-embryonic growth. Upon amputation transverse to the main body axis, if axial regeneration occurs, it usually restores the missing ends including the terminal regions and a variable number of intervening segments. This implies that after amputation, axial regeneration is achieved by restoring the terminal regions and rebooting the segmentation process.