Annelids have a relatively well-conserved nervous system morphology (Figure 10.5A) (Bullock and Horridge 1965; Purschke 2015; Helm et al. 2018). The central nervous system (CNS) is characterized by an anterodorsal cerebral ganglion or brain, and a ventral nerve cord (VNC) that runs along the whole body down to the posterior end. In many groups, the nerve cord has a clearly segmented organization reminiscent of a rope ladder, with iterated sets of transverse connectives linking two longitudinal
FIGURE 10.5 Annelid nervous system ground pattern, and regeneration of neuronal and muscular components. (A) Diagrammatic representation of a typical annelid nervous system. Brain and ventral nerve cord ganglia are shown in blue, and simplified nerves and neural tracts shown in green. Non-segmental terminal regions shown in shades of brown, and posterior growth zone (pgz) shown in graded dark red. (B-I) Neural development during anterior and posterior regeneration in the naidid clitellate Dero (Aulophorus) furcata. Maximum intensity projection of confocal laser scanning microscopy (CLSM) Z-stacks of acetylated alpha-tubulin immunoreactive structures (axons of the nervous system, gut ciliation. and nephridia; green), serotonin immunoreactive structures (central nervous system axons and perykaria; yellow), and DAP1 as a nuclear counterstain (DNA; blue). Filled arrowheads show neurites extended from the old ventral nerve cord; empty arrowheads show the anterior neural loop that becomes the new circumenteric connectives; arrows point at axonal extensions of the peripheral nerves over the blastema. (J-Q) Muscle development during anterior and posterior regeneration in the naidid clitellate Dero (Aulophorus) furcata. Color-coded projections of CLSM Z-stacks of Alexa-Fluor 488 phalloidin showing muscle development. In B-K, color-coded bars indicate new tissues: brown, prostomium and peristomium; orange, pygidium; graded dark red, new posterior growth zone; solid green, undifferentiated tissue; striped green, developing segments. A modified after Zattara and Bely (2015); B-Q modified after Zattara (2012).
sets of neurites. The number and arrangement of longitudinal connectives that form the neuropil of the VNC varies across annelid groups, but it might include an unpaired median connective running along the ventral midline, inner paired paramedian connectives, and outer paired main connectives (Muller 2006). The dorsal brain and ventral cord are connected by paired sets of circumenteric connectives that circumvent the foregut. Connectives are usually seen as single bundles of nerves, but sometimes they split before reaching the brain in two roots, one ventral and one dorsal. At the anterior end, a variable number of prostomial nerves branch off the circumenteric connectives and arborize to innervate the sensory organs at the front of the worm. Similarly, peripheral nerves branch off at the posterior end of the worm to innervate the pygidial structures. Within the segmental region, bilateral pairs of peripheral segmental nerves branch off perpendicularly from the VNC, go through the body wall, and then extend dorsally, sometimes reaching the dorsal midline and forming transverse rings. These peripheral nerves usually innervate the parapodia and numerous epidermal sensory structures. Peripheral nerves are segmentally iterated, although some groups show a reduction in the number of nerves per segment at the anteriormost segments; furthermore, these patterns are conserved within taxonomic groups but vary across the phylum (Zattara and Bely 2015; Purschke 2015).
Early studies on the regeneration of the annelid nervous system were based on microscope observation of histological sections labeled using traditional staining techniques; newer studies have taken advantage of immunohistochemical detection of specific molecular components of neurons, such as acetylated a-tubulin, serotonin, and FMRF-amide-like peptides, using fluorescent tags and laser scanning confocal microscopy (Purschke 2015).
After amputation, the process of regeneration needs to restore the nervous structures at the new terminal regions and new intercalated segments, including a new stretch of ventral nerve cord and terminal and segmental peripheral nerves. Anterior regeneration additionally implies growing a new brain. In most species where the process has been studied, neural regeneration is first evidenced by early invasion of the wound site by neurites originating from the old ventral nerve cord (Figure 10.5B and F, filled arrowheads). At least some of these neurites are axonal extensions of existing neurons: labeling experiments using the cell membrane-diffusible tracer Dil in the clitellate Pristina leidyi show that axons from neurons labeled at the nerve cord ganglion proximal to the wound site invade the blastema and even become part of the circumenteric connectives linking to the regenerated brain (Figure 10.4I-K) (Zattara 2012).
During anterior regeneration, neurites extending toward the anterior end from the amputated nerve cord fuse in a series of loops (Figure 10.5C, empty arrowhead). In most species, paired sets of neurites extend from the paramedian and main connectives after wound healing (Yoshida-Noro et al. 2000; MUller, Berenzen, and Westheide 2003; Muller and Henning 2004; Muller 2004; Zattara 2012; Weidhase, Bleidorn, and Helm 2014; Weidhase, Helm, and Bleidorn 2015; Weidhase et al. 2017). As the anterior blastema forms, each set merges at the anterior tip forming a loop; the medial loop formed by paramedian neurites usually forms first, while the lateral loop formed by main neurites closes later. In the only two species of cirratulids studied so far, Cirratulus cf cirratus and Timarete cf. punctata, each main neurite extension instead forms a separate lateral loop, resulting in a trefoil-like structure; lateral loops later fuse with the medial loop. Some species also extend neurites from an unpaired median connective (Mtiller and Henning 2004; Weidhase et al. 2017; Kozin, Filippova, and Kostyuchenko 2017), but they do not participate in the distal loops, perhaps due to the suggested role of the median nerve in innervating the longitudinal musculature (Mtiller 2006). At this stage, anterodorsal epidermal cells from the prospective prosto- mium become internalized and congregate around the distal end of the loops, forming the brain anlagen (Figure 10.5D and E, asterisk), while the lateral sides of the loops fuse forming the circumenteric connectives (Figure 10.5D and E, empty arrowhead). In groups where the circumenteric connectives split in ventral and dorsal roots, neurites in the ventral root derive from the medial loop, while neurites in the dorsal root derive from the lateral side of the loop (or lateral loops, in cirratulids). The nerve plexus that innervates foregut structures of many annelids also develops around this time. Development of the prostomial/pygidial CNS precedes development of the segmental features of the ventral nerve cord; these features, which include nerve cord ganglia and peripheral nerves, form as the blastemal tissues that are intercalated between the regenerated peristomium and the anterior boundary of the old tissue differentiate into segments (Ozpolat and Bely 2016; Weidhase et al. 2017). The most likely source of blastemal cells fated to form the ganglia is ingression of cells from the lateroventral epidermis, but this cell origin still needs to be experimentally verified through cell tracing or live time-lapse microscopy.
During posterior regeneration, neurites initially invade the distal end branch and develop into the characteristic innervation of the rudiments of pygidial structures (Figure 10.5H, filled arrowheads); thus, differences across species reflect differences in pygidial morphology. In general, regeneration of pygidial neural structures precedes development of species-specific segmental features of the nervous system, which occurs at later stages, after the formation of a new segment addition zone between the regenerated pygidium and the old tissue. These segmental features, which include segmental ganglia of the nerve cord and associated peripheral nerves, develop in the same way as during normal posterior growth; the rate of segment addition, however, is usually much higher during regeneration than during growth.
In many annelids, peripheral nerves from the proximal segment have been reported to extend numerous subepidermal axonal extensions toward the w'ound site and form a neural plexus over the developing blastema (Figure 10.5B-D, F-H, filled arrowheads). In contrast to neurites extending from the ventral cord, these axons are seen only during the regeneration process and disappear during its later stages. While neurites extending from the VNC have a clear role in building the new CNS, the function of the temporary neural plexus formed by neurite outgrowth from peripheral nerves is unknown. Interestingly, earlier studies found this plexus only in clitellate annelids and concluded that its presence was a clitellate-specific regeneration trait (Mtiller 2004; Zattara and Bely 2011). However, more recent studies have since described a similar phenomenon for a non-clitellate Sedentaria, Capitella teleta (Jong and Seaver 2016, 2017) and two Errantia, Platynereis dumerilii and Alitta virens (Kozin, Filippova, and Kostyuchenko 2017; Planques et al. 2018). Although variations at a fine taxonomic scale cannot be discarded, it is also possible for this trait to have been overlooked by earlier studies. Indeed, most reports that do not find formation of a peripheral plexus also fail to explicitly state its absence (Ozpolat and Bely 2016). Furthermore, peripheral nerve neuronal extensions are usually much finer and harder to detect than outgrowths from the VNC. Thus, it is quite possible that variation in the reports of the presence of a transient peripheral plexus during annelid regeneration results more from observation bias than actual phylogenetic variability of this trait (Kozin, Filippova, and Kostyuchenko 2017).
Two alternative, non-mutually exclusive hypotheses have been proposed for the role of the transient peripheral nerve plexus (Jong and Seaver 2017): one states that the plexus promotes cell proliferation and patterning of the blastema (Herlant- Meewis 1964; Muller, Berenzen, and Westheide 2003; Varhalmi et al. 2008); the other proposes that the axons of the plexus serve as cell migration tracks (Stephan-Dubois 1954; Cornec et al. 1987). Although data are still inconclusive, the relative timing of plexus formation and cell proliferation during regeneration supports a role in promoting proliferation (Jong and Seaver 2017). In contrast, the only available study tracking cell migration in vivo shows most cell migration taking place in the coelomic cavity but rarely near the surface of the body wall, which is where most of the neurites forming the plexus are located (Zattara, Turlington, and Bely 2016). Nonetheless, future experimental studies combining live cell labeling and tracing of both migrating cells and neural components with laser ablation of either cells or neurites might shed light on the function of the peripheral nerve plexus.
Expression of genes related to neural fate determination and neurogenesis has been described only for posterior regeneration in Platynereis dumerilii (Planques et al. 2018). The pro-neural gene elav, encoding an RNA-binding protein that promotes neuronal fates in metazoans (Pascale, Amadio, and Quattrone 2007), is expressed early on at the ventralmost part of the wound epithelium, next to the cut end of the ventral nerve cord. During blastema formation, its expression domain extends to many ventral cells, and splits into a pygidial domain and a segmental domain, separated by the newly formed segment addition zone. Neurogenin (ngn), another regulator of neuronal differentiation (Seo et al. 2007), is first expressed during the wound-healing stage at paired sets of cells located on the lateral body wall of the stump, posterior to the parapodia. After blastema formation, expression expands to more ventral cells located at the developing pygidium and segments. Pygidial expression disappears at later stages but remains at the posterior growth zone. Early expression of the axonal guidance gene slit has a salt-and-pepper pattern, but once new segments begin to differentiate, it is expressed at the ventral midline and as ventrolateral stripes. Expression of the neurogenic transcription factor рахб, which precedes ngn, slit, and elav during Platynereis larval development (Denes et al. 2007), is not observed at early stages of regeneration but is later seen as two longitudinal ventrolateral bands at the posterior growth zone. Neurogenic gene expression patterns thus support that despite some shifts in the order of gene expression, neurogenesis proceeds through the same processes in regeneration and embryonic/larval development.