The brain and major neuroanatomical properties in Terebelliformia
Based on our comprehensive investigations, the terebellid Lanice conchilega and the ampharetid Hypania invalida possess circumoesophageal connectives (cc) which are similarly connected to the brain by two strands (roots), a ventral and a dorsal one. These strands or roots of the cc show a clear serotonin-like immunoreactive signal, but cannot be seen in Azan-sections. Due to their general location and the patterning in anti-serotonin-staining, we assume them as being homologous to the two roots former authors described using histological sections (Orrhage and Müller 2005). Anatomical differences, in terms of course and extension of the mentioned roots in both investigated taxa, are caused by morphological transitions related to adaptive changes of the entire anterior end in Terebelliformia. Although both roots of the cc are closer associated in H. invalida than in L. conchilega, the transition of the latter into the two paired neurite bundles of the ventral nerve cord is comparable when observing their serotonin-like immunoreactivity. A closer examination of the neuroanatomical characteristics of both taxa shows many similarities in this respect. In anterior direction, the cc fuses with the dorsal, more prominent region of the brain. Antero-ventral to the dorsal brain region, two connectives - one splitting from the cc (connective of ventral brain region-cvbr) and another proceeding from the dorsal brain region (cdbr) - fuse by forming the ventral, more slender region of the brain. Thereby, the ventrally-oriented brain region splits up distally and forms a distinct ventral and dorsal root, which both innervate the buccal tentacles in H. invalida and L. conchilega.
Contradictory, earlier investigations focussing on Terebelliformia describe a unified, ribbon-like brain (Orrhage 2001) and postulate the ventral and dorsal connective (cvbr and cdbr) as the two roots of the “common tract” – the latter not being part of the brain. Instead, both roots were described to innervate the lateral part of the “tentacular membrane” and the buccal tentacles.
Nevertheless, our data reveal a protrusion of a slender brain region in ventral direction. This ventral region exhibits two neuronal roots and innervates the buccal tentacles in Terebelliformia. The investigations presented herein demonstrate that nerves of the so-called “tentacular membrane” (see Orrhage 2001) are in fact loop-like neurite bundles, which connect the dorsal brain region with the dorsal root of the ventral brain region (prn). Caused by a hypothesized evolutionary transition of the dorsal lip and associated buccal tentacles from the mouth region towards dorsal in Terebellidae (as summarized in Zhadan and Tzetlin 2002), these prostomial nerves are comparable with those in the ampharetid H. invalida. Nevertheless, they are much shorter. The same loop-like nerves innervate the ampharetid buccal tentacles as well as the “dorsal ridge” in Terebellidae (according to Orrhage, 2001) and potentially even the “cephalic veil” in Pectinariidae.
Due to the comparable position of the brain in Terebelliformia (Zhadan and Tzetlin 2002) and similar innervation patterns – including prostomial loop-like nerves connecting the dorsal to the ventral brain region – our data generally confirm previous assumptions (Orrhage 2001; Zhadan Anna E. 2002; Orrhage and Müller 2005). Accordingly, the “cephalic veil” of Pectinariidae, the “dorsal ridge” of Terebellidae and the hood-like “tentacular membrane” of Ampharetidae should be treated as being homologous structures.
In contrast, our conclusions concerning the potential prostomial or peristomial origin of these structures differ from earlier hypotheses. Fauchald and Rouse (1997) suggested the pectinariid cephalic veil as a fusion of pro- and peristomium. Furthermore, developmental studies support the tentacles of Terebellidae as being part of the prostomium (Bhaud and Gremare 1988), whereas others suggested a peristomial origin of the latter (Zhadan and Tzetlin 2002). Only for Ampharetidae developmental and morphological studies both suggest a peristomial origin of the buccal tentacles (Cazaux 1982; Zhadan and Tzetlin 2002).
The combination of potentially homologous innervation patterns described for Pectinariidae (Orrhage 2001) and our observation concerning terebellid and ampharetid taxa, homologous structures such as the “cephalic veil”, the “dorsal ridge” and the “tentacular membrane” should be defined as prostomial structures.
The earlier erroneous interpretation of terebellid tentacles as being of prostomial origin might have been caused by the developmental transition of character complexes involved into the cephalisation processes. This cephalisation is obvious for the ontogenetic transition of branchiae and their inclusion in the formation of the anterior end in Ampharetidae, Alvinellidae, Melinnidae and Terebellidae (Stiller et al. 2020), but is not responsible for the localization of terebelliform tentacles. The latter are characterized by a steady lateral addition of tentacles during ontogenesis. Both processes are independent and will be discussed in more detail below. Nonetheless, additional detailed morphological as well as comprehensive developmental analyses are necessary for a better understanding of the role of cephalisation and multiplication processes in the formation of morphological features in Annelida.
Branchiae – anterior transition during cephalisation
The ampharetid Hypania invalida bears four pairs of digitate branchiae grouped dorsally on the head, while the two terebellid species Lanice conchilega and Terebella lapidaria show three pairs of dichotomous branchiae serially arranged along the trunk (Capa and Hutchings 2006; Stiller et al. 2020). The putative basally-branching Pectinariidae also exhibit four pairs of branchiae along the segments II-V like shown for many other Terebelliformia (Nogueira et al. 2010, 2013). For all species investigated herein, the branchiae were easily identifiable in histological sections by the appearance of an afferent and an efferent blood vessel surrounded by the coelomic cavity.
In Terebelliformia, an ancestral number of four branchiae is assumed, while several reductions, transitions and even multiplication processes took place (Stiller et al. 2020). Due to developmental studies (e.g., Cazaux 1982) the branchiae in larval Ampharetidae occur from segment II-VI and shift towards anterior during ontogenesis. Therefore, branchia in adults are located at segment II and III (Stiller et al. 2020). All branchial appendages in Terebelliformia are innervated by a more prominent anterior and a slender posterior, segmentally arranged lateral neurite bundle originating from the anterior end of the vnc. They proceed in dorsal direction along the trunk musculature and terminate in a huge neuron, situated at the base of each branchia. Notably, this observed pattern is comparable to the neuronal innervation pattern known from parapodial appendages in the errant annelids Neanthes arenaceodentata (Moore, 1903) and Platynereis dumerilii ( Audouin & Milne Edwards, 1833) (see Winchell et al. 2010; Starunov et al. 2017). Therefore, an evolutionary scenario including a parapodia-linked origin of the branchial structures and the later involvement in cephalization events has to be assumed. Notably, cephalisation seems to be a widespread evolutionary phenomenon in annelids, and is described for several errant as well as sedentary taxa (Åkesson 1962; Faroni-Perez et al. 2016). Such an ontogenetic transition of larval trunk-associated appendages towards anterior in adult specimens seems to represent an important mechanism. Cephalisation seem to represent one major evolutionary process responsible for the diversity of the sensorial apparatus and even physiological adaptations of the anterior end in Terebelliformia and Annelida in general. Unfortunately, we were not able to compare our observation with the neuronal innervation of the branchiae in adult Terebellidae. The branchiae in the investigated aulophora larvae of L. conchilega were still not fully developed and can therefore not be used for detailed interpretations.
The anterior-most appendages – a multiplications of palps?
In the ampharetid H. invalida each tentacle is innervated by four neurite bundles, whereas in total eight prominent neurite bundles proceed along each tentacle in late larvae of L. conchilega. However, in both species the tentacles are similarly innervated by neurite bundles, which originate from the ventral and dorsal root of the circumoesophageal connectives. In H. invalida, tentacular nerves originate from the ventral root of the more delicate ventral brain region, which is connected to the circumoesophageal connectives (cc) by the ventral brain region commissure (cvbr). Additionally, they always originate from the dorsal root of the ventral brain region, which is connected to the dorsal, prominent region of the brain by the dorsal brain region commissure (cdbr). In L. conchilega, nerves of the buccal tentacles originate from neurons arranged arc-like on the ventral region of the brain. They are innervated by neurite bundles of both roots of the cc. These neurite bundles split up directly before entering the more prominent dorsal brain region. Such an arc-like set of innervating neurons was never described for terebellids so far.
According to various authors, annelid feeding palps are defined as being innervated by nerves originating from the ventral and dorsal main neurite bundles (roots) of the cc (Fauchald and Rouse 1997; Orrhage 2001; Rouse and Pleijel 2001; Orrhage and Müller 2005; Purschke et al. 2014). As described for the terebellid Pista cristata (Müller, 1776) and the ampharetid Amphicteis cf. gunneri (M. Sars, 1835) (Orrhage 2001) a quite similar innervation pattern of the buccal tentacles can be assumed. Furthermore, they seem closely associated with the nerves of the alimentary canal (Orrhage 2001; Orrhage and Müller 2005). By comparing earlier investigations (see Orrhage and Müller 2005) and our results, the branching pattern of all involved neurite bundles is similar. It is shown that stomatogastric nerves only originate from the cc or/and its ventral root and thus belong to the ventral brain region. A pattern we can observe in H. invalida and L. conchilega as well. Accordingly, the ventral strand (root) of the cc and the innervation of the entire stomatogastric system are closely associated. A close neuroanatomical connection of the “common tract”/ ventral region of the brain and the stomatogastric neurite bundles seems to be a widespread phenomenon. An alimentary origin of the buccal tentacles is therefore unreasonable based on the presented data.
Furthermore, no obvious developmental connection of the pharynx and the buccal tentacles can be observed for terebelliform taxa so far (Eckelbarger 1974; Cazaux 1982; Bhaud and Gremare 1988).
According to the current knowledge (Orrhage 2001) and our presented data, a similar neuroanatomy can be assumed for Pectinariidae as being the sister taxon of all other Terebelliformia (Stiller et al. 2020). Many authors also support the homology of the buccal tentacles in all Terebelliformia investigated so far (Holthe 1986; Orrhage 2001; Zhadan and Tzetlin 2002). In conclusion, our data strongly promote a homologization of the buccal tentacles in Terebelliformia with the feeding palps in the remaining Annelida.
Accordingly, in Oweniidae (exemplarily shown for Owenia fusiformis (Delle Chiaje, 1844) and Myriowenia sp. (Hartmann, 1960)), two main neurite bundles innervate the palps or the tentacular crown. These prominent neurite bundles originate dorso-laterally and medio-dorsally from the brain (Beckers et al. 2019b). Together with data from Magelonidae - which show a much more structured brain with a clear interpretation of the palp neurite bundles coming from the ventral and the dorsal region of the brain - a homologization of the feeding palps in Paleoannelida (Oweniidae and Magelonidae) and Terebelliformia is plausible (Beckers et al. 2019a) . In Sedentaria, comparable palps and palp nerves fulfilling these neuroanatomical criteria can also be found in Orbiniidae (Wilkens and Purschke 2009), Siboglinidae (Worsaae et al. 2016; Rimskaya-Korsakova et al. 2018), Cirratuliformia (Orrhage and Müller 2005) and Spionidae/Sabellidae (Orrhage 1980; Orrhage and Müller 2005; Purschke 2016). A detailed comparison of the neuroanatomy of all mentioned taxa highly supports our hypothesis concerning the terebelliformian anterior-most head appendages as sharing the same evolutionary origin like the feeding palps of other taxa. Comparable investigations for errant annelids are still pending.
During the ontogenetic formation of the anterior end, a clear differentiation has to be made in respect of “cephalisation” and “multiplication” – not only in Terebelliformia.
In contrast to adult Terebellidae, the aulophora larvae of L. conchilega possesses only a few tentacles, which are arranged antero-dorsally on the head. During ontogenesis, multiplication – an increase in the number of similar structures – in this case of the tentacles, leads to a lateral increase of tentacles until adulthood. Inside the tentacular bud, loop-like nerves originate from the ventral brain region and differentiate from lateral in median direction. This multiplication process is also known from other sedentary polychaetes, such as Sabellariidae and Sabellidae (Wilson 1936; Faroni-Perez et al. 2016) and results in an anterior concentration of numerous identical structures, like e.g. feeding palps, which are therefore involved into the formation of the anterior end. In contrast, the anterior clustering of other structures – such as branchiae – is shown to be the result of ontogenetic transition of parapodia-associated structures during cephalisation (see above). Although the result of both processes – multiplication and cephalisation – highly contributes to the formation of the anterior end in annelids (or at least the realisation of the sensorial pecularities of the head), they have to be considered as independent processes that should be interpreted separately.