As summarized above, functional similarities support the relatedness of the Platynereis photosensory cells of the hiomt region and the cells of the pineal complex and hypothalamus of vertebrates. The molecular and developmental data present a more complex picture, which overall doesn’t indicate the presence of 1:1 homolo-gies between annelids and vertebrates. Moving away from the vertebrate dogma
“melatonin equals pineal” helps to decipher the evolutionary relationships of the annelid and vertebrate hiomt expressing cells. Looking at the function of selected transcription factors during development can help to understand how the changes in regulatory interactions between transcription factors might drive changes of cell
types during evolution.
The transcription factor rx (rax in mammals) is one of the very few genes expressed in retina, pineal and deep brain photoreceptors (Casarosa et al., 1997; Deschet et al., 1999), although species-specific differences exist (like the lack of rx expression in the fish pineal).
A question that has not been addressed yet in any vertebrate species is how the early pineal anlage, at the border of the neural plate, is established. At very early stages, this anlage expresses several genes that label also the eye field, like rx, otx genes and tbx2/3. Indeed, fate map studies in Xenopus at the neurula stage showed that the pineal precursors lie in continuity with the more medial eye field (Eagleson and Harris, 1990; Rubenstein et al., 1998). So it is possible that the same patterning mechanisms establish very early the diencephalic domain that gives rise to photore-ceptive structures.
A model for the establishment of the eye field has been proposed for Xenopus (Zuber et al., 2003), and it is likely to be valid in general for vertebrates. The earliest events involve the induction of otx2 expression in the anterior neural plate by neural induction signals, like noggin. Genes of the tbx2/3 family are subsequently activated;
these genes contribute to induction of rx and pax6 expression.
Since the transcription factors otx, rx and tbx2/3 are also expressed in the early anlage of the pineal complex, it is likely that the segregation of the pineal and reti-nal fields takes place after this point. The pineal anlage is distinguished from the retina by the expression of some transcription factors. One of them, not (or flh), is never present in the eye, and plays an important role in the maintenance and pro-liferation of the pineal (but not parapineal) progenitors (Masai et al., 1997; Snelson et al., 2008a). Later, other pineal-specific genes, like bsx and foxD3, are found in the differentiating pineal cells (D’Autilia et al., 2010). Additionally, sens is expressed specifically in the zebrafish parapineal neurons, which loose not expression after migrating away from the midline (Dufourcq et al., 2004; Snelson et al., 2008a).
In Platynereis, the same genes are expressed dynamically in the dorsal brain dur-ing development, and in different combinations in the differentiated hiomt cells.
However, none of the Platynereis cell types expresses the combination of all these transcription factors at any point during development.
From the analysis of the early development with time lapse movies and in situ hy-bridization, three interesting photoreceptor lineages, with a medio-lateral arrange-ment, can be identified.
eage” in fig. 3.6) is tracked from two otx+ NSCs (one on each side of the dorsal episphere) at 12hpf. At 15hpf, the same NSCs turn off otx expression and start to express tbx2/3, and later rx. This is reminiscent of the vertebrate eye field situa-tion, where at onset of rx expression otx2 is downregulated, both in Xenopus and zebrafish (Andreazzoli et al., 1999; Chuang and Raymond, 2001).
This lineage has very dynamic division patterns, which are paralleled by dynamic changes of gene expression in neural progenitors. The initial otx+ NSC starts with some asymmetric cell divisions, producing differentiating neurons, but then divides symmetrically to produce new NSCs. From the comparison of WMISH and time lapse movies it seems likely that the onset of rx expression correlates with this sym-metric cell division. Mapping the expression of rx and otx in this lineage with an higher resolution will be especially insightful and clarify how the changes of the NSCs behaviour correlate with gene expression.
The functional studies indicate that in this lineage Rx plays some role in the main-tenance of neural progenitors. Knock down of Rx produces loss of cPRCs and tll expression, despite rx is neither expressed in immediate progenitors of the cPRCs, nor later during their differentiation. This means that Rx cannot have a role in the terminal differentiation and maintenance of the cPRCs phenotype, but it must main-tain the proliferative capacity of the NSCs where they come from. A similar role for Rx has been proposed in vertebrates, during the early stages of eye development (Casarosa et al., 2003; Andreazzoli et al., 2003; Zhang et al., 2000; Loosli et al., 2001).
Consistently, in Platynereis a smaller number of cells, and a decrease of EdU incorpo-ration in the brain have been observed in rx morphants (not shown). Moreover, the expression of the proneural marker ngn is severely affected in rx morphants. Also the expression of tbx2/3 is strongly reduced (but it does not disappear completely) in morpholino-injected larvae. These results can be interpreted either as an indirect consequence of the reduced mitotic capacity of the rx+ NSCs, or as a requirement of a functional Rx to sustain tbx2/3 expression. In medaka, rx3 was shown to be up-stream of tbx2 in the retina, but not in the hypothalamus (Loosli et al., 2001). It will be definitely of great interest to study in more detail the regulatory relationships between rx and tbx2/3 in Platynereis.
Later in development, more cell types populate the “clock area”, so they are likely to be generated by the same lineage, although the time lapse movies do not last enough to cover later stages. These cells are all distinct, as shown by their molecular fingerprint analysis (described in Chapter 7). Among them, the “lateral not cells”
express an interesting combination of markers: not, foxD and otx, which are pineal markers, and rfx4/6, a gene that labels specifically the SCN within the nervous sys-tem. The function of these cells is currently unknown: they express perops and hiomt, indicating that they participate to the melatonin system.
The asymmetric serotoninergic cell comes from a different lineage, which is also expressing tbx2/3, but neither rx nor otx. At 15-16hpf, bsx is found in some cells of the left side of the tbx2/3+ region; only one of them turns on the expression of sens and differentiates into the asymmetric serotoninergic cell. Despite rx is not expressed in this lineage at 15hpf, the expression of c-ops1 in the asymmetric serotoninergic cell requires a functional Rx. Thus, this requirement might be non-cell autonomous and needs to be investigated further. Interestingly, the presence of serotonin, c-ops1, tbx2/3 and sens in this cell is highly reminiscent of the molecular fingerprint of the parapineal cells (it remains to be shown if bsx is expressed also in the parapineal, besides its pineal expression).
Finally, a third cell population might be extremely interesting for evolutionary comparisons with vertebrate cell types. This small cell population starts from a very lateral position in the episphere, and is demarcated by the coexpression of rx, otx and pax6; tbx2/3 is excluded from this region. Later in development, these cells acquire a more medial position; two of them become the two lateral serotoninergic cells, which retain rx expression after differentiation. Overexpression of Rx induces an increase of the number of lateral serotoninergic cells, together with the expansion of the lateral otx expression domain. This is consistent with a role of Rx, upstream Otx, in the development of cells of the lateral episphere.
Despite their expression profile is so similar to pineal and retinal progenitors, these rx+ otx+ pax6+ cells were not found to express ops1. However, another c-opsin gene (c-ops2) exists in Platynereis and is transcribed at larval stages (RT-PCR), but unfortunately it was not possible yet to localize its expression. Thus the expres-sion pattern of c-ops2 might reveal new cell types, more interesting for evolutionary comparisons.
In Xenopus, a cocktail of transcription factors otx2, tbx3, pax6, six3, rx1, tll, optx2 -is completely sufficient to induce ectopic eyes outside the nervous system (Zuber et al., 2003). All the orthologs of these genes in Platynereis1, except pax6, are expressed at some point in the lineage producing the cPRCs and most of the perops+ cry1+ CNGa+
1Platynereis six3 is the ortholog of the vertebrate genes six3 and optx2/six6.
and tll, but not otx, in this lineage, consistent with the different onset of expression of these transcription factors during development. The hypothesis of homology of vertebrate retinal photoreceptors and annelid cPRCs proposed by Arendt et al.
(2004) was based on the expression of two differentiation markers, c-ops1 and bmal.
These new developmental data add a new level to the comparison, indicating that the same players (transcription factors) are involved in the development of ciliary photoreceptors within different anatomical contexts.
9.4. Differences in the organization and development of annelid and