Numerous vertebrate and ecdysozoan animal models are used in developmental biology. By contrast, only a few species have been studied in lophotrochozoan groups. Cephalopods are now emerging as novel lophotrochozoan model systems; among them, Sepia officinalis is particularly suitable because of its availability and regarding its commercial value. Cephalopods have been essentially studied for the complexity and originality of their nervous system, as they show a central nervous system presenting both structural and functional convergences with vertebrates (Packard,1972; Hochner et al.,2006; Williamson and Chrachri,2007). However, other particularities are as remarkable, especially regarding the muscular system. For example, the cephalopods cardiac muscular system involves three hearts (at the base of each of the two gills, a branchial heart pumps blood from the body up to the gills where it is oxygenated; the axial systemic heart pumps oxygenated blood from both gills to the rest of the body). However, despite of this particularity, physiological convergences of the cephalopods' closed circulatory system with the vertebrate one have been underlined years ago (reviewed in Schipp,1987; Wells and Smith,1987), justifying the interest of this model in muscle structure and function studies.
Different histological types of muscles are described in molluscs (Millman,1967). Added to smooth and cross-striated muscles known in vertebrates, a third class, obliquely striated muscles, is noticeable. In these muscle cells, the structures anchoring actin filaments are not linked in continuous Z-lines. Moreover, the contractile apparatus is organized in structures reminding of cross-striated muscle sarcomeres, but these “sarcomeres” are not laterally aligned; instead, the Z-line analogues are aligned obliquely to the longitudinal axis of the cell (Gonzalez-Santander and Socastro Garcia-Blanco,1972). Obliquely striated muscles are present in a few metazoan groups such as nematodes (Reger,1964; Rosenbluth,1965), annelids (Kawaguti and Ikemoto,1957a), and molluscs (cephalopods: Ballowitz,1892; bivalve: Marceau,1904). Although they have often been misinterpreted as smooth muscle (Kawaguti and Ikemoto,1957b; Lowy and Millman,1962), almost all muscles in cephalopods are obliquely striated. Cross-striated muscles have also been described in tentacles (Kier,1985,1991; Kier and Thompson,2003; Kier and Schachat,2008); the presence of this type of muscle in the systemic heart is discussed (Schipp and Schäfer,1969; Kling and Schipp,1987; Budelmann,1997).
Few data are available concerning the genetic control of muscle development in cephalopods. In vertebrates, skeletal cross-striated muscle development depends of MRFs (myogenesis regulating factors), a highly conserved family of four transcription factors (review in Buckingham et al.,2003; Tapscott,2005): MyoD, myf-5, myogenin and MRF4 (myf-6). Using vertebrate anti-MRF antibodies, Grimaldi et al. (2004a,b) have shown myf5- and MyoD-like positive staining in muscle cells of arms and tentacles in Sepia officinalis embryos from stage 26 (according to Lemaire,1970) to hatching.
The development of cardiac muscle, the other type of striated muscle in vertebrates, does not depend on MRF, but rather from numerous different transcription factors (review in Bruneau,2002; Brand,2003; Buckingham et al.,2005; Olson,2006). Among them, NK4 is particularly interesting. This gene belongs to the NK gene cluster, which is probably the most ancient homeobox gene cluster, dating at least from the base of Metazoa (Larroux et al.,2007), and NK4 gene homologues have been identified in numerous species (Harvey,1996; Elliott et al.,2006). Moreover, the main function of these genes seems to be highly conserved within Metazoa. The vertebrate NK4 homologue, Nkx2-5 (Lints et al.,1993), is implicated in heart development (review in Akazawa and Komuro,2005; Olson,2006). Tinman/NK4, the Drosophila founder member of this family (Kim and Nirenberg,1989), is involved in mesoderm regionalization and is required for cell fate specification of the dorsal vessel, the equivalent of the vertebrate heart (Bodmer et al.,1990; Bodmer,1993; review in Cripps and Olson,2002). In Caenorhabditis elegans, the tinman homologue ceh-22 is involved in pharyngeal muscle differentiation (Okkema and Fire,1994; Okkema et al.,1997); these muscles are responsible for intrinsic rhythmic contractions of the pharynx, and thus considered as analogs of vertebrate cardiac muscle (Pilon and Morck,2005). In cephalopods, NK4 homologue expression in the systemic heart and the adjacent muscular ink sac of Loligo paelii has been pointed out in one late embryonic stage (Elliott et al.,2006). No data are available regarding the expression in earlier stages when muscle determination and differentiation occur.
We determine in this study the expression pattern of NK4 during Sepia officinalis development. We show by whole-mount in situ hybridization that NK4 expression begins before morphogenesis, and is not restricted to the prospective cardiac muscles but concerns above all prospective locomotory muscles areas.
Sepia officinalis develops directly without metamorphosis. Organogenesis proceeds during 2 to 3 weeks (at 25°C), from stage 15 to hatching resulting in adult anatomy (Boletzky,2006). Zygote cleavage gives rise to a disk-shaped embryo at the animal pole of the egg, whereas the vegetal pole is made of thin layer of “extra-embryonic” ectoderm cells that covers the yolk. After a disk-shaped phase where prospective organs start delineating, the embryo expands as all organs gain volume. Describing these steps of development is often confusing due to an unusual orientation of the embryo when compared with other well-known models (Fig. 1): the stomodeum is dorsal, the oral (or anterior) pole of the future adult lies at the periphery of the embryo (arm crown and stomodeum), whereas the future aboral (or posterior end of the adult) pole is central (mantle, gills, funnel). The final adult arrangement is reached at stage 21, where the whole embryo straightens: eyes, mouth, and the arm crown are then located at the yolk side (cephalopodium) and the visceral mass, mantle cavity, and surrounding mantle (visceropallium) at the opposite side. The arm crown finally comprises eight arms and two tentacles (which are numbered from the dorsal side); they are morphologically identical until stage 22/23, after which tentacles (= “arms” IV) begin to show a recognizable tentacular club.
SoNK4 expression has been detected at the earliest studied stage (early stage 15), before any visible structure emerged (Fig. 2-15b). The mantle area was stained, except in its central part that corresponds to the shell sac depression primordia. The prospective arm areas were also stained, except for arm III that is known to appear later than the others. The intensity of expression was different between arms, being particularly weaker in arm I. A weak but noticeable expression was observed as an anterior extension of arm II (early stage 15, Fig. 2-15b) and I (late stage 15, Fig. 2-15c), which could correspond to migrating cells, as it was not observed on these arms in later stages. As soon as the gill primordia were discernable, they were NK4-positive.
At stage 16, arm morphogenesis began as an arm crown, over which arm buds grew. Arm I and III were not morphologically perceptible. In this continuous structure, SoNK4 expression was still restricted to the future arm buds, and as in stage 15, prospective arm III was devoid of staining. As arm buds grew, SoNK4 expression did not concern all the bud structure, the ventral part of the bud being devoid of staining (Fig. 2-16b). The mantle was also strongly underlined except a depression in the ventral median line; the interpretation of this structure (already signaled in morphological studies by Naef,1928) is not clear. At this stage, expression was clearly noticeable in the gill primordia. An expression was also detectable at the base of the gills, which corresponds to the prospective branchial heart area (Fig. 2-16c).
At stage 17, all arm buds were clearly SoNK4 positive, including arm III (Fig. 2-17b). Of interest, the staining of this arm showed an anterior extension similar to that observed for other arms at early stages. The already stained structures were all still underlined.
Sections at these stages (16/17) showed in all cases a SoNK4 expression below the epithelium, corresponding to the mesodermal layer (Fig. 2-16c, 2-17c), and confirmed a restricted expression in the dorsal part of arm buds (Fig. 2-17d).
From stage 18 to 20, expression was still strong in the mantle and in the arms (Fig. 3). Arm buds were clearly separated into two adjacent parts, with the dorsal-most part stained. This arm bud separation is underlined by the venous system (Naef,1928), and was not SoNK4-positive, suggesting that SoNK4 was not required for the early development of this peripheral vascular system. By contrast, the branchial hearts as the gills were still stained. A sparse but clear staining was also noticed in the funnel tube primordia that begin to take place from stage 16/17, but the funnel pouch primordia remained negative: this difference is in agreement with the facts (i) that these two structures are known to differentiate independently (Naef,1928) and (ii) that their embryological origins are really different, the funnel tube rudiments belonging to the cephalopodium, whereas the funnel pouch has a pallio-visceral origin (Boletzky,1988b).
At stage 18 and 19, the anus area located between the gills was very transiently stained as the staining disappeared after stage 19. A scattered expression was also noticed at stage 18 as a line in the middle of the forming eyes, which could correspond to the closing of the primary optic vesicles. The contrast observed at later stage in the eye was due to the pigmentation of the retina, as the controls show similar staining (compare Fig. 4-24 and Fig. 4-24T).
At stage 21 to 23, as the two funnel tube primordia went in contact, gills and funnel tube were no longer stained (Fig. 3 insert). By contrast, expression in the mantle and in the arms persisted. It was progressively less intense in the arms. This dynamics of SoNK4 expression in arms was underlined by the fact that the latest appeared arms (III) presented at this stage a higher and more regionalized signal than the others, in agreement with its delayed morphogenesis (Fig. 3-22b). As the arm bud extended, SoNK4 expression was restricted to their distal end (Fig. 3-22+). No obvious difference was noted between arms and tentacles (IV). The decrease of the signal was also evident in mantle. Sections showed that SoNK4 expression was restricted to the future muscle layer (Fig. 3-23).
From stage 24 to 27, only the mantle was stained; SoNK4 was not expressed in the arms anymore (Fig. 4). Finally, in the later tested stages (28 to hatching), all specific staining disappeared (not shown).
We show in this study that SoNK4 is expressed during early development in Sepia officinalis not only in cardiac tissue but also in muscle containing structures such as the mantle, the funnel, and the arms. To our knowledge, this is the first description of NK4 expression during the development of a Mollusca.
In vertebrates, the NK4 homologue Nkx2-5 is initially expressed in mesoderm and anterior endoderm (Lints et al.,1993), and the endodermic expression has been linked to a role in the pharyngeal region patterning. By contrast, the expression of NK4/Tinman in Drosophila is strictly mesodermic (Bodmer et al.,1990; Bodmer,1993). In Sepia officinalis, we did not detect any indication of SoNK4 expression in putative endodermic territories (such as the so-called mesentodermic areas; Naef,1928), except for a very transient expression in the anal prospective area at stages 18–19. The setting up of the digestive tract takes place relatively late in the development as the animal gains volume (Boletzky,1978); it becomes functional after hatching. No expression has been detected in these tissues even in the latest stages (stages 25 to 30) where the tract is fully developed. Thus, it appears that SoNK4 is not implicated in the development of anterior endodermic tissues in Sepia officinalis. As in Drosophila, its role seems to be restricted to mesodermic development in cephalopods.
The main accepted role of NK4 and its homologues is their implication in cardiac/rhythmically contractile muscle development (Cripps and Olson,2002; Zaffran and Frasch,2002; Akazawa and Komuro,2005; Olson,2006). According to Okkema and coworkers (Okkema and Fire,1994; Okkema et al.,1997), the expression of ceh-22, the NK4 homologue in Caenorhabditis elegans, is restricted to the rhythmically contractile pharyngeal muscle, whereas bodywall (somatic) muscles development require other myogenic factors such as MRF-related, MADS-related, or Hand-related factors (Baugh and Hunter,2006; Fukushige et al.,2006). In vertebrates, expression of the NK4 homologue Nkx2-5 is rapidly restricted to cardiac muscle (Olson,2006); Nkx2.5 is not directly involved in cardiac cells determination but participates to the correct morphogenesis of the heart (Lints et al.,1993; Lyons et al.,1995). In Drosophila melanogaster, NK4/Tinman is involved in the determination of cardiac cells (Bodmer,1993) and it has been shown recently that NK4 could successively intervene in cardiac cells determination and differentiation (Zaffran et al.,2006). Thus, NK4 may be responsible for cardiac cells determination and/or differentiation, depending on the species.
In Sepia officinalis, NK4 expression is observed at early stages (stages 16 to 18) in the gills and in the branchial hearts prospective areas. Branchial hearts differentiate earlier than the arterial (systemic) heart. In a relatively close species (Octopus), they begin to pulse at stage 22–23 (before the gills function) irregularly then regularly. At the end of this period, the systemic heart begins to contract and adopts a synchronous contraction rhythm with branchial hearts at stages 24–25, as arterial and venous circuits are established (Boletzky,1987). At early stages (16–18), the gill rudiments are not yet vascularized in Sepia officinalis (Schipp,1987) and no vascular/heart muscle structure are present. After stage 20, SoNK4 expression has been detected neither in branchial or systemic hearts nor gills. In a late stage of development (probably stage 26, personal communication), Elliott et al. (2006) reported an expression of SoNK4 in the systemic heart but not in branchial hearts of Loligo paelii. These results confirm that expression of SoNK4 in gills and branchial hearts prospective areas is transient and restricted to the earlier stages of development in cephalopods. Therefore, SoNK4 could play a role in branchial heart determination, but does not seem to take part in branchial heart morphogenesis. By contrast, the expression observed in systemic heart of Loligo, late in development, suggests that NK4 could intervene in the final morphogenesis as the circulatory system takes place only at the end of the development. Even if we cannot exclude that branchial hearts and the systemic heart do not have identical muscle cell structure (Schipp and Schäfer,1969; Kling and Schipp,1987; Budelmann,1997), it appears clearly that the pulsatory system development takes place by the way of different genetic mechanisms, suggesting a NK4 role at two different developmental stages, one for the branchial hearts (the venous system) and one for the systemic heart (the arterial system). Thus, NK4 expression in cephalopods does not correlate with its proposed general role in pulsatory cell determination (Olson,2006), and NK4 should be interpreted as part of a pulsatory gene network rather than as a pulsatory master gene. This is also the case in other species. In C. elegans embryo Ceh22 (NK4 homologue) is not expressed in all the putative contractive pharyngeal muscular cells and requires later other genes to specify the contractile characteristic of these cells (Okkema and Fire,1994). The expression domain of tinman in Drosophila is much broader than the region of heart precursor formation, but its ectopic expression does not produce ectopic cardiac tissue (Yin and Frasch,1998); other genes like the GATA factors family are required to promote the cardiac cell lineage (Gajewski et al.,1999).
The main expression of SoNK4 in Sepia officinalis is essentially located in structures known to be particularly rich in muscles, namely arms and mantle. The arms' buds are positive from stage 16 to 19 and the mantle from stage 15 to 27. Taken together, these results suggest that a transient expression of SoNK4 could be implicated in early myogenesis of noncardiac muscle in Sepia officinalis. This early expression of SoNK4, even before any detectable morphogenesis, suggests a role in muscle cell determination. As no expression is detected at later stages (stages 27 to 30), a putative role in muscle differentiation (which occurs after stage 26 at least in tentacles, Grimaldi et al.,2004b) or function can be discarded.
Cardiac, arm, and mantle musculature of Sepia officinalis are composed of obliquely striated muscles, suggesting that SoNK4 expression in cephalopods somatic muscle could be linked to this particular type of muscle cell. We also found that tentacles (arms IV), like the arms (arms I, II, III, and V), are positive for SoNK4. It has been shown by Kier and coworkers (Kier,1985; Kier and Schachat,2008) that adult cephalopods tentacles are largely composed of transversal cross-striated muscles, but striated muscles of the tentacles are known to appear after embryonic development, around hatching time in Sepia officinalis (Grimaldi et al.,2004b), and 4 to 5 weeks after hatching in Loligo paelii (Kier,1996). Thus, other gene(s) are likely to be implicated in the differentiation of cross-striated muscle in this species, and MRF are obviously good candidates: we are currently exploring this possibility. However, it should be emphasized that the muscular wall of the buccal mass and the oesophagus, as the plexiform layer of the stomach of Sepia officinalis have also been described as obliquely striated (Amsellem and Nicaise,1980). We were unable to detect any SoNK4 expression in these structures during Sepia officinalis development. Therefore, it is unlikely that SoNK4 expression is strictly related to the structural type of muscles in Sepia officinalis.
NK4 gene homologues expression has been strongly linked to rhythmically contractile muscle development, but it is clear from our results that SoNK4 is not strictly linked to this specific muscular function, namely intrinsic rhythmic contractility. By contrast, it is noteworthy that structures strongly expressing SoNK4 during Sepia officinalis development are involved in the locomotory function, including part of the muscular hydrostat (Kier,1989; Kier and Thompson,2003). Locomotory muscles in cephalopods (such as mantle, funnel, or arm muscles) can be functionally assimilated to skeletal muscles of vertebrates or body wall muscle of Drosophila, which do not express NK4. A role in specific parts of somatic muscle development has also been proposed for Tinman in Drosophila (Bodmer et al.,1990; Azpiazu and Frasch,1993; Bodmer,1993). Although SoNK4 expression is not strictly restricted to muscular structures in Sepia officinalis, suggesting an implication of SoNK4 in different mesodermal areas, it is clearly preferentially linked to the locomotory system, thus underlying this model as a particularly original one.
Last but not least, our results about SoNK4 expression in a cephalopod allow evolutionary inference. Indeed, arms and funnel are synapomorphies of cephalopods among mollusks, as derived from the foot of a mollusc putative ancestor (Naef,1923; Boletzky,1988a). In association with the funnel, the mantle is also modified for a locomotory function (the jet propulsion) in cephalopods. NK4 could have been recruited in this clade for the apparition of morphological novelties implicated in the locomotory function. Alternatively, NK4 recruitment could be interpreted as a molluscan synapomorphy implicated in the foot morphogenesis. To our knowledge, no data are currently available regarding the expression of NK4 homologues in other molluscan species. This could be an interesting perspective to explore.
Collection of Sepia officinalis Embryos
During spring and summer (April to September), fertilized eggs were laid by captive Sepia officinalis females maintained in the biological stations of Luc-sur-mer (France) and Banyuls-sur-mer (France). Eggs were kept in artificial see water (Red Sea) with aeration at room temperature (RT). Development occurred into the chorion, a tough secreted membrane that surrounds the egg. In these conditions, the development was normal, as assessed by its timing and the morphological aspect of the embryos. From eggs batches, individual eggs were detached and embryos were taken out by removing some of the numerous surrounding envelopes using forceps in sea water. Then, embryos were visually staged using Lemaire's (1970) system for Sepia officinalis. As we focused on organogenesis, embryos at stages 15 to 30 were selected for in situ hybridization (ISH). Embryos were fixed one night within the chorion in 3.7% paraformaldehyde (PFA) in phosphate buffered saline (PBS), then dechorionated. A second step of fixation (3.7% PFA, 24 hr) followed. After three rinses in PBS, embryos were dehydrated in increasing concentrations of methanol (25% to 100%). They were conserved at −20°C until use.
In Situ Hybridization
The full-length SoNK4 (Sepia officinalis NK4) cDNA (EMBL accession number AY298768) was kindly provided by Dr Richard P. Harvey from the Developmental Biology unit of Victor Chang Cardiac Research Institute, Sydney Australia.
RNA probes (1,150 bp) were obtained with the Sp6-T7 kit from Roche as recommended by the manufacturer. Antisense probes were obtained with Sp6 polymerase. Sense probes obtained with the T7 polymerase were used as a control.
ISH was done in hemolyze tubes, each containing one embryo in large volumes of solution (approximately 2 ml), and under agitation. At least 3 embryos from each studied stage have been treated, and controls were done for each stage. Progressive rehydration of the embryos was done by successive immersions (5 min, RT) in PTW/methanol solution (PBS, Tween20 1%, methanol from 60% to 0%). Embryos were then treated by proteinase K in PTW (0.05%, 20 to 50 min, depending on the stage of the embryos). They were then post-fixed 1 hr in PFA.
A prehybridization step was done in SH solution (formamide 50%, standard saline citrate 20×, Tween20 1%, sodium dodecyl sulfate 2.5%) allowing progressive impregnation of the tissues. Embryos were then incubated 1 night at 65°C in this solution to which the probe had been added. Control embryos were incubated with a sense probe. Excess of probe was eliminated by five rinses (30 min, 65°C) in SH solution.
Embryos were impregnated with the buffer of blocking solution (BS: Maleic acid 100 mM, NaCl 150 mM, pH 7.5, Tween20 1%). Saturation was then done in blocking solution (BS, BB 4% Roche, fetal bovine serum [FBS] 15%, 1 hr, RT), followed by incubation 1 night at 4°C with anti-digoxigenin antibodies (Roche) coupled to alkaline phosphatase (AP) in blocking solution (BS, BB 2.4%, FBS 20%).
Revelation of AP activity was done using NBT-BCIP (Roche) as a substrate. After complete development, the reaction was stopped by several washing in PTW solution. A final fixation in PFA (2 hr, RT) was then realized. To obtain more precise localizations, some of these embryos were included in paraffin using standard protocols, and cut in 7-μm sections.
Observations were done with a Leica M16 2F binocular stereomicroscope and Leica microscope.
We thank the biological stations of Luc sur mer, Banyuls sur mer, Joel Henry (Université de Caen), Ludovic Dickel (Université de Caen) for providing Sepia brood, Aude Andouche for help in Sepia officinalis embryos dissection, and referees for helpful commentaries.