A hedgehog homolog is involved in muscle formation and organization of Sepia officinalis (mollusca) mantle



Our study focuses on the possible involvement of the Hedgehog (Hh) pathway in the differentiation of striated muscle fibres in cuttlefish (Sepia officinalis) mantle. We show here that both an hh-homolog signalling molecule and its receptor Patched (Ptc) are expressed in a specific population of myoblasts which differentiates into the radial fast fibres. To evaluate the functional significance of hh expression in developing cuttlefish, we inhibited the Hedgehog signalling pathway by means of cyclopamine treatment in cuttlefish embryos. In treated embryos, the gross anatomy was considerably compromised, displaying an extremely reduced mantle with a high degree of morphological abnormalities. TUNEL and BrdU assays showed that the absence of an hh signalling induces apoptosis and reduces the proliferation rate of muscle precursors. We therefore hypothesize a possible involvement of Hh and its receptor Ptc in the formation of striated muscle fibres in cuttlefish. Developmental Dynamics 237:659–671, 2008. © 2008 Wiley-Liss, Inc.


Members of the hedgehog (hh) gene family have been found in vertebrates (practically in all chordates) (Shimeld,1999) and in several invertebrate taxa such as echinoderms (sea urchin) (Chang et al.,1994), insects (Drosophila melanogaster) (Ingham et al.,1991; Heemskerk and DiNardo,1994), annelids (leeches) (Kang et al.,2003), and molluscs (limpet) (Nederbragt et al.,2002).

In both vertebrates and invertebrates, members of the hedgehog (hh) gene family are expressed in both embryonic and adult tissues, where they function as key mediators of many fundamental processes such as development, growth, patterning, and morphogenesis. Although the final outcome might look quite different, several studies in invertebrates and vertebrates have unveiled a striking conservation in the deployment of members of the same signalling families to regulate development (Ingham and McMahon,2001).

Transduction of Hh signalling in the target cells occurs through two transmembrane proteins, Patched (Ptc) and Smoothened (Smo). According to several authors (Hooper and Scott,1989; Nakano et al.,1989), Ptc, the receptor of Hh, regulates Smo activity by influencing its interaction with small cellular molecules (Chen et al.,2002; Masdeu et al.,2006). Indeed, Smo is known to activate downstream Hh gene targets. However, in the absence of Hh, Ptc represses the signalling activity of Smo. By contrast, when Hh binds to Ptc, Smo inhibition is released and Hh signal is transduced within the target cell. ptc, found in Drosophila, in nematods (Zugasti et al.,2005) and in vertebrates as well (Goodrich et al.,1996; Hahn et al.,1996; Marigo et al.,1996), is itself a target gene of hh signalling.

Within the Hh gene family, Sonic hedgehog (Shh) has been defined as a key regulator of myogenesis in vertebrates. Indeed, Shh affects the decision between fast and slow muscle formation (Blagden et al.,1997; Grimaldi et al.,2004b), the correct positioning of slow muscle fibers during chick muscle limb development (Duprez et al.,1999), and the survival and proliferation rates for myogenic precursor cells during mouse hypaxial muscle (Kruger et al.,2001) and chick limb muscle development (Duprez et al.,1998).

Even if substantial data show that Hh/Ptc signalling is involved in patterning vertebrate muscle differentiation, little information is available about the role played by this pathway in invertebrates. Only recently, it has been shown that an hh homolog is required for normal development of obliquely striated muscle fibres, a peculiar type of striated muscles, in leech proboscis (Kang et al.,2003).

In the present study, we chose the cuttlefish Sepia officinalis (Mollusca, Cephalopoda) as an invertebrate animal model to further investigate the extent of evolutionary conservation of the role of Hh signalling in myogenesis. To this end, the expression of Hh signalling proteins during cuttlefish muscle mantle development and differentiation was investigated, based on the recent finding of Hh expression in other molluscs such as Patella vulgata (Nederbragt et al.,2002). In addition, we have attempted to elucidate the role of Hh/Ptc signalling in the control of the precise positioning of obliquely striated muscle fibres during development.

The rationale for choosing the cuttlefish mantle for our studies is due to the fact that this structure is mainly composed of a large mass of obliquely striated muscles. These fibres form a preponderant circular layer that is subdivided in regular fields by thin radial fibres (cartoon in Fig. 1). The circularly disposed fibres, providing the power stroke contraction to expel water from the mantle cavity, are in turn divided into two types: oxidative mitochondria-rich fibres, localized in the outer zones of mantle, and mitochondria-poor fibres placed in the central region of mantle. Radial fibres, which are antagonist of the central circular fibres, are mitochondria-poor fibres as well (Bone et al.,1981; Budelmann et al.,1997). Even though the muscular mantle in adult cephalopods has been largely described, neither the increase in size of this muscular district nor the final spatial organization that it achieves during development have been comprehensively described.

Figure 1.

Color-coded cartoon showing the structure of an adult cuttlefish mantle in frontal view. The integument is in green, the connective tissue is in pink. The muscular layer is formed by the mass of circular fibres (squares in light blue) subdivided in regular fields by thin radial muscle fibres (in red).

Our expression data, combined with results from the experimental perturbation of the Hh signalling pathway by two independent approaches, suggest that hh signalling acts to provide both a muscle organizer factor for the mantle muscle layer and a survival and proliferation factor for myogenic precursors.


Morphological Analysis of Developing Mantle

Morphological analysis of cuttlefish mantle was carried out at different stages of development starting from stage 21, corresponding to an embryo showing the typical architecture of a cephalopod. At this stage, the embryonic body is made of a head, largely occupied by optic lobes with arms surrounding the mouth, and a small mantle sac (less than 1 mm in length) that envelopes the visceral mass (Lemaire,1970).

The mantle organization was quite simple and formed a cup-shaped structure (Fig. 2A) that was lined on both sides (external side facing the water environment and the innermost side proximal to the palleal cavity, i.e., proximal to the visceral mass) by a multilayered integument (Fig. 2B). This integument was formed by ciliated cells interspersed with rounded cells showing the typical organization of blast cells (large nucleus and scarce cytoplasm, Fig. 2C). Moreover, both outer and inner integument portions outlined a compact mass of blast-like cells (Fig. 2D).

Figure 2.

Frontally sectioned S. officinalis embryos stained with crystal violet and basic fuchsin. AD: Stage 21. The mantel sac (outlined in A) is lined by a multilayered integument (i, in B). C, D: Electromicrographs. Detail of ciliated integument cells (arrowhead in C), a centriole (arrowhead in inset), and blast-like cells (C, D arrows) forming the “myoblast layer” (square in B). EG: Stage 23. Dark elongated cells (F, G arrowheads) are distinguishable in the multilayered integument in close contact with the rounded cells (arrows). The myoblast layer (E, F, squares) is separated from the integument by a loose connective tissue (asterisk). HJ: Stage 25. I: The dark elongated cells (arrowheads) in the integument detach from adjacent rounded cells (arrows) and migrate into the connective tissue (inset Ia). J, Ja (electromicrograph): the elongated cells (arrowheads) form a very thin sheet close to the “myoblast layer” (square in H, J). KN: Stage 26. Elongated cells (arrowhead in K) migrate inside the “myoblast layer”. L: Electromicrograph. Myosin filaments (arrowheads) are visible in the elongated cells differentiating in radial muscle fibres. M,N: Detail of the rounded myocytes (arrows in M) and differentiated circular muscle fibres (encircled area in M, myosin filaments obliquely sectioned shown magnified in N). O: Stage 28. The mass of rounded myocytes is disposed in rows (brackets) separated by thin radial fibres (arrowheads). P: Stage 30. The mantle is chiefly composed of circular muscle fibres divided into regular fields (brackets) by thin radial fibres (arrowheads). The connective tissue (asterisk) thickness is reduced in comparison with the earlier stages. n, nucleus; i, integument, a, anterior region; p, posterior region. Scale bar = 600 μm in A, C, E, H; 50 μm in B, F,O,P; 4 μm in D; 1 μm in G; 25 μm in I, J, K, M; 3 μm in Ia, Ja; 200 nm in L, N.

Starting from stage 23 (Fig. 2E), two kind of cells were clearly distinguishable in the multilayered integument from a morphological point of view, since elongated, darker cells were now observed in contact with the rounded cells already described above (Fig. 2F,G). At this stage, a well-defined spatial organization of the mantle cells was achieved, with the compact layer of blast-like cells (around 50 μm in thickness) being separated from the outer integument by a central area of loose connective tissue (Fig. 2E,F). We will refer to the thick layer of blasts as “myoblast layer,” since the complex multilayered sheath of circular muscles will develop from this region. During further development, besides the extensive visceral mass growth, lengthening of the mantle in a postero-anterior direction becomes evident.

Starting from stage 25 (the mantle sac having reached 2 mm in length, Figs. 2H–J), the dark elongated cells of the integument became irregular in shape and lost contact with the neighbouring rounded cells (Fig. 2I). Moreover, both dark elongated and rounded cells were detected in the connective tissue (Fig. 2I, Ia). The dark and elongated cells reached and lined the inner “myoblast layer” (Fig. 2J, Ja), which at this stage reached around 70 μm in thickness. Two different cell types following a different fate during mantle muscle development were, therefore, unveiled and give rise to two distinct muscle fibres (as shown below). By stage 26 (Fig. 2K–N), the elongated cells, previously lining the myoblast layer, migrated inside it (Fig. 2K) giving rise to the radial muscle fibres. Their differentiation into spindle-shaped myocytes was confirmed by the presence of myosin filaments in their cytoplasm (Fig. 2L). At the same time, the rounded cells, also migrating from the integument, increased the thickness of the “myoblast layer,” differentiating into small circular muscle fibres (Fig. 2M,N).

Most cells forming the “myoblast layer” (approximately 130 μm in thickness) showed a more complex and precise organization at stage 28 (Fig. 2O). Indeed, groups of rounded myocytes, disposed in parallel rows, were now separated by the differentiating radial muscle fibres (Fig. 2O). At the time of hatching (stage 30), all the rounded myocytes were completely differentiated as circular fibres that were subdivided in regular fields by thin radial muscle fibres (Fig. 2P), similar to that previously described in adult animals by Bone et al. (1981) and as shown in the colour-coded cartoon (Fig. 1). The final structural organization of the mantle showed a largely reduced connective tissue under the integument, whereas the muscular layer reached approximately 180 μm in thickness (Fig. 2P).

Molecular Characterization of Muscle Precursors

In order to demonstrate that the cuttlefish mantle's muscle precursors derive from the external integument layer, we performed immunolocalization assays using an antibody raised against the muscular myogenic factor MyoD, which is specifically expressed in the muscular precursors in both vertebrates and invertebrates (Krause et al.,1992; Buckingham,1994; Chen et al.,1994; Krause et al.,1994; Ontell et al.,1995; Misquitta and Paterson,1999; Muller et al.,2003; Grimaldi et al.,2004c) together with an antibody raised against desmin, one of the earliest myogenic markers (Choi et al.,1990).

Starting from stage 21, a MyoD-like signal was mainly detected in cells populating the integument (Fig. 3A). By stage 25, the signal was now detected in those cells migrating into the connective tissue (Fig. 3B). Finally, at stage 28 the MyoD-like signal was detected in the myoblasts of the inner layer (Fig. 3C). Significantly, many MyoD+ cells co-expressed desmin as well (Fig. 3F–H), demonstrating that these cells were differentiating in muscle fibres. No signal for the antibodies anti-MyoD and desmin was detected in the control sections (Fig. 3D,E,I,J). By stage 30 (Fig. 3K–N), the MyoD signal was localized in the nuclei of MyHC+ muscle fibres (Fig. 3K) and in the integument migrating cells (Fig. 3L), in which a low positivity for the MyHC antibody was detected as well.

Figure 3.

A–E: Combined DAPI (blue) and MyoD (red) immunofluorescence images of frontally sectioned mantle. A: Stage 21. MyoD signal is detected in the integument (i). B: Stage 25. Red positive cells are visible in the connective tissue. C: Stage 28. Labelled red nuclei are visible in the myoblast layer (square). D, E: Negative controls for immunofluorescence protocol. No red staining was detected (E). Nuclei were counterstained in blue with DAPI (D). FJ: Combined MyoD (red) and desmin (green) immunofluorescence images of frontally sectioned mantle at different stage of development. Cells co-expressing MyoD and desmin are visible in the integument (F), the connective tissue (G), and the myoblast layer (square in H). I, J: Negative controls for immunofluorescence protocol. KN: Stage 30. Combined MyoD (red) and skeletal MyHC (green) immunofluorescence. Muscle fibres co-express MyoD and MyHC (K). A low signal for MyHC antibody is detected also in the integument (i) and in the migrating cells (L, arrowheads). M,N: Negative controls. Scale bar = 50 μm in A–J, L; 40 μm in K, M, N.

Characterization of Muscle Fibre Types in the Developing Mantle

In order to detect the two major types of muscle fibres (slow and fast) in the developing mantle, we focused on the mantle region that will give rise to the muscular layers, as indicated in the cartoon of cuttlefish frontal section (Fig. 4A).

Figure 4.

A: Schematic drawing of a frontally sectioned cuttlefish embryo. B–M refer to studies performed on the myoblast/myocytes layer (squared area) of each stage. BE: Stage 26. FI: Stage 28. JM: Stage 30. Oxidative slow fibres, as indicated by positive blue reaction with NADH (B,F) and anti-slow MyHC antibody (C,G), are localized in the outer superficial zone of the “myoblast/myocytes” layer. The radial fibres (arrowheads in D,E,H,I) instead are PAS positive (dark pink staining in D,H) and express the fast MyHC (E,I), differentiating into fast glycolytic fibres. J–M: Stage 30. The mantle is mainly formed by oxidative slow circular muscle fibres (J, K) that are regularly divided in fields by thin fast radial fibres (arrowheads in L, M). Fibres differentiated in fast glycolytic fibres (arrows in L,M) are visible in the middle area of the muscular layer. Scale bar = 25 μm in B–E; 50 μm in F–M.

Starting from stage 26 (Fig. 4B–E) up to stage 28 (Fig. 4F–I), within the mass of myocytes the cells facing the external and visceral side of the muscular layer were NADH-positive (Fig. 4B,F) and expressed slow MyHC (Fig. 4C,G), indicating differentiation into slow oxidative fibres, whereas the radial fibres were PAS positive (Fig. 4D,H), expressed fast MyHC (Fig. 4E,I), and were differentiating into fast glycolitic fibres.

Slow fibre differentiation took place centripetally and at hatching (stage 30, Fig. 4J–M). Nearly all the circular muscles were differentiated into slow fibres (Fig. 4J,K) and were regularly divided in blocks by thin fast radial fibres (Fig. 4L,M). In the middle area of the muscular layer, few cells, differentiating into fast fibres, were visible (Fig. 4L,M), in agreement with Bone's description for adult animals (Bone et al.,1981).

Expression of hh Homolog in Different Districts Including the Mantle

In order to evaluate the role of hh signalling in the differentiation of striated muscle fibres in S. officinalis, RNA in situ hybridisation experiments were performed to determine the site of expression of cuttlefish hh ortholog in different stages of development. From stage 23 (Fig. 5A), hh-like transcripts were strongly localized in the head/foot region (which is the locomotory and sensory portion of the body region formed by optical lobes, arms, tentacles, and the funnel tube) and in the mantle. The hybridization signal was stronger in the superficial layer of the large optic lobes, in the ventral arms, in the apical region of the funnel tube, and in the mantle, where the profile of this cup-shaped structure was bordered. Starting from stage 26 (Fig. 5B) up to stage 28, expression of the hh-like transcript persisted in the cephalic region and became even more intense in the ventral side of the mantle (Fig. 5C). At the pre-hatching stage (stage 30), the hybridization signal was fading in the head-foot region, whereas in the mantle it was mainly localized in the anterior region (Fig. 5D). No hybridization signal was detected using a control sense probe either in the early or the late stages of development (Fig. 5E,F).

Figure 5.

A–F: Whole mount in situ hybridisation of S. officinalis embryos with So-hh probe. Embryos are shown in ventral view and with the anterior region at the top. A: Stage 23. So-hh mRNA is highly expressed in the head-foot region formed by the large optical lobes (arrowheads), the arms (asterisk), and the funnel tube (encircled area). In the mantle, So-hh expression was restricted to the lateral edges (arrow). Stage 26 (B) and stage 28 (C): So-hh is located in the arms (asterisk), optical lobes (arrowheads); funnel tube (encircled) and in the ventral side of the mantle (arrows). D: Stage 30. So-hh is localized in the ventral arms (asterisks) and in the funnel tube (encircled), while transcripts are not present in the optical lobes (arrowhead). In the mantle, So-hh expression is fading and a faint signalling is detected at the edges of the mantle and in the apical regions, corresponding to the growing areas (arrows). E,F: Negative control of in situ hybridisation protocol. No signal was detectable in embryos treated with a sense probe. Scale bar = 500 μm A; 1.5 mm in C, F; 2.5 mm in D; 3.5 mm in E, G.

The whole mount in situ stained embryos were then cryosectioned to precisely identify the cells expressing the hh-like transcript. At an early stage of development (stage 23), expression was localized in a subset of superficial ciliated cells of integument and in the scarce cytoplasm of the innermost integument cells (Fig. 6A). Starting from stage 25, transcripts were also detectable in flattened, elongated cells dispersed in the connective tissue (Fig. 6B). At stage 28, the hybridization signal was still present in the integument cells and in those cells migrating through the connective region, but was also localized in the radial differentiating fibres that subdivided the myocytes to form the perspective muscular layers (Fig. 6C). Again, no signal was detected using a sense probe (Fig. 6D–F).

Figure 6.

Cryosections of mantle shown with the integument (i) on the left and the myoblast/myocyte layer on the right. AF: Whole mount in situ stained embryos. hh-homolog transcripts are detected in the integument cells (arrowheads), in the elongated cells dispersed in the connective tissue (empty arrowheads) or lining the myoblast/myocyte mass (B, arrows), and in the radial fibres inside the myoblast/myocyte layer (C, arrows). D–F: No signal was detected using a control sense probe. GI: Combined DAPI (blue) and immunofluorescence for Ptc (red) images. The integument cells (G arrowheads), the cells migrating in the connective tissue (H, arrowhead), those lining the “myoblast layer” (H, arrow), and the differentiating radial fibres (I, arrows) are Ptc+. i: integument. J: Western blot from monodimensional SDS-PAGE. The anti-Ptc antibody detects a band of about 135 kDa in the cuttlefish extract (Lane c) and a band of about 160 kDa in mouse embryo extract (Lane b). Lane a, standard. Scale bar = 50 μm in A–I.

Expression of hh Receptor Patched (Ptc)

We then tried to identify the target cells of hh-like signalling by using an antibody raised against the N-terminal region of the vertebrate Hh receptor, Ptc. This region of the Ptc protein was chosen since a BLASTP search carried out using the entire amino acid sequence identified a high grade of conservation in this region in metazoans. In cuttlefish, the expression pattern of a Ptc-like receptor was initially localized in the ciliated and in the innermost cells of the mantle integument (Fig. 6G). Starting from stage 25 (Fig. 6H), Ptc-like signal was detected also in the migrating cells. At stage 28, Ptc-like signal was expanded in the “myoblast layer” and was specifically localised in the radial differentiating fibres, while the mass of myocytes (forming the circular layer) was Ptc negative (Fig. 6I).

To support the specificity of antigen recognition, a Western blot analysis was performed to compare a protein extract from a stage-28 cuttlefish (when the Ptc-like signalling was highly expressed) with that from a mouse embryo, used as positive control. Figure 6J (lanes b, c) showed that the anti-Ptc antibody detected, as expected, a band of about 160 kDa in mouse (lane b) and a band of about 140 kDa in cuttlefish embryo (lane c). Significantly, the molecular weight of Ptc observed in cuttlefish was very similar to that reported for both zebrafish (141 kDa) and Drosophila (146 kDa) Ptc proteins.

The hh Homolog Is Required for Normal Cuttlefish Development

To evaluate the functional significance of hh expression in the cuttlefish developing mantle, we inhibited the Hh signalling pathway with cyclopamine, a known inhibitor of hh signalling in vertebrates (Cooper et al.,1998; Incardona et al.,1998,2000) and invertebrates (Kang et al.,2003). Cyclopamine treatment from stage 23 (when a small mantle was already present) until stages 26 and 28 led to a high grade of embryo malformation both in the head/foot (formed by arms, funnel tube, optical lobes) and mantle regions. In particular, when compared to controls (Fig. 7A–F), the treated embryos (Fig. 7G–L) showed shorter arms, eyes without transparent corneas, and funnel tube directed towards the posterior region. Moreover, the mantle size was greatly reduced in 80% of embryos (16 embryos out of 20) exposed to 10 μM cyclopamine, especially in the ventral side (Fig. 7G), where the mantle could not completely cover the internal organs, such as gills and gut, and did not reach the head/foot region.

Figure 7.

Comparison between control (AF) and cyclopamine-treated (GL) embryos at stage 28. Embryos in A and G are shown both in ventral and lateral view and with the anterior region on the top of the figure. Treated embryo (G) shows shorter arms (asterisks) and high reduced mantle (outlined in the ventral view), exposed funnel tube (arrowheads), gut (g), and gills (white arrows) as compared with control one (A), where the apical region of the mantle (outlined in ventral view) reach the anterior region. B,C,H,I: Sectioned mantle of control (B,C) and cyclopamine-treated (H,I) embryos. The precise arrangement of the myoblast/myocyte layer (squares) is missing and optical (H, arrowheads) and TEM analysis (I) show the presence of apoptotic nuclei in the cyclopamine-treated embryos. D,J: TUNEL assay in control (D) and treated (J) embryos. Arrowheads in J indicate nuclei stained with DAPI and characterized by DNA fragmentation. E,K: Negative control of TUNEL assay. F,L: BrdU staining in control (F) and treated (L) embryos. Labelled nuclei (arrowheads) are visible in the integument (i) and in the connective tissue (asterisk) of untreated embryos only (F). M,N: Combined DAPI (blue) and immunofluorescence for Ptc (red) images of frontally sectioned mantle in control (M) and cyclopamine-treated (N) embryos. Ptc signal is down-regulated by cyclopamine treatment and is practically absent in the inner muscular layer (square). O,P: Immunofluorescence staining using the antibodies anti-fast (O) and anti-slow (P) MyHC shows that cyclopamine reduces muscle fibre formation. Scale bar = 2.5 mm in A, G; 50 μm in B, D, E, F, H, L, J, K, M–P; 4 μm in C, I.

Morphological abnormalities in the mantle were also investigated in frontally sectioned, stage-28 cyclopamine-treated embryos. The gross anatomy was considerably compromised, due to the reduction in number of cells in the muscle layer (Fig. 7H). In particular, the myoblasts/myocytes, which in control embryos at stage 28 (Fig. 7B) were normally organized in a compact sheath and disposed in rows separated by the differentiating radial fibres, in cyclopamine-treated embryos were scattered and randomly distributed in a loose connective tissue (Fig. 7H). Moreover, many cells migrating from the integument to the connective tissue (Fig. 7H) and several cells in the myoblast layer (Fig. 7H, I) showed apoptotic features, as validated by TUNEL staining (Fig. 7J). In contrast, no apoptotic nuclei were detected in the mantle of control embryos (Fig. 7B–D) and in the negative control sections (Fig. 7E,K).

In order to demonstrate that hh signalling acts both as a myoblast survival factor and as myoblast mitogen, we performed a proliferation assay using BrdU incorporation. In untreated control embryos (stage 28), we detected mitotic cells mainly in the integument and in the connective tissue (Fig. 7F). On the contrary, in cyclopamine-treated embryos at the same developmental stages, mitotic muscle precursors were absent in all the districts mentioned before (Fig. 7L).

As expected, Ptc expression, which is normally up-regulated by Hh, was greatly reduced in cyclopamine-treated embryos, thus providing further evidence that such a treatment interferes with the hh pathway in cuttlefish. The scaffold made of elongated radially disposed fibres, typical of the control animals (Fig. 7M), was almost absent in treated embryos (Fig. 7N) where also the expression of fast and slow MyHC in the mass of myoblast/myocytes was highly reduced (Fig. 7O,P).

Taken together, the experiments described above provide a strong support for a role of Hh signalling in the proper development of the cuttlefish mantle muscular compartment.


We have employed morphological, immuno-histochemical, histoenzymatic, and in situ hybridization assays to obtain information on the mechanisms involved in muscle patterning during development of the cuttlefish S. officinalis. Since the Hh signalling pathway in vertebrates regulates differentiation and positioning of muscle fibres (Blagden et al.,1997; Duprez et al.,1999; Barresi et al.,2000; Grimaldi et al.,2004a), we have considered it as a good candidate for regulating muscle patterning in cuttlefish mantle.

Muscle Fibres Differentiation in Cuttlefish Development

We first delineated the sequential events during development of cuttlefish mantle muscles, which involves a complex network of cell migration and differentiation.

From its early stages (stages 21, 23) of development, the mantle is composed of a compact mass of blast cells enveloped by a multilayered integument formed of ciliated and rounded blast cells.

A key step of muscle mantle differentiation and organization consists of the morphological change and migration of several integument cells. By stage 25 (and continuing until pre-hatching stage 30), several cells located in the superficial integument become irregularly shaped, lose contact with the neighbouring cells, and migrate through the connective tissue towards and among the inner myoblast layer. These integument-derived cells are morphologically different: some elongated and others rounded. Starting from stage 26, differentiating muscle fibres are visible in the muscular layer: the fast glycolitic elongated radial fibres, deriving from the elongated myocytes, and the slow oxidative circular muscle fibres, deriving from the migrating rounded myocytes and from the pre-existing rounded myoblast in the muscular layer.

As development proceeds, the rounded myocytes are subdivided in parallel rows by the radial fibres and differentiate in a centripetal direction, starting from the more superficial regions of the layer, towards the inner region, giving rise to the superficial slow and deeper fast circular fibres. At the pre-hatching stage (stage 30), in the mantle muscular layer, the well-developed fibres, belonging to obliquely striated fibres (Bone et al.,1981), show their metabolic characteristics and express distinct isoforms of myosin heavy chain proteins (MyHC). In fact, fibres of both outer and inner zones of the developing mantle are oxidative (NADH positive) and express slow MyHC, while the central and the radial fibres are glycolitic (PAS positive) and fast MyHC positive. Such distribution, organization, and different mitochondrial enzyme activity of differentiating muscle fibres during mantle development is in accordance with the muscle organization already described by Bone and colleagues (1981) in the mantle of adult animals. Moreover, the arrangement of the muscle fibres of mantle cuttlefish shows impressive similarity with the organization of fibres types in the myotomal locomotor muscle fish (Bone et al.,1981) and can be considered another example of convergence between the two taxa as already described by Packard (1972). Summarizing, the development of muscle fibres in cuttlefish mantle apparently occurs in distinct phases: differentiation and migration of muscle precursor cells from the external integument, maturation of the muscle fibres inside the muscular layer. In cuttlefish, the integument seems therefore to represent a source of muscle precursors, as suggested by the presence of cells co-expressing the muscular regulatory factor MyoD, the intermediate filament desmin and skeletal MyHC within this compartment. These muscle precursor cells subsequently migrate into the “myoblast mass” and contribute to mantle muscle growth and spatial organization. For this reason, we speculate that such an embryonic structure could play a similar role of the layer of cells that Devoto et al. (2006) have described in fish. This cell layer located underneath the epithelium and lining the surface of the somite is formed by myogenic precursors that migrate into the myotome, where they differentiate in muscle fibres. This structure could be compared to the dermomyotome of amniotes (Kahane et al.,1998), which is an ancient and conserved structure that provides a simple and adaptable mechanism for sequestering stem cells and at the same time contributing to muscle growth (Devoto et al.,2006).

In cuttlefish, the radial fibres may act as a scaffold for the spatial organization of myocytes that will give rise to the mass of slow and fast circular fibres. The formation of a primary muscle grid and its subsequent use as a template for differentiation of myoblasts are two basic developmental phases that have already been described in vermiform Spiralia such as Platyhelminthes (Ladurner and Rieger,2000) and Anellids (Jellies and Kristan,1988), and the patterning of body wall muscles is achieved without using originally positional information of the nervous system (Ladurner and Rieger,2000).

A hh Signalling Pathway Is at Work in the Mantle and It Is Involved in Muscle Patterning

The recruitment of hh signalling has long been established as an evolutionary conserved mechanism for the proper developmental patterning of the muscle fibres. We detected the expression of a cuttlefish hh homolog transcript also in S. officinalis embryos, in ectodermal and mesodermal tissues (head-foot and mantle region), as demonstrated for other animals (Ingham and McMahon,2001). In situ hybridization experiments using an S. officinalis cDNA probe highlighted that hh expression is detectable early in cuttlefish development, associated to the eyes, arms, and mantle. Moreover, the overlapping expression of both hh and its receptor Ptc in cuttlefish suggests that Hh signalling is mediated through the interaction with Ptc, as already described in several systems such as zebrafish fin regeneration (Laforest et al.,1998), posterior mesenchymal cells of chick (Marigo et al.,1996), mouse limb buds (Goodrich et al.,1996), and developing mouse retina (Jensen and Wallace,1997).

Expression of both Ptc and the hh in S. officinalis, starting from stage 26 in the differentiating radial myocytes, seems to be critically required for the normal development of the muscle fibres of the mantle, as validated by experiments with cyclopamine, which blocks the hh pathway (Cooper et al.,1998; Kang et al.,2003).

Cyclopamine-treated embryos show a high degree of malformation of several structures, including arms and eyes, and a peculiar disorganization of mantle muscular mass following the down-regulation of Ptc/hh signalling. In fact, the reduced number of Ptc+/hh+ cells leads to a loss of the radial network of fibres responsible of spatial organization of the whole muscular layer. Moreover, Hh signalling pathway appears to be essential for morphogenetic processes including cell survival. The impairment of the Hh signalling pathway leads to a massive apoptotic phenomenon involving integument migrating cells. As a consequence of this defect in cell survival, there is a loss of thickening and patterning in the muscular layer. In addition, following interference of this Hh signalling pathway, no proliferation activity was detected in those cells migrating from the integument through the connective tissue. Thus the loss of organization of the mantle muscle layer may be an indirect effect due to the loss of muscle cells in this layer. Taken together, this evidence explains the observed reduction of the muscular layer in cyclopamine-treated embryos.

In conclusion, we hypothesize that during cuttlefish development an hh signal may act both as a survival and a proliferation factor for myogenic precursor cells during muscle mantle growth paralleling previous findings in vertebrates, where Hh acts primitively as a survival and proliferation factor for myogenic precursor cells and not as a primary inducer of myogenesis (Duprez et al.,1998; Marcelle et al.,1999; Kruger et al.,2001). The absence of an Hh-like signal leads to an apoptotic process, which could be Ptc-mediated, as already shown during chick neural tube development (Thibert et al.,2003). Taken together, our data suggest that the Hh/Ptc signalling pathway is highly conserved in mollusc muscle development controlling patterning, cell death, and survival processes. Thus, we speculate that during evolution the Hh pathway played a pivotal role in the arrangement of body-wall musculature even in soft-bodied animals, like molluscs and leeches (Kang et al.,2003) and in creating different body plans.


Embryo Preparation and Manipulation

Fertilized eggs of cuttlefish (Sepia officinalis, Mollusca, Cephalopoda) were kept in water at 18°C. Embryos belonging to different stages of development were selected according to Boletzky (1987), by measuring their dorsal mantle length (DML), and to Lemaire (1970). by characterisation for developmental landmarks.

The cuttlefish were anaesthetized with a 1:1 solution of 7.5% MgCl2.6H20 and seawater (Messenger et al.,1985) prior to carrying out any procedures or fixations.

Light Microscopy and Transmission Electron Microscopy (TEM)

Embryos were fixed for 2 h in 2% glutaraldehyde in sea water at 4°C. The specimens were then washed in sea water and post-fixed at 4°C for 2 h with 1% osmic acid in cacodylate buffer (pH 7.2). After standard dehydration in an ethanol series, specimens were embedded in an Epon-Araldite 812 mixture. Semi-thin and thin sections were stained by conventional methods, using crystal violet and basic fuchsin or uranyl acetate and lead citrate, respectively. The observations were made with a light microscope (Olympus) or with a Jeol 1010 electron microscope (Jeol).


Embryos were embedded in polyfreeze cryostat embedding medium (OCT; Polysciences) and rapidly frozen in liquid Nitrogen. Cryosections (7 μm) of cuttlefish embryos at different stages were obtained with a Reichert Jung Frigocut 2800 (Leica) and were stained with purple-blue formazan utilizing histoenzymatic kits (Bio-Optica) for NADH-diaphorase to detect the mitochondrial activity, and with Schiff's Reagent utilizing histoenzymatic kits (Bio-Optica) for periodic acid Schiff (PAS) to highlight the presence of glycogen.


Embryos were embedded in polyfreeze cryostat embedding medium (OCT; Polysciences) and rapidly frozen in liquid Nitrogen. Cryosections were treated for 30 min with PBS containing 2% bovine serum albumin (BSA) before the primary antibody incubation (37°C for 1 h). The primary antibodies used were the polyclonal antibodies goat anti mouse Ptc (1:50, Santa Cruz Biotechnology), rabbit anti mouse MyoD (1:20, Santa Cruz Biotechnology), which have been demonstrated already to cross-react with obliquely striated muscles fibres in cuttlefish (Grimaldi et al.,2004c), goat anti-mouse desmin (1:20 Santa Cruz Biotechnology), rabbit anti-human skeletal Myosin (1:20 Sigma), and the monoclonal antibodies mouse anti-human fast and slow myosin MyHC (1:20, Sigma) that have been proved already to react with cuttlefish muscles (Grimaldi et al.,2004a). The washed specimens were incubated for 1 h at room temperature with the appropriate secondary antibody Cy3 or FITC conjugated (dilution 1:100 in blocking solution) (Jackson, Immuno Research Laboratories). Control sections were incubated with PBS/BSA without the primary antibody and then with the secondary antibody. Nuclei were stained by incubating for 15 min with 4′,6-Diamidino-2-Phenylindole (DAPI, 0.1 μg/ml in PBS). Coverslips were mounted in Vectashield mounting medium for fluorescence (Vector Laboratories); slides were examined with a confocal laser microscope Olympus. Images were combined with Adobe Photoshop.

5-Bromodeoxyuridine (BrdU) Labeling

Cell proliferation was monitored using DNA incorporation of the substituted nucleotide, 5-BrdU. Five cuttlefish embryos treated with cyclopamine and five untreated control embryos from stage 28 were immersed in BrdU dissolved in seawater at a final concentration of 0.05% for 6 hr. After BrdU incubation, the animals were prepared using the procedure described above for histochemistry and immunochemistry.

Cryosections were treated for 30 min with PBS containing 2% bovine serum albumin (BSA) before the primary antibody incubation (37°C for 1 hr). They were incubated overnight at 4°C with a monoclonal antibody against BrdU (Amersham, Buckinghamshire, UK) diluted 1:100 in PBS. After washing, they were treated with 0.3% H2O2 in PBS to exclude the potential activity of endogenous peroxidase. Washed semi-thin sections were incubated for 3 hr with peroxidase anti-mouse IgG at room temperature and, after washing in PBS, incubated with 0.05% 3,3′-diaminobenzidine and 0.03% H2O in PBS. They were then rinsed in distilled water. Control reactions were carried out by omitting the primary antibody.

Biochemical Procedures

Cuttlefish embryos (stage 28) were homogenised in liquid nitrogen with a mortar. Mouse embryos, used as positive controls, were homogenised following the same protocol. For SDS-polyacrilamide gel electrophoresis (SDS-PAGE), homogenates were suspended in extraction buffer (0.4 M NaCl, 5 mM MgCl2, 1 mM EDTA, 0.5 μg/ml pepstatin, 50 mM Tris HCl, pH 7.2, with 1 mM phenylmethylsulphonyl fluoride and 0.2 mM ATP freshly added); particulate material was removed by centrifugation at 13,000 rpm for 10 min at 4°C in a refrigerated Sorvall RCM14 microcentrifuge. Supernatants were denatured in sample buffer 2× (2% SDS, 10% β-mercaptoethanol, 0.002% bromophenol blue, 20% glycerol, 0.02 M Tris HCl, pH 6.8) at 100°C for 5 min.


Analytical SDS-PAGE using 15% acrylamide minigel was made according to (Laemmli,1970). Molecular weights were determined by concurrently running the standard PageRuler™ Prestained Protein Ladder (Fermentas). Electrophoresis was made at 200 V for 1 hr.

Western Blot

Proteins separated by SDS-PAGE were transferred onto Bio-Rad nitrocellulose filters according to (Towbin et al.,1992). Before immunostaining, the membrane was saturated with 5% BSA in Tris-buffered saline (TBS, 20 mM Tris-HCl buffer, 500 mM NaCl, pH 7.5) at room temperature for 2 h. Nitrocellulose sheets were incubated overnight at 4°C with a polyclonal goat anti-mouse Ptc as primary antibody (Santa Cruz Biotechnology) at 1:250 dilution in 2% TBS-BSA. After three washes of the membrane with TBS, antigens were revealed with a donkey anti-goat secondary antibody (1:500) peroxidase-conjugated (Jackson ImmunoResearch Laboratories). After washing with TBS, signal was developed with chemiluminescent substrate (Pierce).

Cloning of So-hh Gene

A BLASTP search using the P. vulgata Hh amino acid sequence as a query was performed in order to identify the most highly conserved region across several species. Within this region (spanning residues 103–175 in the snail Hh protein), several degenerate primers were designed to amplify cDNA fragments from the cuttlefish ortholog Hh gene by RT-PCR. Experimental conditions were as follows: 3 μg of cuttlefish embryonic total RNA were reverse transcribed with random hexamers using the High Capacity cDNA Reverse Transcription kit (Applied Biosystems) at 37°C for 2 h. Thereafter, 45 cycles of PCR were performed with the following parameters: denaturation: 94°C for 30”; annealing: 45°C for 30”; extension: 72°C for 30”. In a preliminary effort, we could obtain just a short 219-bp-long cDNA fragment from cuttlefish stage-28-embryo total RNA (accession number EF651782) using the following primer pair for RT-PCR:



However, further attempts to clone larger S. officinalis Hh cDNA fragments were carried out subsequently and resulted in the cloning of a 465-bp-long fragment. A combination of 5′/3′ RACE and DOP-PCR was used to extend the previously cloned 219-bp sequence in both directions. A nested-DOP PCR performed with both degenerate and cuttlefish-specific primers was particularly useful for this purpose and was carried out with the following primers:


  • Sepia O. HH rev 1 5′-CTTCGACTGCTAATCTGG-3′

  • Sepia O. HH rev 2 5′-CTCGTAATTTCACCCCAGGC-3′

By means of these experiments, a 465-bp cDNA fragment was obtained (accession number EF651782), and subsequently used for in situ hybridization and Northern blot assays.

Whole Mount In Situ Hybridization

Embryos were fixed in MEMFA (0,1M MOPS, pH 7.4, 2 mM EGTA, 1 mM MgSO4, 3.7% formaldehyde) for 1 h and stored in methanol at −20°C until use. Whole mount in situ hybridization assays were performed using digoxigenin-labeled probe representing the sense and antisense RNAs obtained following T7- and SP6-mediated in vitro transcription (MEGAscript kit, Ambion) from the linearized 465-bp-long S. officinalis hh cDNA clone described above (accession number EF651782). Hybridization was carried out as previously described (Kang et al.,2003). Embryos were embedded in OCT and rapidly frozen in liquid Nitrogen. Cryosections were mounted with glycerol/PBS.

Cyclopamine Treatment

Cyclopamine (BIOMOL Research Laboratories), a known blocker of hh signalling both in vertebrates and in invertebrates (Cooper et al.,1998; Incardona et al.,1998,2000; Kang et al.,2003), was diluted to a final concentration of 10 μM in sea water (from a stock solution 10 mM in ethanol). Experimental embryos (early stage 21) were cultured in cyclopamine for 10–15 days (up to stages 26–28). Control embryos were cultured in sea water with 0.1 % ethanol.


Cryosections from unfixed embryos were treated with DeadEnd™ colorimetric TUNEL System assay kit (Promega Corporation) according to the manufacturer's instructions to visualize the fragmented DNA of apoptotic cells. Nuclei were counterstained with DAPI. For negative controls, Terminal Deoxynucleotidyl Transferase Recombinant (rTdT) was omitted.


The authors thank Prof. Douglas Noonan and Dr. Andrea Moriondo for critical reading and editing the manuscript, Dr. Patrizia Basso, Manuela Viola, and Gabriella Greco for their excellent technical assistance, and Prof. Giulio Lanzavecchia and Enzo Ottaviani for helpful discussion on the manuscript. This work was supported by University of Insubria FAR (Fondi dell'Ateneo per la Ricerca) 2004 and 2005 to A.G.