Evidence that gastropod torsion is driven by asymmetric cell proliferation activated by TGF-β signalling

1. Introduction

Triploblastic animals have a mostly bilaterally symmetric bodyplan; however, the symmetry is not always complete and most show greater or lesser degrees of left–right asymmetry. Gastropods transform the symmetric bodyplan into a highly asymmetric one through a unique process called torsion.

Torsion involves the 180° rotation of the velum and the foot of the developing larval body relative to the larval shell and visceral mass (see the electronic supplementary material, figure S1). This unique developmental process is a characteristic feature of gastropods, and has intrigued many developmental and evolutionary biologists for over a century as a model case of transition to a novel bodyplan [1–6]. Although the onset of torsion and its completion are readily observed in several species, the underlying mechanism of morphological movement remains to be determined [5–7].

Traditionally, it was widely accepted that torsion was driven by contraction of the larval retractor muscle connecting the foot and shell plate [2,3,6]. However, some recent studies have questioned this idea, because torsion occurred even in larvae where the muscle failed to attach to the shell [7,8]. As an alternative hypothesis, it was proposed that torsion is caused by asymmetric growth of the mantle epithelium which is the surface region of trunk ectoderm posterior to the foot [5,9]. To further investigate asymmetric growth during gastropod torsion, we investigated cell proliferation activity during development in a limpet species, Nipponacmea fuscoviridis.

Gastropods possess another distinct aspect of left–right asymmetry in the handedness of shell coiling. The handedness of shell coiling is known to correlate with the modes of the spiral cleavage pattern (i.e. if the third cleavage micromere is displaced clockwise, as observed in most gastropods, shell coiling is dextral, and vice versa) in most species, with the possible exception of hyperstrophic gastropods [10]. Additionally, shell coiling correlates with the direction of torsion. While dextral species undergo torsion in a counter-clockwise manner, torsion in sinistral species occurs in the opposite direction [4]. A recent study by Grande & Patel [11] found that the asymmetric activation of the Nodal signal is involved in the left–right specification and the formation of shell coiling in gastropods. Interestingly, they reported that the asymmetric expression of Nodal is also observed in limpets, which do not show an obvious handedness in shell coiling even in early larval stages [12]. Thus, we reasoned that the Nodal pathway may also be involved in the direction of torsion.

In this study, we report that in N. fuscoviridis cell proliferation in the larval mantle epithelium is significantly higher on the right-hand side. Additionally, this asymmetric proliferation was sensitive to pharmacological inhibition of transforming growth factor-β (TGF-β) signalling, including Nodal. Thus, we conclude that asymmetric cell proliferation, regulated by TGF-β signalling, is a major driving force of gastropod torsion.

2. Material and methods

Sexually mature individuals of N. fuscoviridis were collected in Yoshidahama harbour, Miyagi Prefecture, Japan, during the breeding season (March–July and October–November). In vitro fertilization was performed following Deguchi [13]. Embryos were cultured in filtered sea water (FSW) at 22°C. Inhibition of TGF-β signalling was performed using SB-431542 (TOCRIS Bioscience) dissolved in dimethyl sulfoxide (DMSO) and cell proliferation was analysed using 5-bromo-2-deoxyuridine (BrdU; Roche). Statistical analysis of cell proliferation was performed using a Mann–Whitney U-test or two-way ANOVA followed by Tukey's test [14].

Nfnodal and Nfpitx were amplified by PCR using the primers shown in the electronic supplementary material, table S2. In situ hybridization was performed as described in Kin et al. [15]. More detailed methodology is shown in electronic supplementary material, S3.

3. Results

First, we briefly describe the time course of the torsion of N. fuscoviridis cultured at 22°C. In this study, we describe the orientation of the larval body axis by referring to the axis of the velum and foot (i.e. lateral view from the right-hand side indicates that an image is viewed from the right-hand side of the velum and foot). Overall, torsion occurred more-or-less synchronously and was completed within 5 h. Until about 22 hours post fertilization (hpf), we did not observe any signs of torsion (see the electronic supplementary material, figure S1a). At 23 hpf, the torsion process had begun and showed an approximate 30–45° rotation of the visceral mass and larval shell relative to the velum and foot. At 25 hpf, the rotation was approximately 90° (see the electronic supplementary material, figure S1b). Larvae completed the 180° rotation by 27 hpf (see the electronic supplementary material, figure S1c).

To examine whether asymmetric cell proliferation occurs during torsion, we measured cell proliferation activity separately in the mantle epithelium and the mantle edge (refer to the electronic supplementary material, figure S1a), and compared the cell proliferation activity between the left and right sides. We examined cell proliferation activity during the first 90° of rotation (from 22 to 25 hpf) because at that stage it is easy to align the degree of rotation, and thus, the number of BrdU-positive cells was counted accurately. We found asymmetric growth in the mantle epithelium, but not in the mantle edge. While an average of about 8.8 BrdU-positive cells (n = 10) were detected on the left-hand side of the mantle epithelium, there was a significantly larger number (approx. 43.2; n = 10) of BrdU-positive cells in the right half of the mantle epithelium (figure 1a). The mantle edge possessed BrdU-positive cells at similar levels to those in the left side of the mantle epithelium (approx. 7.4: left and 7.9: right; n = 10 each), and there was no significant difference between the left and right sides (figure 1b).

• Download figure
• Open in new tab
• Download PowerPoint

• Download figure
• Open in new tab
• Download PowerPoint

We wanted to know whether the Nodal signalling pathway was involved in the asymmetric cell proliferation. As reported for other gastropod species [11], asymmetric expression of nodal and pitx was observed in N. fuscoviridis. The earliest expressions were detected at 8 hpf for both of them, and no expressions detected before 7 hpf (figure 2a,b). The expressions were retained until 11 hpf, and downregulated by 12 hpf. Thus, asymmetric cell proliferation may not be directly controlled by Nodal signalling. We used the chemical inhibitor SB-431542, a known inhibitor of the TGF-β type I receptor and reported to inhibit the Nodal signalling pathway during early development in the snail [11]. When early trochophore larvae of N. fuscoviridis (8 hpf) were treated with SB-431542 until fixation at 27 hpf, torsion was completely suppressed at a concentration of 5 µM (table 1 and figure 2c,d). Thus, we examined whether TGF-β signalling contributed to asymmetric cell proliferation in the mantle epithelium of the limpet; we measured cell proliferation activity during torsion (22–25 hpf) in 5 µM SB-431542 treated larvae from 8 to 22 hpf. We found that the 5 µM SB-431542 treatment in the pre-torsion stage (22 hpf) was sufficient to suppress torsion (100%; n = 28) and resulted in symmetric growth of the mantle epithelium. No activation of cell proliferation was observed in the right-hand side, and cell proliferation occurred at a level comparable with the left-hand side of the control larvae (figure 1c–i). No effect was observed on the left-hand side in drug-treated larvae (figure 1e,h). It is also notable that the retractor muscle apparently developed normally in SB-431542-treated larvae (figure 2d).

4. Discussion

Since the first description of gastropod torsion by Lankester [1], the driving force behind it has been intensively discussed [2,3,5–7]. Traditionally, it was widely accepted that torsion is driven by contraction of the larval retractor muscle that rigidly attached to the calcified shell. Wanninger et al. [6] reported that muscle contraction indeed occurred during torsion. However, this idea was questioned by recent studies which reported that torsion occurred before the larval shell becomes rigidly calcified [7]. Furthermore, Page [8] reported that torsion still occurred in drug-treated larvae that did not form muscle attachments to the shell. We found that asymmetric cell proliferation occurred during torsion in the mantle epithelium. High cell proliferation activity on the right-hand side of the mantle epithelium may be a driving force of torsion, by rotating the shell and visceral mass to the left-hand side relative to the velum and foot (figure 2e).

We found that torsion was prevented by the TGF-β signalling inhibitor SB-431542, and these drug-treated larvae did not show a high cell proliferation rate on the right-hand side of the mantle epithelium. In a previous report, nodal expression of limpet and snail was detected at 2c descant cells [11] which develop into the ectoderm of the right side of the foot and the mantle fold in veliger larval stage [16]. Thus, this expression matches with the region that shows asymmetric cell proliferation. Our report provides evidence that TGF-β signalling (perhaps the Nodal pathway among them) is not only involved in the asymmetry of shell coiling [11], but also in torsion, another aspect of asymmetry, by means of activating asymmetric cell proliferation in the mantle.

Footnotes