Ontogenetic torsion in two basal gastropods occurs without shell attachments for larval retractor muscles.

Results of this study on two species of vetigastropods contradict the long-standing hypothesis, originally proposed by Garstang (1929), that the larval retractor muscles power the morphogenetic movement of ontogenetic torsion in all basal gastropods. In the trochid Calliostoma ligatum and the keyhole limpet Diodora aspera, the main and accessory larval retractor muscles failed to establish attachments onto the protoconch (larval shell) when the antibiotics streptomycin sulfate and penicillin G were added to cultures soon after fertilization. Defects in protoconch mineralization were also observed. Despite these abnormalities, developing larvae of these species accomplished complete or almost complete ontogenetic torsion, a process in which the head and foot rotate by 180 degrees relative to the protoconch and visceral mass. Analysis by using phalloidin-fluorophore conjugate and transmission electron microscopy showed that myofilaments differentiated within myocytes of the larval retractor muscles and adherens-like junctions formed between muscle and mantle epithelial cells in both normal and abnormal larvae. However, in abnormal larvae, apical microvilli of mantle cells that were connected to the base of the larval retractor muscles failed to associate with an extracellular matrix that normally anchors the microvilli to the mineralized protoconch. If morphogenesis among extant, basal gastropods preserves the original developmental alteration that created gastropod torsion, as proposed by Garstang (1929), then the alteration involved something other than the larval retractor muscles. Alternatively, the developmental process of torsion has evolved subsequent to its origin in at least some basal gastropod clades so that the original alteration is no longer preserved in these clades.