A Biomolecular Survey of Calcification 117 



structure and characteristic bulbous termini, which may become associated with the 

 inner surface of the shell to be incorporated into the familiar lacelike mosaic of 

 conchiolin reported by Gregoire (1957). 



Calcium phosphate 



Raphide formation in plants is essentially an intracellular process and shell for- 

 mation in molluscs in possibly so, but in bone, dentine and enamel, calcification is 

 generally considered to be an extracellular process in which the role of the cell is to 

 provide the matrix. Precisely how the mineral gets into the "matrix" is not clear, 

 but it is assumed that the ions migrate somehow from the serum fluids. In other 

 structures containing bone salts, however, the relationship between the cell and the 

 mineralization front is quite clear. In Lingula, where the crystallites develop in 

 laminae within the shell, the mantle cells are firmly attached to the inner wall. In 

 baleen, the deposits of salt are within the cell and there is no evidence for dis- 

 continuity between mineral and keratin (Pautard, 1965). The close connection 

 between the intracellular deposits of apatite in baleen and the extracellular deposits 

 in bone leaves us with an ultrastructural anomaly which now seems to be resolved. 

 In baleen, the crystallites are usually surrounded by less dense, leaf-like structures 

 (arrowed D in Fig. 3 a inset) whereas in the past no comparable structures have been 

 observed in bone, although opinion as to the fine structure of the calcifying front has 

 varied. Earlier reports (Robinson and Cameron, 1956) suggest that crystals are laid 

 down "within a fraction of a micron of the bone cells"; later, Sheldon and Robin- 

 son (1957) reported that the osteoblast could carry out metabolic transfer between 

 the extracellular space and the calcified regions. Recently, it has been stated that 

 there is a clear zone between osteoblast and mineral (Ascenzi et al., 1963; Frank, 

 1963) while Cameron (1961) comments that mineral appears in cartilage beyond 

 invading capillaries and at a distance from any cell. On the other hand, Dudley and 

 SpirO' (1961) observed structural modifications of the cell surface at the site of 

 contiguity, and Hancox and Boothroyd (1964) have described short processes 

 extending from the cell to the bone edge. 



Our present experiments on newborn mouse bone (Arnott et al., in press) using 

 modifications of the fixation techniques recommended by Sabatini et al. (1964) 

 suggest that there is a considerable amount of ultrastructure within the osteoblast 

 and within numerous filamentous processes which spread from the cell to the 

 mineralized front. Moreover, where a section has been cut by good fortune in the 

 plane of the filament, the filopodia can be traced directly into the body of the 

 crystallites as a delta of electron-dense material (arrowed D in Fig. 3 b) which closely 

 resembles the leaf-like structures (c. f. Fig. 3 a) so often observed in baleen. In these 

 areas, a faint electron diffraction pattern of apatite can be observed; nowhere can 

 we find any periodic deposition of "nuclei", either on the collagen fibres which are 

 always present, or as rows of electron-dense dots. The calcifying delta simply appears 

 to spread out, like a stain, into the organic substances adjacent to the filopodia. 



In enamel, present opinion favours the view that the ameloblast is separated from 

 the mineral front by a continuous membrane. Some authors (Frank et al., 1960; 

 Travis and Glimcher, 1964) state that the organic phase forms a framework which 

 faithfully copies the outlines of the crystallites; there is no suggestion as to whether 

 this framework is created by the crystals or by the cell. Ronnholm (1963), on the 



