Archs oral Bid. Vol. 31, No. 5, pp. 287-296, 1986 0003-9969/86 $3.00 + 0.00 Printed in Great Britain. All rights zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA reserved Copyright 0 1986 Pergamon Press Ltd zyxwvu PENETRATION OF VARIOUS MOLECULAR-WEIGHT PROTEINS INTO THE ENAMEL ORGAN AND ENA M EL OF THE RAT INCISOR M. D. MCKEE, B. MARTINEAU-DOIZE and H. WARSHAWSKY Department of Anatomy, McGill University, 3640 University St, Montreal, Quebec, Canada H3A 2B2 Summary-During enamel maturation, most of the organic matrix is removed as the mineral content increases; it is postulated that proteolytic enzymes within enamel break down large proteins into more mobile fragments. To predict how such fragments might leave the enamel, the entry and penetration of various proteins into it was examined. Rats (1OOg) were injected via the external jugular vein with ‘251-iodinated calcitonin (3600), insulin (5700) epidermal growth factor (EGF; 6100) and albumin (68,000). They were killed after 10 min and radioautographs made to visualize these molecules in the incisor enamel organ and enamel. In addition, dissected incisors were wiped free of their enamel organs, dipped in the iodinated protein solutions for 10 min, and processed for radioautography. In all dipped teeth, except those exposed to albumin, there was a gradient of silver-grain density over the entire thickness of enamel in both the secretion and maturation zones. In all injected animals, enamel labelling in the secretion zone was only slightly above background. In the maturation zone of animals injected with calcitonin and insulin, many grains were over enamel adjacent to smooth-ended ameloblasts but not ruffle-ended ones. Animals injected with EGF and albumin had no labelled enamel in the maturation zone. Thus dipped rat incisor enamel was permeable to proteins with molecular weights as high as 6100. Localization of injected proteins indicates that the enamel organ restricts their passage into enamel, but proteins with molecular weights as high as 5700 may pass into enamel through or between smooth-ended ameloblasts. As exogenous proteins readily diffused into the enamel, it seems likely that enamel proteins of similar size can leave enamel by a similar route. INTRODUCTION Enamel formation requires a high degree of cellular interaction in the enamel organ (Warshawsky and Smith, 1974). In the secretion zone, the primary function of the organ is to elaborate the organic matrix of enamel (reviewed by Weinstock and Leblond, 1971; Slavkin, Mino and Bringas, 1976; Warshawsky, 1979; Karim and Warshawsky, 1979). In the maturation zone, most of the organic matrix is removed whereas the mineral content, in the form of hydroxyapatite, is increased (Deakins, 1942; Weinmann, Wessinger and Reed, 1942; Allan, 1967; Reith and Cotty, 1962; Robinson, Lowe and Weatherell, 1977). The mechanism whereby organic matrix is removed and the mineral content increased is still not clear. Immediately adjacent to the matu- ring enamel are two distinct forms of maturation ameloblast (Suga, 1959), each with a different distribution and type of intercellular junction (Warshawsky and Smith, 1974; Josephsen and Fejerskov, 1977; Boyde and Reith, 1976, 1977). Several studies have demonstrated the band-like arrangement of smooth-ended and ruffle-ended am- eloblasts across the rat incisor (Takano and Ozawa, 1980; Reith and Boyde, 1981a; Warshawsky, 1985). Other studies have correlated various banding pat- terns in rat incisor enamel with the distribution of the two types of maturation ameloblast (Boyde and Reith, 1981, 1982; Reith and Boyde, 1981b; Takano et al., 1982; Reith, Boyde and Schmid, 1982; Josephsen, 1983; Reith, Schmid and Boyde, 1984). These correlations indicate that the maturation pat- tern of rat incisor enamel is under strict control of the overlying organ. Regulation of the local environment at various body surfaces is an important homeostatic function of epithelial tissues. The tight junction is the structure that limits epithelial permeability through the inter- cellular spaces. The tightness and leakiness of such junctions has been assessed by examining the extra- cellular penetration of tracers such as lanthanum (Garant, 1972; Takano and Crenshaw, 1980; Shaklai and Tavassoli, 1982) and horseradish peroxidase (Skobe and Garant, 1974; Takano and Ozawa, 1980; Kallenbach, 1980a, b). However, some controversy exists as to the dimensions, ionic charges and effects of these molecules. Revel and Karnovsky (1967) assumed that, in their technique, lanthanum is in colloidal form and therefore acts as a passive tracer of extracellular space. However, Schatzki and Newsome (1975) showed that at physiological pH, at least 70 per cent of lanthanum is in the ionic form, and that the deposits observed electron- microscopically may be due to it having either collo- idal or non-colloidal dimensions. In addition, due to its cationic nature, lanthanum may bind to cell glycoproteins, cell-membrane phospholipids and calcium-binding sites (reviewed by Shaklai and Tavassoli, 1982). Horseradish peroxidase (HRP) ex- hibits anomalous behaviour relative to its molecular size (Mazariegos, Tice and Hand, 1984; Mazariegos and Hand, 1985); it may damage cell membranes and junctions through its peroxidative activity and gly- coproteinaceous nature. Furthermore, certain rat strains suffer a histamine reaction with increased vascular permeability after an injection of HRP (Simionescu, Simionescu and Palade, 1975). To avoid the pitfalls of using exogenous tracers, we have employed various normally-occurring proteins 287