The effect of osmium postfixation and uranyl and lead staining on the ultrastrucure of young enamel in the rat incisor.код для вставкиСкачать
THE ANATOMICAL RECORD 207:l-16 (1983) The Effect of Osmium Postfixation and Uranyl and Lead Staining on the Ultrastructure of Young Enamel in the Rat Incisor A. NANCI, P. BAI, AND H. WARSHAWSKY Department ofAnatomy, McGill University, Montreal, Quebec, H3A 2%’ Canada ABSTRACT Enamel crystallites are electron opaque without osmium or heavy metal staining and give a crystalline electron diffraction pattern. Since the opacity and diffraction pattern are abolished from ultrathin sections of young enamel by floating on distilled water (Bishop and Warshawsky, 19821, the possibility that aqueous staining may also remove crystallites was tested. In addition, the effect of osmium postfixation on crystallite structure was examined. Rat incisors fixed by perfusion with a mixture of aldehydes were either nonosmicated or osmicated prior to dehydration. Incisor segments in the region of inner enamel secretion were embedded in the same Epon block to ensure reliable comparison. Osmicated enamel was more intensely stained with toluidine blue and more electron opaque than nonosmicated enamel. No other structural differences were seen. However, crystallites in osmicated enamel were more resistant to grid demineralization and electron beam damage. Routine staining was done by floating sections on solutions of uranyl acetate and lead citrate; sections were also floated on similar solutions from which the heavy metals were omitted. These solutions removed the electron opaque crystallites from the youngest enamel. Stained sections showed electron opaque crystallite-like structures similar to unstained enamel. When sections that were extracted by the solutions from which the metals were omitted were restained, they appeared identical to routinely stained enamel. It was concluded that staining of young enamel removes the crystallites and reveals only the organic matrix. Enamel crystallites are intrinsically electron opaque and therefore do not require staining to be visualized. Yet, few electron microscope studies of young enamel have taken advantage of this property (Nylen and Omnell, 1962; Nylen et al., 1963; Frazier, 1968; Bishop and Warshawsky, 1982). Most studies have used osmium tetroxide as either the primary and only fixative (Quigley, 1959; Reith, 1960; Watson, 1960; Fearnhead, 1961; Ronnholm, 1962a; Travis and Glimcher, 1964; Glirncher et al., 1965; Boyde, 1967; Frank and Nalbandian, 1967), or as a secondary fixative after primary fixation with aldehydes (Selvig and Halse, 1972; Bernard, 1972; Decker, 1973; Kallenbach, 1973; Daculsi and Kerebel, 1978; Leblond and Warshawsky, 1979; Warshawsky et al., 1981). In addition, most of the work used staining with 0 1983 ALAN R. LISS, INC heavy metals. However, no evaluation has been made of the contribution by these agents in producing the final appearance of the enamel. The organic matrix of young enamel is mainly protein and lipid (Stack, 19671, and like other organic constituents, requires heavy metal contrasting to be seen. Thus, stained sections of calcified enamel theoretically should reveal the intrinsic opacity of the inorganic crystallites together with the stained organic matrix. Except for Decker Received September 27, 1982; accepted May 15, 1983. Dr. Nanci’s present address is Departement de Stomatologie, Faculte de Medecine Dentaire, Universite de Montreal, Case Postale 6209, Succ. A, Montreal, Quebec, H3C 3T9 Canada. Address reprint requests to Dr. H. Warshawsky. 2 A. NANCI, P. BAI, AND H. WARSHAWSKY (19731, none of the works using staining could describe the presence or distribution of formed constituents other than crystallites'. Thus, it can be questioned whether it is indeed possible to demonstrate both components simultaneously in the same preparation. In this study, the effect of osmium tetroxide on enamel preservation and structure was evaluated. In addition, the contributions made by the commonly used electron microscope stains, aqueous uranyl acetate and lead citrate, were tested in view of the results of Bishop and Warshawsky (1982) that distilled water can remove crystallites from thin sections of enamel. The results indicate that osmium tetroxide stabilizes the crystallites, but the solutions commonly used to stain sections for electron microscopy dissolve them. Consequently, staining with aqueous solutions of uranyl and lead is incompatible with crystallite preservation, and such procedures cannot reveal the relationship between the young inorganic crystallites and the organic matrix. tion (Bishop and Warshawsky, 1982). These were considered to be calcified sections, that is, they contained electron opaque crystallite-like profiles with no further contrast enhancing procedures. The effect of postfixation with osmium was assessed by comparing aldehyde-fixed nonosmicated with aldehydedxed osmicated specimens. Calcified sections from separate blocks were examined unstained or stained in a standardized way with aqueous solutions of 4% uranyl acetate, pH 4.7, for 10 minutes (Watson, 1958) followed by lead citrate, pH 12.5 (Reynolds, 19631, for 5 minutes. In addition, nonosmicated and osmicated segments of two separate incisors from the same rat were embedded together in a common block such that the enamel organs were in contact. Thus, a single section contained enamel subjected to both treatments and comparisons could be reliably made. In this material, the region of inner enamel secretion was examined. The effect of staining either nonosmicated or osmicated enamel was assessed by comparing unstained calcified sections to stained MATERIALS AND METHODS calcified sections. Male Sherman rats weighing approxiTo identify the organic matrix, calcified mately 100 gm were anesthetized with a n sections were decalcified by floating the gridintraperitoneal injection of nembutal. After mounted sections (Nalbandian and Frank, perfusion with lactated Ringer's solution for 1962; Ronnholm, 1962a; Travis and Glimabout 1 minute through the left ventricle, cher, 1964) on drops of 1%formic acid in 10% the animals were perfused for 10 minutes sodium citrate pH 5.0 (Greep et al., 1948), for with a n aldehyde mixture consisting of 2% 1hour at room temperature. acrolein, 2.5% glutaraldehyde, and 3% forFinally, the effect of the solutions in which maldehyde in 0.06 M sodium cacodylate the heavy metals are incorporated was tested buffer (pH 7.3) containing 0.05% CaC12 or by floating calcified sections for a total of with 5% glutaraldehyde in the same buffer. 15 minutes on solutions identical in compoThe mandibular incisors were dissected and sition to the stains but containing no heavy washed in 0.1 M sodium cacodylate buffer metals (solution-extraction). The pH of the containing 0.05% CaC12, pH 7.3. Some inci- solution for uranyl acetate was 7.6 and for sors were not osmicated while others were lead citrate 12.9. postfixed for 2 hours a t room temperature in All material was examined with a Philips 2% aqueous osmium tetroxide or osmium re- EM 400 operated at 80 kV. duced with potassium ferrocyanide (KarnovHigh resolution electron probe microanasky, 1971). After dehydration in graded lysis was carried out on conventionally preconcentrations of acetone the teeth were pared specimens, mounted on formvar-coated embedded in Epon. No deliberate decalcifi- grids. Grids were examined in a modified cation was performed prior to embedding. JEOL JSM50A scanning electron microscope Thick sections (1 Fm) were cut with glass capable of scanning-transmission operation knives and stained with toluidine blue. These and high spatial resolution electron prob? Xwere used to identify the regions of inner and ray microanalysis, in the range of 100 A or outer enamel secretion that were selected for analysis. Thin sections (gold interference 'The only exception is the so-called stippled material believed color) were cut with a diamond knife and by some to be the organic precursor to mineralized enamel. material is discussed by Nanci and Warshawsky (1983). quickly removed from the distilled water af- Stippled These authors conclude that stippled material is an artifact ter sectioning to prevent crystallite dissolu- caused by breakdown of enamel during fixation. EFFECT OF HEAVY METAL STAINING ON ENAMEL less. The instrument incorporates a n 80 mm2 solid-state Si(Li) Princeton Gamma Tech Xray detector (energy resolution 177 eV) located beneath the specimen stage. Vacuum in the specimen chamber was maintained in the range of 5 x lop7to 4 x lo-' Torr for Xray microanalysis. The microscope was operated at 25 kV with a fixed beam current of 1 x 10-llA. Selected area electron diffraction was done on calcified unstained sections, on calcified stained sections, on solution-extracted sections, and on solution-extracted stained sections. The selected area aperture was 100 pm in diameter, covering approximately 5 pm diameter at the specimen. Diffraction was done with a camera length of 940 mm at 80 kV in the Philips 400 electron microscope. 3 of separate crystallites (Ronnholm, 1962a) or central dissolution (Boyde, 1979). The appearance of the enamel crystallites and the degree of electron opacity in both nonosmicated (Figs. 8, 10) and osmicated (Figs. 9, 11) teeth were similar. In both cases crystallites appeared as either sharp dark lines or flat gray profiles. Effect of Staining Calcified Sections The presence or absence of osmium had only a slight influence on the overall uptake of heavy metal stains by the young enamel. The osmicated enamel was somewhat more dense after staining with uranyl and lead (Figs. 13, 15) than nonosmicated enamel (Figs. 12, 14). In these stained, calcified sections crystallite-like structures appeared as sharp dark lines and flat gray profiles in both nonosmicated (Figs. 12, 14) and osmicated RESULTS (Figs. 13,15) enamel. Staining does, however, Postfixation With Osmium Tetroxide impart a granularity to both the sharp dark The regions of inner and outer enamel se- lines and the flat gray profiles (Figs. 12-15). cretion were examined. In the case of inner Grid Decalcification With Formic Acid enamel secretion, a nonosmicated tooth segWhen grid-mounted sections from regions ment was embedded together with a n osmicated segment (Fig. 1)to avoid any differ- of inner (Fig. 16) and outer (Fig. 18) enamel ences that may result from variations in sec- secretion were floated on formic acid all intion thickness and the subsequent handling trinsic electron opacity was removed. Stainof the sections. One-micrometer thick Epon ing the formic acid demineralized sections sections stained with toluidine blue showed with aqueous uranyl acetate and lead citrate that osmicated enamel (Figs. 1,3)bound more restored sharp dark lines and flat gray prostain than nonosmicated (Figs. 1, 2). Also, files to the sections (Figs. 17, 19). These crysthere was a marked shrinkage in cell height tallite-like profiles were similar in size, shape when osmium was omitted (cf. Figs. 2,3). and distribution to crystallites in calcified Young enamel crystallites from nonosmi- unstained (Figs. 5, 7, 24, 27, 30, 32) and calcated teeth (Figs. 4,6) were more susceptible cified stained preparations (Figs. 13, 15, 31, to dissolution in the distilled water of the 33). Since formic acid dissolves enamel crysknife trough than crystals from osmicated tallites, the electron opaque profiles that teeth (Figs. 5, 7). Nonosmicated crystallites were restored to decalcified sections must were also more susceptible to electron beam represent binding of the heavy metals to the damage than osmicated crystallites. Crystal- organic matrix. lites in the commonly embedded segments of Effect of Solutions Without Heavy Metals inner enamel secretion were systematically (Solution-Extraction) exposed to beam irradiation for 5 minutes. The nonosmicated crystallites (Fig. 8) were Enamel crystallites in calcified sections extensively sublimated, whereas osmicated from regions of inner (Figs. 24, 30) and outer crystallites showed few sublimation holes (Figs. 27, 32) enamel secretion appeared as (Fig. 9). Similar results were obtained on sharp dark lines or pale gray profiles. When crystallites from separately embedded seg- calcified sections were floated on solutions ments in the region of outer enamel secretion without heavy metals (Figs. 20, 21, 25, 28) (Figs. 10, 11). Sublimation created small dis- the sharp dark lines and pale gray profiles crete holes over the flat surfaces of the rib- were abolished in the region of young enamel bonlike crystallites. Sublimated crystallites crystallites. The opacity of older crystallites, viewed on edge gave the illusion of a central located away from Tomes' processes, was only clear space (Fig. 8). This central clear space sporadically abolished (Figs. 20,21,25). Conwas previously interpreted as either fusion ventional staining of these solution-ex- 4 A. NANCI, P. BAI, AND H. WARSHAWSKY Fig. 1. A nonosmicated tooth segment (top) is embed- was stained with toluidine blue. Note that osmicated tisded together with an osmicated segment (bottom) in a sues bind more stain than nonosmicated tissues. A, Amecommon Epon block. Both incisor segments arc from the loblasts; D, dentin; E, enamel; 0, odontoblasts. ~ 8 0 . region of inner enamel secretion. This l-Km thick section Figs. 2 , 3 . The enamel organs from the nonosmicated (Fig. 2) and the osmicated (Fig. 3) tooth segments shown in Figure 1. Equal magnification of comparable regions shows a marked shrinkage in cell height when osmium is omitted. The layer of enamel (El is more intensely stained in the osmicated tooth. A, Ameloblasts. X 800. EFFECT OF HEAVY METAL STAINING ON ENAMEL Figs. 4-7. Electron micrographs illustrating the susceptibility of young nonosmicated crystallites to dissolution in distilled water (Figs. 4, 6) and the degree of protection provided by osmication (Figs. 5,7). The youngest crystallites make up the interrod enamel prongs (ir) adjacent to Tomes’ processes (T) in the regions of inner (Figs. 4, 5) and outer (Figs. 6, 7) enamel secretion. Figures 4 and 5 are from commonly embedded incisor segments. The section was floated for a routine length of 5 time on the trough water. Electron opaque crystallites have been removed from the nonosmicated specimen leaving clear spaces in the interrod enamel prongs (Fig. 4). No clear spaces are noted in the osmicated segment in that section (Fig. 5). Instead, opaque crystallites are seen throughout the enamel. A similar loss of crystallites occurs in nonosmicated outer enamel (Fig. 6) but not in the osmicated counterpart (Fig. 71. R, Rod. x 10,023. 6 A. NANCI, P. BAI, AND H. WARSHAWSKY Figs. 8-11. Detailed view of somewhat older enamel crystallites (further removed from Tomes’ processes) from nonosmicated (Figs. 8, 10) and osmicated enamel (Figs. 9, 11). Crystallites in the commonly embedded segments of inner enamel secretion were systematically exposed to beam irradiation for 5 minutes. The nonosmicated crystallites (Fig. 8 ) are extensively sublimated, whereas the osmicated crystallites show few sublimation holes (Fig. 9). Similar resistance to beam damage is seen in the crystallites of outer enamel secretion (compare Fig. 10 to Fig. 11).Sublimination creates small discrete holes over the flat surfaces of crystallites. Sublimated crystallites viewed on edge give the illusion of a central clear space (Fig. 8, arrows) previously interpreted as either fusion of separate crystallites or central dissolution. ~45,562. EFFECT OF HEAVY METAL STAINING ON ENAMEL Figs. 12-15. Effect of staining calcified sections with aqueous solutions of uranyl acetate and lead citrate. The presence or absence of osmium appears to have only a slight influence on the overall staining density of the young enamel. The nonosmicated enamel (Figs. 12, 14) is paler than the slightly more dense osmicated enamel (Figs. 13,15). Heavy metal staining does not profoundly 7 alter the unstained images, as sharp dark lines and flat gray profiles are still seen. Evidence of beam damage was never seen in these very young “stained crystallite” images. Figures 12 and 13 are from commonly embedded segments of inner enamel secretion close to Tomes’ processes (TI.Figures 14 and 15 are from outer enamel secretion. ~ 4 5 , 5 6 2 . 8 A. NANCI, P. BAI, AND H. WARSHAWSKY Figs. 16-19. Extraction of calcified sections with formic acid removes all intrinsic crystallite opacity both in inner enamel secretion (Fig. 16) and outer enamel secretion (Fig. 18).The density seen in Figure 18(arrows) is due to potassium ferrocyanide reduced osmium. Staining these formic acid decalcified sections with aqueous uranyl acetate and lead citrate restores sharp dark lines and flat gray profiles (Figs. 17, 19).The granular texture of these gray images is identical to that seen in stained calcified sections (Figs. 12-15). Indeed, the images are virtually indistinguishable. Since the profiles in Figures 17 and 19 cannot be due to hydroxyapatite, they must, therefore, represent the organic matrix. T, Tomes’ process. ~ 4 5 , 5 6 2 . EFFECT OF HEAVY METAL STAINING ON ENAMEL Figs. 20, 21. Effect of the solutions in which uranyl and lead are incorporated on inner enamel as seen in the commonly embedded nonosmicated (Fig. 20) and osmicated (Fig. 21) segments. This procedure employed solutions and times identical to staining with uranyl and lead, but without these heavy metals. The opacity due to the young crystallites is removed from the interrod prongs (ir)of both preparations. T, Tomes’ process; R, rod. x 13.750. 9 Figs. 22,23. When solution-extracted sections are restained with heavy metal-containing solutions (Figs. 22, 23), electron opacity is restored and the images are identical to those seen in formic acid decalcified and stained sections (see Figs. 17, 19).It must again be concluded that such opacity cannot be due to hydroxyapatite, but must represent the binding of heavy metals to the organic matrix of enamel. T, Tomes’ process. ~ 4 6 , 1 2 5 . 10 A. NANCI, P. BAI, AND H. WARSHAWSKY tracted sections restored electron opaque images similar in size, shape, orientation, and distribution (Figs. 22, 23, 26, 29) to the sharp dark lines and pale gray profiles found in calcified sections of enamel (Figs. 4, 5 , 6, 7, 24, 27,30, 32). Since the staining solutions without heavy metals abolished electron opacity, the dense crystallite-like structures revealed by the stains must represent heavy metal binding to the organic matrix. Similarly, the opaque crystallite-like profiles seen after staining calcified sections (Figs. 12, 13, 14, 15, 31, 33) must also represent prior removal of crystallite opacity followed by binding of heavy metals to the organic matrix. Electron Probe Microanalysis High resolution electron probe microanalysis of calcified sections such as those shown in Figures 5, 24, and 27 revealed a peak for calcium in the enamel layer (Fig. 34). When such sections were treated with formic acid, electron opacity was removed (Figs. 16, 18) and only a minimal amount of calcium was detected with the probe (Fig. 34). Therefore, a loss of electron opacity can be correlated with a loss of calcium and both parameters must reflect dissolution of enamel crystallites. Selected Area Electron Diffraction Calcified sections of young inner enamel close to Tomes’ processes (Fig. 35a), similar to areas shown in Figures 5 and 24, were analyzed by selected area electron diffraction (Fig. 35b). The diffraction pattern confirmed the presence of crystalline material. Adjacent sections stained for a total of 15 minutes with aqueous uranyl and lead presented similar morphology (Fig. 36a), but selected area electron diffraction showed that no crystalline material was present (Fig. 36b). Solution-extraction removed the electron opacity from calcified sections (Fig. 37a), and the absence of crystalline structure was confirmed by electron diffraction (Fig. 37b). When solution-extracted sections were stained with urany1 and lead, electron opacity was restored (Fig. 38a), but such material failed to give a crystalline diffraction pattern (Fig. 38b). DISCUSSION It is commonly held that there are two constituents in enamel, the hydroxyapatite crystallite and the organic matrix. While the crystallites are intrinsically electron opaque and require no contrasting to be visualized, the matrix presumably behaves as other organic constituents and needs to be stained in order to be seen. Therefore, it should be possible to visualize both enamel constituents in undecalcified, but stained, sections. Previous attempts to achieve this simultaneous visualization were reviewed by Leblond and Warshawsky (1979). Two major views emerged. The first postulates that the matrix exists as a tubule, often visualized as hexagonal in cross-sectioned profile, with the interior containing the crystallite (Jessen, 1968; Travis and Glimcher, 1964; Travis, 1968; Decker, 1973). This idea was recently refined and it is now proposed that a coat of organic matrix is adsorbed to the outer surface of the enamel crystallite (Nylen, 1979; Yanagisawa and Nylen, 1980). The second view postulates that, a t least in young enamel, the matrix forms a “crystal ghost” similar to calcified cartilage (Bonnucci, 1969);that is, it occupies the same site as the crystallite (Ronnholm, 1962b; Frank, 1979; Nanci and Warshawsky, 1981; Warshawsky and Nanci, 1982). In order to explore the crystal-matrix relationship it seemed necessary to obtain a better understanding of the role played by the various contrasting agents normally used to prepare electron micrographs of the inorganic and organic constituents of enamel. Visualization of the Inorganic or Crystalline Constituents Although the inorganic crystallites can be seen with no contrast-enhancing treatments, it has been well documented for bone that mineral loss occurs from thin sections prepared for transmission electron microscopy (Boothroyd, 1964; Landis and Glimcher, 1978). Bishop and Warshawsky (1982) have reported that enamel crystallites are also susceptible to dissolution even by the water in the trough of the diamond knife. In addition, they showed that the electron diffraction pattern of young enamel is lost after staining with aqueous uranyl acetate and lead citrate. The present work confirmed the findings of Bishop and Warshawsky (1982) that young enamel crystallites are very susceptible to dissolution in distilled water and to aqueous staining solutions. Also, this work confirmed the well-known observation that crystallites are vulnerable to sublimation under electron beam irradiation (Boyde, 1979).The only role attributed to osmication of enamel is that crystallites are rendered less soluble in thin Figs. 24-29. Summary of enamel appearances under various conditions. The region of inner enamel secretion is at the top (Figs. 24-26, ~ 1 0 , 3 9 5 and ) outer enamel secretion is at the bottom (Figs. 27-29, ~ 4 7 , 2 5 0 )Arrows . point to electron density due to potassium ferrocyanide. R, Rod; ir, interrod prongs; T, Tomes’ process. Figs. 24, 27. In these calcified unstained sections, enamel crystallites are intrinsically electron opaque and appear as thin flat ribbons or as sharp needle-like densities. Figs. 25, 28. Solution-extractionresults in an abolition of electron opacity due to the loss of young enamel crystallites. In Figure 25 older crystallites (*) are only partially removed. Figs. 26, 29. When solution-extracted sections are conventionally restained, electron opaque images similar in size, orientation and distribution to those in calcified sections are now present. Since solution-extraction dissolves enamel crystallites and abolishes electron opacity these opaque crystallite-like structures must represent organic matrix. 12 A. NANCI, P. BAI, AND H. WARSHAWSKY Figs. 30-33. Comparison between osmicated enamel in a calcified unstained section (Figs. 30,32) and stained section (Figs. 31, 33) in the region of inner enamel secretion (Figs. 30, 31) and outer enamel secretion (Figs. 32, 33) very close to Tomes’ processes. Staining does not significantly alter the images seen in unstained preparations. Since staining solutions dissolve young enamel crystallites, the images in Figures 31 and 33 must represent the site of organic matrix. Since the pattern of the organic matrix is identical to the unstained crystallites, it is concluded that the matrix occupies the same site as the hydroxyapatite crystallites. T, Tomes’ process; R, rod; ir, interrod enamel. x 100,237. EFFECT OF HEAVY METAL STAINING ON ENAMEL 280 13 .. 200 -. 160 -. La--+---?- -i ---f--40 E m 100 120 140 160 180 200 220 240 260 CHRNNEL NO, I ZEl I Fig. 34. Energy dispersive X-ray spectra of young inner enamel, examined at a screen magnification of x 6,000 over areas on the sections of 20 X 25 pm (raster mode) and 300-second integrated detecting time periods. The recorded X-ray intensities (abscissa)were calibrated at 20 eV per channel. The upper trace was generated from osmicated, unstained sections not exposed to formic acid. The lower trace was obtained from similar sections decalcified on formic acid. The principal Cak,, peak was not detected in the formic acid decalcified enamel (arrow). The Cak, peak (normally occurring at - channel 201) was too small to be detected in either of these spectra. The X-ray peak positions of osmium and phosphorus (if present) overlap in the region of channel 100. Detectable sulfur is intrinsic to the tissue; chlorine derives from the support film over the grid. Other details of the microanalysis method have been published (Landis and Glimcher, 1982). sections and osmium reduces their vulnerability to electron beam damage. No alteration was noted in the actual appearance of the enamel with or without osmication. Although the ability of the enamel to bind toluidine blue and heavy metals was increased after osmication, uranyl and lead staining showed no major difference in crystallite appearance. Except for a granularity, no additional material was revealed in the stained sections when compared to the unstained sections, a n observation similar to that reported by Ronnholm (1962b). electron opacity formerly attributed to crystallites. High resolution electron probe microanalysis showed a loss of calcium from the enamel layer, thus confirming that formic acid had decalcified the thin section. Uranyl acetate and lead citrate staining of the formic acid-decalcified thin sections should reveal the organic constituent. Indeed, staining restored electron opaque images to the sections, but these were indistinguishable from the original crystallites. Since distilled water and formic acid can decalcify thin sections of enamel, the aqueous staining solutions themselves might also be capable of decalcifying enamel sections. Thus, utilizing staining solutions of the exact composition, but without uranyl and lead, also removed the electron dense crystallite images from thin sections and abolished the crystalline electron diffraction pattern. Urany1 and lead restaining of these extracted sections produced images indistinguishable from formic acid decalcified and stained enamel. From these experiments it was concluded Visualization of the Organic Matrix Constituents Staining of calcified sections of enamel did not reveal a n organic constituent that differed from the crystallites seen without staining. Therefore, in order to visualize the organic matrix alone, sections were decalcified with formic acid. Floating grid-mounted thin sections of calcified enamel for 1 hour on formic acid resulted in a complete loss of 14 A. NANCI, P. BAI, AND H. WARSHAWSKY Fig. 35. a) Area of inner enamel close to Tomes’ processes (T) selected for electron diffraction from a calcified, unstained section. x 11,700. b) Diffraction pattern from the selected area shown in (a). The concentric lines or maxima confirm the presence of crystalline material. The discontinuity in the concentric lines demonstrates a preferred crystallite orientation. Fig. 36. a) Area of inner enamel close to Tomes’ processes (T) from a calcified section stained with aqueous uranyl acetate and lead citrate for a total of 15 minutes. No major structural differences are seen when compared to Figure 35a. x 11,700. b) Diffraction pattern from the area shown in (a). The absence of lines or maxima indicates that no crystalline material is present. that routine staining cannot be used to re- Leblond and Warshawsky (1979). Yanagiveal the relationship between the crystallite sawa and Nylen (1980) favor the view that and the matrix, because these staining solu- “crystals in developing enamel bind organic tions decalcify the section leaving only the material to their surfaces, a n d . . . it is this stained organic matrix. It is thus not possible material which is visualized a s structural to visualize both constituents of young en- elements in decalcified sections of enamel.” The present work could not demonstrate a amel simultaneously on the same section. structured organic matrix that was separable from the crystallite images. No stained Crystallite-Matrix Relationship material was present between crystallite imThe question of how the hydroxyapatite ages and no coat of material was detected a t crystallite relates to the organic matrix has their surface. Thus, only a “ghost” of organic been recently discussed by Nylen (1979) and material within the crystallite image could EFFECT OF HEAVY METAL STAINING ON ENAMEL Fig. 37. a) Area of inner enamel close to Tomes’ processes (T)from a calcified section that was solutionextracted for 15 minutes. Note the absence of needle-like densities. x 11,700. b) The diffuse haloes in this diffraction pattern confirm the absence of crystalline material in the selected area shown in (a). 15 Fig. 38. a) Selected area of inner enamel close t o Tomes’ processes (T) from a calcified section that was solution-extracted and stained for 15 minutes with urany1 and lead. The needle-like densities have been restored. X 11,700. b) The diffuse diffraction pattern shows that the needle-like densities in (a) are not crystalline. be demonstrated. On theoretical grounds, an organic matrix septa. This organic material organic stroma within a ckystallite is dis- would not interfere with the visualization of puted because of space restrictions within crystal lattice fringes by high resolution electhe lattice of the crystal structure (Nylen and tron microscopy. Omnell, 1962). However, the crystalline structure could grow unimpeded on the surACKNOWLEDGMENTS face of an organic template. Supposing that such a template is a thin flat strip of matrix This work was supported by grants from capable of initiating or promoting crystallite the Medical Research Council of Canada. The growth, then crystalline hydroxyapatite authors are deeply indebted to Dr. William could grow on all surfaces. Thus, as previ- J. Landis, Harvard Medical School, for his ously proposed by Ronnholm (1962b), the collaboration with the electron probe and for young crystal would grow on both sides of a n providing Figure 34. 16 A. NANCI, P. BAI, AND H. WARSHAWSKY LITERATURE CITED Bernard, G.W. (1972) Ultrastructural observations of initial calcification in dentine and enamel. J. Ultrastruct. Res., 41:l-17. Bishop, M.A., and H. Warshawsky (1982)Electron microscopic studies on the potential loss of crystallites from routinely processed sections of young enamel in the rat incisor. Anat. Rec., 202t177-186. Bonnucci, E. (1969) Further investigation on the organic/ inorganic relationships in calcifying cartilage. Calc. Tissue Res., 3:38-54. Boothroyd, B. (1964) The problem of demineralization in thin sections of fully calcified bone. J. Cell Biol. 20: 165-173. Boyde, A. (1967) The development of enamel structure. Proc. R. SOC.Med., 6Ot923-928. Boyde, A. (1979) Carbonate concentration, crystal centers, core dissolution, caries, cross striations, circadian rhythms and compositional contrast in the SEM. J. Dent. Res., 58(‘B):981-983. Daculsi, G., and B. Kerebel (1978) High resolution electron microscope study of human enamel crystallites: Size, shape and growth. J. Ultrastruct. Res., 65163-172. Decker, J.D. (1973)Fixation effects on the fine structure of enamel crystal-matrix relationships. J. Ultrastruct. Res., 44:58-74. Fearnhead, R.W. (1961) Electron microscopy of forming enamel. Arch. Oral Biol.. 4:24-28. Frank, R.M. (1979) Tooth enamel: Current state of the art. J. Dent. Res., 58(B):684-693. Frank, R.M., and J. Nalbandian (1967)Ultrastructure of Amelogenesis. In: Structural and Chemical Organization of Teeth. A.E.W. Miles, ed. Academic Press, New York, Vol. 1,pp. 399-466. .Frazier, P.D. (1968) Adult human enamel: An electron microscopic study of crystallite size and morphology. J. Ultrastruct. Res., 22:l-11. Glimcher, M.J., E.J. Daniel, D.F. Travis, and S. Kamhi (1965) Electron optical and X-ray diffraction studies of the organization of the inorganic crystals in embryonic bovine enamel. J. Ultrastruct. Res., 7:1-77. Greep, R.O., C.J. Fischer, and A. Morse (1948) Alkaline phosphatase in odontogenesis and its histochemical demonstration after demineralization. J. Am. Dent. Assoc., 36:427-442. Jessen, J. (1968) Elliptical tubules as unit structure of forming enamel matrix in the rat. Arch. Oral Biol., 13r351-352. Kallenbach, E. (1973) The fine structure of Tomes’ process of rat incisor ameloblasts and its relationship to the elaboration of enamel. Tissue Cell, 5:501-524. Karnovsky, M.J. (1971) Use of ferrocyanide reduced osmium tetroxide in electron microscopy. Proc. 11th Am. SOC.Cell Biol., New Orleans, Louisiana. Abstract 284, 146. Landis, W.J., and M.J. Glimcher (1978) Electron diffraction and electron probe microanalysis of the mineral phase of bone tissue prepared by anhydrous techniques. J. Ultrastruct. Res., 63t188-223. Landis, W.J., and M.J. Glimcher (1982) Electron optical and analytical observations of rat growth plate cartilage prepared by ultracryomicrotomy: The failure to detect a mineral phase in matrix vesicles and the identification of heterodispersed particles as the initial solid phase of calcium phosphate deposited in the extracellular matrix. J. Ultrastruct. Res., 78227-268. Leblond, C.P., and H. Warshawsky (1979) Dynamics of enamel formation in the rat incisor tooth. J. Dent. ~ ~~~ Res., 58(B):950-975. Nalbandian, J., and R.M. Frank (1962)Microscopic electronique des gaines, des structures prismatiques et interprismatiques de l’email foetal humsin. Bull. Group. Int. Rech. Sc. Stomat., 5t523-542. Nanci, A., and H. Warshawsky (1981)Renioval of enamel crystallites by routine procedures such as staining with uranyl and lead. Microw. Soc. Can., 8:74-75. Nanci, A., and H. Warshawsky (1983) Relationship between the quality of fixation and the presence of stippled material in newly formed enamel of the rat incisor. Anat. Rec., in press. Nylen, M.U. (1979)Matrix-mineral relationships. A morphologist’s viewpoint. J. Dent. Res., 58(B):922-926. Nylen, M.U., E.D. Eanes, and K.-A. Omnell(1963) Crystal growth in rat enamel. J. Cell Biol., 18:109-123. Nylen, M.U. and K.-.&.Omnell (1962) The relationship between the apatite crystals and the organic matrix of rat enamel. In: Fifth International Congress for Electron Microscopy. Philadelphia S.S. Breese, ed., Academic Press, New York, Vol. 2, pp. QQ-4. Quigley, M.B. (1959) Electron microscopy of developing enamel matrix in the Syrian hamster. J. Dent. Res., 38t180-187. Reith, E.J. (1960) The ultrastructure of ameloblasts from the growing end ofrat incisors. Arch. Oral Biol., 2: 253-262. Reynolds, E.S. (1963) The use of lead citrate at high pH as an electron-opaque stain in electron microscopy. J. Cell Biol., 17:208-212. Ronnholm, E. (1962a) The amelogenesis of human teeth as revealed by electron microscopy. 11. The develop ment of the enamel crystallites. J. Ultrastruct. Res., 6:249-303. Ronnholm, E. (1962b)The amelogenesis of human teeth as revealed by electron microscopy. 111. The structure of the organic stroma of human enamel during amelogenesis. J. Ultrastruct Res., 6t368-389. Selvig, K.A., and A. Hake (1972) Crystal growth in rat incisor enamel. Anat. Rec., 173t453-468. Stack, M.V. (1967) Chemical organization of the organic matrix of enamel. In: Structural and Chemical Organization of Teeth. A.E.W. Miles, ed. Academic Press, New York, Vol. 2, pp. 317-346. Travis, D.F. (1968) Comparative ultrastructure and organization of inorganic crystals and organic matrices of mineralized tissues. In: Biology of the Mouth. P. Parson, ed. Am. Assoc. Adv. Sciences, Washington, D.C., pp. 237-297. Travis, D.F., and M.J. Glimcher (1964)The structure and organization of, and the relationship between the organic matrix and the inorganic crystals of embryonic bovine enamel. J. Cell Biol., 23:447-497. Warshawsky, H., K. Josephsen, A. Thylstrup, and 0. Fejerskov (1981)The development of enamel structure in rat incisors as compared to the teeth of monkey and man. Anat. Rec., 200:371-399. Warshawsky, H., and A. Nanci (1982) Effect of osmium, uranyl and lead staining on enamel crystallites. J. Dent. Res., 61[ZADR Abst.J;143:194. Watson, M.L. (1958) Staining of tissue sections for electron microscopy with heavy metals. J. Biophys. Biochem. Cytol., 4t475-478. Watson, M.L. (1960) The extracellular nature of enamel in the rat. J. Biophys. Biochem. Cytol., 7:489-492. Yanagisawa, T., and M.U. Nylen (1980)The relationship between matrix and mineral in rat enamel. J. Dent. Res., Special Issue A, AADR Abst., 59t303, 343.