Electron microscopic studies on the potential loss of crystallites from routinely processed sections of young enamel in the rat incisor.

THE ANATOMICAL RECORD 202:177-186 (1982)
Electron Microscopic Studies on the Potential Loss
of Crystallites From Routinely Processed
Sections of Young Enamel in the
Rat lrrcisor
M.A. BISHOP A N D H. WARSHAWSKY
Department of Anatomy, McGill University, Montreal, Quebec,
Canada H3A 2B2
ABSTRACT
Newly formed rat incisor enamel was fixed aqueously by perfusion with xlutaraldehyde and anhydrously by immersion in ethylene glycol. Ultrathin sections were studied using transmission electron microscopy and electron
diffraction Aqueously processed enamel was shown to lose its mineral content
when sectioned on distilled water. This mineral loss was minimized by limiting the
exposure of sections to the water. In such preparations, enamel crystallites were
seen by virtue of their intrinsic electron density only. Selected area electron diffraction provided corroborative evidence for the presence or absence of crystallites in
the sections. Observations on mineralized sections and on stained mineralized and
distilled-water-demineralizedsections revealed organic material apparently in the
same location as the crystallites. Anhydrously processed enamel which was sectioned on ethylene glycol showed a similar appearance of the crystallites. This appearance was not obviously altered after staining despite evidence that organelles
in the ameloblasts were stained. In view of the observations that both methods
yielded similar crystallite morphology, it was concluded that aqueous techniques
can be used to study the relationship between organic and inorganic components.
However, valid description of crystallites in such preparations requires minimal
exposure of ultrathin sections to water.
The structural relationship between organic
and inorganic components of young tooth
enamel remains unclear (Nylen, 1979). The
problem is complicated by inadequate understanding of alterations produced by the techniques involved. For example, since the earliest
ultrastructural investigations on tooth enamel
and amelogenesis (Watson and Avery, 1954)
conventional aqueous methods of fixation and
staining have been used. I t has been suggested
more recently that these methods may cause
inadvertent demineralization of the enamel tissue (Watson, 1960; Decker, 1973; Glick and
Eisenmann, 1973). Therefore at least part of
the inorganic component may be lost during
processing and the validity of observations on
the organic-inorgayicrelationships can be questioned. The crystaUites of enamel possess intrinsic electron density. Thus, descriptions of
the inorganic elements do not require the use of
stains. Yet, little work has been performed on
young enamel without the interference of
heavy metal fixatives or stains such as osmium
tetroxide, uranyl acetate, and lead citrate.
Nylen (1979) and Leblond and Warshawsky
(1979) suggested that the organic matrix of
0003-276X/8212022-0177$03.00 0 1982 Alan R. Liss, Inc.
enamel may lie in close contact with the surface
of the crystallites. In addition, Frank (1979)
and Leblond and Warshawsky (1979)also suggested that matrix may lie directly within the
crystallites. Visualization of the organic matrix
in the electron microscope does require the use
of heavy metal stains. In view of the intimacy
of the proposed relationship it seems questionable whether it is possible to delimit the extent
of the organic matrix in the presence of the
crystallites. The use of heavy metals introduces electron density additional to the intrinsic electron density of the crystallites and it is
then difficult to determine where the crystallite
ends and the organic matrix begins.
This work attempts to explore the reliability
of several procedures in preserving both components of immature enamel. Using transmisReceived July 20. 1981; accepted August 12. 1981
M.A. Bishop's present address is Anatomy Department. Umversity College. Cardiff, CFI 1XL. Wales. U K.
Send reprint requests to H. Warshawsky. Department of Anatomy,
McGill University. 3640 University St.. Montreal. Quebec. H3A 2R2.
178
M.A. BISHOP AND H. WARSHAWSKY
sion electron microscopy coupled with electron
diffraction to provide more direct evidence for
the presence or absence of crystalline material,
it provides evidence for the loss of crystallites
from glutaraldehyde fixed, aqueously prepared
young enamel during sectioning on distilled
water. Furthermore, it demonstrates that extreme caution must be used when making interpretations from so-called “undemineralizedsections.”
MATERIALS AND METHODS
Aqueous preparation of tissue using
glutaraldehyde fixation
ide),and (4) stained according to (2)followed by
(3).
Anhydrous preparation of tissue using
ethylene glycol
The techniques used were based on those employed for bone by Landis et al. (1977). Male
rats weighing 80-115 gm were decapitated and
the right hemimandible was rapidly dissected
and placed in a dish of ethylene glycol on ice or
at room temperature (2-5 minutes after decapitation). The incisor was dissected from the
mandibular bone and the apical third was
sliced at right angles to its long axis using a
scalpel or a tissue sectioner (Sorvall).Segments
thus produced were placed in scintillation vials
containing ethylene glycol (at room temperature or on ice). The last segment reached a vial
within 15 minutes of decapitation. The vials
were placed in a desiccator on a shaker (at room
temperature or 4°C) and the desiccator was
evacuated continously for up to 24 hours. The
ethylene glycol was then replaced by Cellosolve
at atmospheric pressure and 4°C. The Cellosolve was changed after 12 hours. After a further 1 2 hours the Cellosolve was replaced with
a mixture of propylene oxide-Epon (1:l) at
room temperature. Thus the tissue remained
for 1 week in the desiccator, the latter being
shaken throughout processing. Finally the
specimens were embedded in Epon.
Regions of inner enamel in which two to
three rows of rod profiles would be present
were located. Block faces were trimmed to less
than 0.5 mm square and sections were cut with
a diamond knife using ethylene glycol as the
trough fluid. The chances of “wetting” the
block face during sectioning were minimized by
adding glycol to the trough after approaching
the block face with the diamond knife. When
wetting occurred the knife was removed,
washed with distilled water and dried and the
block face was wiped with 100% ethanol. Sections were often compressed but on expansion
(by gently warming the air above the sections)
they revealed interference colors of silver to
light gold. Sections were removed from the
trough using the slot grid method which produced fewer folds. Excess glycol was removed
with filter paper and the grids were placed under vacuum for several hours or overnight.
Sections were either unstained or stained according to each of the schedules described
above.
Male rats weighing approximately 100 gm
were anesthetized by intraperitoneal injection
of sodium pentabarbitone and ventilated via a
tracheal cannula. After perfusion with lactated
Ringer’s solution for about 1 minute through
the left ventricle of the heart, the rats were
fixed with a 2.5% solution of phosphatebuffered glutaraldehyde (pH 7.0) for 10-15
minutes. The right half of the mandible was divided into segments of about 1-mm thickness
by cutting transverse to the long axis of the
incisor tooth. The slices were made with a rotating carborundum disc in a straight dental
handpiece. Each segment was placed immediately in 30% acetone and the segments were
dehydrated in graded acetone and embedded in
Epon. After some trials it was decided to examine enamel in the region of inner enamel secretion, where two to three rows of rod profiles
were present in the sections (Fig. 1).Silver to
light gold sections were cut (Reichert Om U4
microtome) with a diamond knife, and floated
on a trough of distilled water (pH 6.9-7.5). The
sections were removed from the trough by two
methods. Either the grid was touched to the
sections floating on the distilled water in the
trough and excess water was removed with
filter paper, or a slot grid was used as a “loop”
to trap sections in the slot together with a film
of the trough water. This was then placed on
top of the supporting grid (held in forceps) and
the water was removed from between the two
grids with tissue paper. In the first method the
sections can be dried faster than in the second.
The time intervals between the moment of section cutting and dryness on the grid were varied and ranged from less than 1 minute to 135
minutes. Sections collected by either technique
were supported on 100-meshgrids coated with
Formvar (0.25%)and carbon.
Electron microscopy and diffraction
Sections were examined (1) unstained, (2)
stained anhydrously for 1 hour in 8% uranyl
The sections were examined in Philips 400
acetate in ethylene glycol (Landis et al., 1977), and 400T electron microscopes at 80 KV. Se(3)stained for up to 5 mkutes in aqueous 0.4% lected area electron diffraction was performed
lead citrate (dissolvedin 0.1 N sodium hydrox- on areas of approximately 5 pm diameter at the
179
LOSS OF ENAMEL CRYSTALLITES
specimen. The areas were selected using
100 pm and 150 pm diffraction apertures in the
E M 400 (objective magnification X 20) and
EM 400T (objective magnification X 30), respectively.
RESULTS
The effect of microtomy technique on the
appearance of unstained aqueously
prepared specimens
At an early stage in this work it became clear
that differences in the electron density of both
enamel and dentin could be produced by allowing sections to float on the distilled water in the
trough of the diamond knife for varying intervals. The unstained section in Figure 1was wet
for less than 2 minutes from the moment of sectioning to dryness on the grid while that in Figure 2 was wet for 110-135 minutes. The electron density in Figure 1 is caused by the
crystallites of dentin and enamel. In the
enamel, noncrystalline areas are present near
the dentinoenamel junction, in the middle of
the enamel, and at its forming surfaces where
Tomes’ processes of the ameloblasts interdigitate with enamel prongs. In Figure 2 both the
dentin and the enamel are electron lucent and
no crystallites are visible. Outlines of areas
equivalent to the noncrystalline regions can be
seen near the dentinoenamel junction and at
the forming surface of the enamel.
Figure 3 is an area of enamel selected from
the section in Figure 1. This area produced the
diffraction pattern shown in Figure 4. The distinct diffraction maxima in the pattern confirm
the presence of crystalline material in the
enamel.
The selected area shown in Figure 5 was
taken from the section in Figure 2 and produced the diffraction pattern shown in Figure
6. The absence of distinct diffraction maxima
indicates that no crystalline material was
present. The diffuse halos in Figure 6 are similar to those in Figure 7 and are produced by
carbon. The pattern in Figure 7 was obtained
from the unstained ameloblast cytoplasm.
Features o f enamel in stained aqueously
prepared specimens
The section shown in Figure 8 floated on distilled water for less than 2% minutes and was
stained anhydrously with uranyl acetate. In
the enamel, electron-dense elements resembling
the crystallites are present. Large areas near
the dentin contain lightly stained filamentous
or flocculent material which is similar in density to the material often seen between the
“crystal-like” elements (Fig. 9). Granules are
stained in Tomes’ processes at the forming
surface of the enamel (Fig. 8). An area selected
from the section in Figure 8 is shown in Figure
10 and gave the crystalline diffraction pattern
shown in Figure 11.
Figures 12 and 13 are taken from a section
that floated on distilled water for 117-119 minutes and was stained anhydrously with uranyl
acetate and with aqueous lead citrate. The
enamel consists of elongated “crystal-like”
structures. Some light stained flocculent material can be seen in the large spaces near the dentin and granules are stained within Tomes’
processes. An area selected from Figure 12 is
shown in Figure 14 and produced the noncrystalline diffraction pattern in Figure 15. A
section exposed to distilled water for 110-115
minutes and stained with anhydrous uranyl acetate only, produced micrographs similar to
those in Figures 12 and 13.
Features of enamel in unstained and
stained specimens prepared anhydrously
The unstained section in Figures 16 and 17
shows electron density produced by the crystallites within dentin and enamel. An area selected from this section is shown in Figure 18
and gave the crystalline diffraction pattern in
Figure 19.
The appearance of the enamel in sections
stained anhydrously with uranyl acetate was
essentially the same as that for the unstained
glycol preparation and the diffraction pattern
was crystalline.
DISCUSSION
In the present study enamel was processed
aqueously using glutaraldehyde as the sole
fixative. Since osmium tetroxide was not used,
no artificial electron density was added in the
preparation of the specimens. Thus the enamel
crystallites were demonstrated by virtue of
their intrinsic electron density alone. However,
it became obvious that specimens of young
enamel processed in this manner and sectioned
on distilled water can lose their crystalline content. Although electron diffraction showed
that crystal loss can be complete after 2 hours
on distilled water, significant loss was clearly
visible in the transmission mode as early as 15
minutes from the moment of sectioning. Minimal exposure to distilled water in the trough
retained the crystalline elements but it is not
clear whether the retention was complete. Evidence for the dissolution of certain synthetic
calcium phosphates subjected to standard electron microscopical procedures from fixation to
embedding (Termine, 1972) suggests that it
was not. Nevertheless, crystalline elements
were present in sections minimally exposed to
distilled water.
This study shows that there can be signifi-
180
M.A. BISHOP AND H. WARSHAWSKY
D, Dentin
En, Enamel
A bbreuiations
S, Noncrystalline area
TP, Tomes’ process
Figs. 1. 2. Unstained sections of glutaraldehyde-fixed
and aqueously processed enamel (En) and dentin (D) from
the same block. The sections were floated on distilled water
in the trough of the diamond knife. X6.450.
Fig. 1. This section was floated for less than 2 minutes.
Dentin (D) and enamel crystallites are seen as a result of
their intrinsic electron density alone.
Fig. 2. This section was cut adjacent to the one shown in
Figure 1. Both enamel (En) and dentin (D) are electron lucent, having lost their crystalline content as a result of long
exposure to distilled water (110-135 minutes).
LOSS OF ENAMEL CRYSTALLITES
181
Fig. 3. Area of enamel selected for diffraction from the
section in Figure 1. X12,OOO.
Fig. 5. Area of enamel selected for diffraction from section in Figure 2. X 12,000.
Fig. 4. Diffraction pattern from selected area shown in
Figure 3. The distinct concentric lines or maxima confirm
the presence of crystalline elements.
Fig. 6. Diffraction pattern from selected area in Figure
5. The absence of distinct lines or maxima indicates that no
crystalline material is present. The diffuse halos are produced by carbon.
182
M.A. BISHOP A N D H.WARSHAWSKY
Fig. 7. Diffraction pattern obtained from an area selected from the ameloblast cytoplasm of unstained
glutaraldehyde-fixed, aqueously processed tissue. Note the
similarity to the pattern given by the demineralized enamel
(Fig. 6).
cant mineral loss from aqueously prepared
young enamel if the sections are exposed to water for a protracted period of time. The problem
of mineral loss is well documented for bone
(Boothroyd, 1964; Thorogood and Gray, 1975;
Landis and Glimcher, 1978)but previously has
only been suggested for enamel (Watson, 1960;
Decker, 1973; Glick and Eisenmann, 1973).
Some attempts have been made to overcome
the problem by using buffers as flotation media
for the sections (Decker, 1973; Glick and Eisenmann, 1973). but no evidence of their effectiveness has been published. The use of electron
diffraction in the present work has provided objective assessment on the presence or absence
of crystals (Beeston, 1973).
Of the previous reports on the specific problem of the organic-inorganic relationships in
enamel (Scott and Nylen, 1962; Travis and
Glimcher, 1964; Decker, 1973; Nylen, 1979;
Fearnhead, 1979; Leblond and Warshawsky,
1979) only Decker (1973) mentioned any attempt to prevent demineralization during sectioning. There have been few micrographs published of enamel crystallites which did not include electron scattering, contrast-enhancing
elements (Nylen, et al, 1963).Thus, most previously published micrographs on this problem
contain the heavy metals osmium, uranium,
and/or lead. While some studies have deliber-
ately utilized grid decalcification (Travis and
Glimcher, 1964; Scott and Nylen, 1962; Nylen,
1979; Leblond and Warshawsky, 1979), claims
have been made that crystals were originally
present in sections without corroborative evidence such as electron diffraction (see Fig. l of
Travis and Glimcher, 1964 and Figs. 2 and 3 of
Decker, 1973).Evidence presented in this paper
shows that an appearance virtually identical to
that of crystallites (“crystal-like’’ structures,
Figs. 12-14) can be produced by adding heavy
metal elements to a considerably demineralized
section as shown by electron diffraction (Fig.
15). Thus, it is not acceptable to assume the
presence of crystallites in young enamel only
on the basis of their appearance in the transmission mode of the electron microscope. Corroborative evidence, as, for example, electron
diffraction, is essential.
The possibility of demineralization during
staining was minimized in the present work by
using uranyl acetate dissolved in ethylene glycol. The staining properties of the aqueously
prepared enamel were such that crystallites
and “crystal-like” structures showed affinities
for both uranyl acetate and lead citrate. The
latter observations imply that the crystallites
and/or the organic material intimately associated with the crystallites take up both stains.
Direct interpretation of these observations
leads to the conclusion that organic material
lies in the same location as the crystallites.
However, the more specific question of whether
the organic material lies on the surface of the
crystals, directly within the crystals, or in both
these locations has not been answered.
In view of the potential problem of demineralization, enamel was prepared anhydrously
using the ethylene glycol method, as described
for bone by Landis et al., (1977). The glycol
technique was thought to be particularly advantageous t o the present work because,
being almost totally anhydrous, it should prevent crystal dissolution. In addition, Pease
(1966a,b),who first described this method for
soft tissues, reported it to be a good preservative of protein. In the glycol-prepared enamel
the crystallites seemed well preserved. Staining did not change the appearance of the
enamel, yet the presence of identifiable organelles such as rough endoplasmic reticulum and
nuclei within the ameloblasts was evidence
that stain had been taken up by other organic
molecules in the section.
Routine aqueous procedures, including fixation with glutaraldehyde, seem to give images
similar to anhydrously prepared enamel using
183
LOSS OF ENAMEL CRYSTALLITES
Fig. 8.9.10. Sections of glutarddehyde-fixed.aqueouslY
processed tissue exposed to distilled water for less than 2%
minutes and stained anhydrously with uranyl acetate.
Fig. 8. Enamel (En)and dentin (D)show marked electron
density but otherwise resemble the unstained aqueously
prepared material in Figure 1. Flocculent materid is Present
in the large crystal-free areas (S) Of enamel. Granules in
Tomes' processes are stained (TP). X6,450.
pig. 9. ~~~~l
cryst&tes
in large crystd.free
a d fine flocculent material
(8are evident. ~22,800.
Fig. 10. Area selected for diffraction from section in Figure 8, x 12.000,
Fig. 1 1 . Diffraction pattern confirming the presence of
in selected area shown in Figure 10.
184
M.A. BISHOP AND H. WARSHAWSKY
Figs. 12-14. Section of glutaraldehyde-fixed. aqueously
processed tissue exposed to distilled water for 117-1 19 minutes. and stained anhydrously with uranyl acetate and
aqueous lead citrate.
Fie. 12. Enamel (En) and dentin (D) show electron dcn
sity resembling the mineralized tissues in Figures 1 and 8.
Flocculent material is present in large areas (S)near the denare stained. X6.480.
tin. Granules in Tomes’ processes (TP)
Fig. 13. The enamel shows “crystal-like” structures
within a flocculent “background” material. X25,OOO.
Fig. 14. Area selected for diffraction from the section in
Figure 12. X15.000.
Fig. 15. Diffraction pattern of the area in Figure 14
showing little trace of crystalline material. Thus, the
“crystal-like” structures are not crystalline.
LOSS OF ENAMEL CRYSTALLITES
Figs. 16-18. Unstained section of tissue processed anhydrously in ethylene glycol. The section was cut on a diamond knife and floated on ethylene glycol in the trough.
Fig. 16. Dentin (D) and enamel (En)crystallites are seen
due to their intrinsic electron density only X6.450.
185
Fig. 17. Higher magnification of the enamel crystallites
and part of a large crystal-free area (Sj. X22.800.
Fig. 18. Area selected for diffraction from the section in
Figure 16. X 12,000.
Fig. 19. Diffraction pattern showing the presence of
crystalline material in the selected area shown in Figure 18.
186
M.A. BISHOP AND H. WARSHAWSKY
ethylene glycol. In both procedures identical
enamel crystallites were preserved judging
from transmission mode electron microscopy
and electron diffraction. Thus, aqueous processing prior to embedding seems not to affect
crystallite preservation. On the other hand,
ultrathin sections containing these crystallites
seem vulnerable to extraction by distilled water. A consideration more important than
aqueous processing seems to be minimal exposure of ultrathin sections to aqueous solutions.
Since ethylene-glycol-processedenamel gave
a picture which was essentially identical to
aqueously processed enamel, particularly if
care was taken to minimize exposure to water,
it is concluded that crystallites are equally well
preserved by both methods. Because the glycol
method is more difficult to use it is suggested
that aqueously processed material may be adequate for morphological studies of the
inorganic-organic relationship. However, it is
quite clear that if aqueous methods are to be
used for examining crystallites in ultrathin sections great care must be taken to minimize
their exposure to water.
ACKNOWLEDGMENTS
This work was supported by a grant from the
Medical Research Council of Canada.
LITERATURE CITED
Beeston, B.E.P. (1973) An introduction to electron diffraction. In: Electron Diffraction and Optical Diffraction
Techniques. A.M. Glauert ed. North HollandiAmerican
Elsevier, New York.
Boothroyd, B. (1964) The problem of demineralization in
thin sections of fully calcified bone. J. Cell Biol., 20:
165-173.
Decker, J.D. (1973) Fixation effects on the fine structure of
~~~
~
~
enamel crystal-matrix relationships. J. Ultrastruc. Res.,
44: 58-74.
Fearnhead, R.W. (1979) Matrix-mineral relationships in
enamel tissues. J. Dent. Res., 58(B): 909-915.
Frank, R.M. (1979)Tooth Enamel: Current State of the Art.
J. Dent. Res., 58(B): 684-693.
Glick P.L., and D.R. Eisenmann (1973)Electron microscopic
and radiographic investigation of a morphologic basis for
the mineralization pattern in rat incisor enamel. Anat.
Rec., 176: 289-306.
Landis, W.J.. M.C. Paine. and M.J. Glimcher (1977) Electron microscopic observations of hone tissue prepared
anhydrously in organic solvents. J. Ultrastruct. Res., 59:
1-30.
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., 63: 188-223.
Leblond, C.P., and H. Warshawsky (1979) Dynamics of
enamel formation in the rat incisor tooth. J. Dent. Res.,
58(B):950-975.
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.R. Omnell (1963) Crystal
growth in rat enamel. J. Cell Biol. 18: 109-123.
Pease, D.C. (1966a) The preservation of unfixed cytological
detail by dehydration with “inert” agents. J. Ultrastruct.
Res., 14: 356-378.
Pease, D.C. (1966b) Anhydrous ultrathin sectioning and
staining for electron microscopy. J. Ultrastruc. Res.. 14:
379-390.
Scott, D.B., and M.U. Nylen (1962) Organic-inorganic interrelationships in enamel and dentin-A possible key to the
mechanism of caries. Int. Dent. J.. 12: 417-432.
Termine, J.D. (1972)Mineral chemistry and skeletal biolom.
Clin. Orthop. Rel. Res. 85: 207-241.
Thorogood, P.V., and J.C. Gray (1975) Demineralization of
bone matrix: Observations from electron microscope and
electron probe analysis. Calcif. Tissue Res., 19: 17-26.
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.
Watson, M.L., and J.K. Avery (1954) The development of
the hamster lower incisor as observed by electron microscopy. Am. J. Anat., 95: 109-161.
Watson. M.L. 11960) The extracellular nature of enamel in
the rat. J. Biophys. Biochem. Cytol., 7: 489-492.
~
~~