ABSTRACT
The fine structure of the surface epithelium of developing palatine shelves in the mouse was studied from days 11 through 14 of gestation. Ruthenium red, a cationic stain used as an ultrastructural indicator of acid mucopolysaccharides, was employed to detect the presence of any surface coat.
Positive staining was first observed on day 12 of gestation and was seen to be present throughout the period of shelf elevation and fusion. Tt was seen over medial and lateral surfaces as well as the inferior to of vertical shelves. The surface coat was found to be present along the entire length of the shelf, extending superiorly up the medial and lateral epithelial borders until it abruptly disappeared.
Since this surface coat first appeared approximately 48 h prior to shelf elevation, it is suggested that its appearance may be associated with the ability of palatine shelves to undergo fusion as shown by previous in vitro experiments. The time of acquisition by the shelves of this ‘fusing potential’ is also in the range of 48 h before shelf elevation.
INTRODUCTION
The cellular components of the palatine shelves that participate in the formation of the secondary palate have been previously described in a variety of laboratory animals both at the light (Walker & Fraser, 1956; Coleman, 1965; Walker, 1971) and electron microscopic levels (Mato, Aikawa & Katahira, 1967b; DeAngelis & Nalbandian, 1968; Farbman, 1968; Smiley & Dixon, 1968; Smiley, 1970; Chaudhry & Shah, 1973). Adherence between epithelia of adjacent shelves is one of the first events that occurs after shelf elevation (Smiley, 1972). Attempts to pull shelves apart that have been in contact results in tearing of the epithelial borders, demonstrating that this adherence is indeed strong (Zeiler, Weinstein & Gibson, 1964; DeAngelis & Nalbandian, 1968; Farbman, 1968).
The basis for this tight adherence has been attributed both to desmosomes (DeAngelis & Nalbandian, 1968; Brusati, 1969; Chaudhry & Shah, 1973) and to an ‘ill-defined’ extracellular surface coat on the epithelial cells covering the palatine shelves (Hayward, 1969; Nanda & Kelly, 1973). The purpose of this study was to establish, with the help of specific stains available for electron microscopy, whether or not an extracellular surface coat is present on the epithelial cells throughout the period of shelf elevation and fusion. An attempt is also made to correlate this information with that obtained by Pourtois (1966) and Vargas (1967) regarding the timing of the so called ‘acquired potential’ to fuse exhibited by palatine shelves in vitro.
MATERIALS AND METHODS
Random-bred TCR/DUB mice (Flow Laboratories, Dublin, Virginia) fed Purina mouse chow and tap water ad libitum were used. Females were mated between 8 a.m. and 12 noon so that the time of fertilization of eggs could be closely estimated. The presence of a vaginal plug immediately afterwards was regarded as evidence of mating. Day 1 was considered to begin at 10 a.m. the following morning.
Animals were killed by cervical dislocation at various times between days 11 and 14 of gestation. Fetuses were removed to a dish containing the fixative and the heads sliced off below the mandible. Using fine dissecting knives, the mandible was then sliced away from the rest of the head and the tongue carefully removed.
The trimmed heads with exposed palatine shelves were fixed for three h at room temperature in 3 % glutaraldehyde, buffered with 0-1 M cacodylate buffer (pH 7-2), containing 0·05 % ruthenium red (K & K Laboratories, Inc., Plainview, N.Y.). They were then rinsed for 15 min in 0·1 M cacodylate buffer (pH 7·2) and post-fixed in similarly buffered 2 % osmium tetroxide containing 0·05 % ruthenium red, in the dark, for 3 h at room temperature (Luft, 1966). Fetal heads were then dehydrated through increasing concentrations of ethyl alcohol, beginning with 30 % and concluding with two changes of absolute alcohol. Dehydration was followed by two 10 min changes of propylene oxide. The dehydrated tissues were then gently agitated for 1 h in a 1:1 mixture of Araldite (Luft, 1961) and propylene oxide, followed by 24 h in a 3:1 Araldite and propylene oxide mixture, and then placed in pure Araldite (Luft, 1961) overnight. The tissue was embedded in Beem capsules kept at 60 °C for 24 h. Fetal heads were oriented during embedding in such a way as to facilitate coronal sectioning of the palatine shelves.
The blocks were trimmed on an LKB 11800 Pyramitome and sectioned with glass knives on an LKB-Huxley Mark 2 Ultramicrotome. One //m sections were stained with a 1 % toluidine blue solution (Trump, Smuckler & Bonditt, 1961) for tissue orientation and identification. Both thick (1 μm) and thin (60–80 nm) sections were taken from various regions of the secondary palate, ranging from the most anterior area to the most posterior. In appropriate regions, 30 μm high mesas were formed on the block face using the pyramitome. Thin sections were cut from these mesas with glass knives and transferred to pre-cleaned, Formvar and carbon-coated 75-mesh copper grids. Thin sections thus obtained were not stained, for maximal ruthenium red contrast, and examined on an RCA EMU-3H electron microscope operating at an accelerating voltage of 50 kV.
Other fetal heads were prepared as above with the omission of ruthenium red from the glutaraldehyde and osmium tetroxide solutions. Thin sections obtained from these blocks were stained with alcoholic uranyl acetate (Watson, 1958) and lead citrate (Reynolds, 1963) before examination in the electron microscope.
RESULTS
On day 11 of gestation the palatine processes are represented by a longitudinally oriented extension of the maxillary arch on either side of the tongue. Each process consists of loosely distributed mesenchymal tissue covered with a double-layered epithelium. At this stage the outer epithelial surface of the entire shelf remained unstained with ruthenium red (Fig. 1). Positive staining was first observed on day 12 of gestation. It was most evident over the inferior tip of the vertical shelf, extended over microvillous processes of the epithelial cells and for a short distance on both the medial and lateral surfaces of the shelf (Figs. 2–4). This surface coat was found to be present along the entire length of the shelf. At higher magnification the epithelial surface appeared to be covered with a layer of electron-dense granules matted together to form a more or less continuous coating (Figs. 5, 6).
Epithelial surface of the tip of a vertically oriented palatine shelf on day 11 of gestation. Note the absence of any continuous surface coating on the cell membrane. Ruthenium red en bloc staining, × 55800.
Palatal shelf epithelium on day 12 of gestation from an area corresponding to that shown in Fig. 1. Note positive staining at the cell membrane extending over microvillous processes of epithelial cells. Ruthenium red en bloc staining, ×38300.
Palatal shelf epithelium on day 12 of gestation from an area corresponding to that shown in Fig. 1. Note positive staining at the cell membrane extending over microvillous processes of epithelial cells. Ruthenium red en bloc staining, ×38300.
Epithelial surface of the medial border of a vertically oriented palatine shelf at 13 days 12 h of gestation. Ruthenium red en bloc staining, × 38300.
The lateral border of a vertically oriented palatal shelf at 13 days, 12 h of gestation. Ruthenium red en bloc staining, × 35200.
At intervals along the epithelial covering of the palatine shelves, surface cells of greater electron density were frequently seen (Fig. 7). These cells were generally found in the tip or medial margin of the vertical palatine shelves occupying the outer of the double-layered epithelial jacket around the shelves. This electron density was never seen in the deeper cuboidal cells of the palatine epithelium. It is questionable whether these cells represent an expression of the programmed cell death thought to occur in this region (Morgan & Harris, 1972; Smiley & Koch, 1972; Hudson & Shapiro, 1973), as they are found as early as day 12 of gestation and, when seen near the time of shelf elevation and fusion, demonstrate normal morphology except for their increased electron density. These cells also are not easily defined in terms of the morphology of physiological cell death (Schweichel & Merker, 1973).
Note a cell of greater electron density than adjacent and subjacent palatal epithelial cells. Ruthenium red en bloc staining, ×22000.
The surface coating found in the inferior region of vertical shelves extended superiorly up the medial and lateral epithelial borders for some distance and abruptly disappeared (Fig. 8). The sudden termination of the surface coating was more evident on the lateral than the medial epithelial surface. On the medial surface the coat merely became irregular and discontinuous as one progressed superiorly.
Lateral epithelial border of day 13 vertical palatal shelf (tip is towards the left of the micrograph). Note abrupt disappearance of surface coating (arrow). Ruthenium red en bloc staining, × 15000.
Sections not stained with ruthenium red also demonstrated some sort of surface substance on palatal epithelial cells. Figure 9 illustrates the epithelial borders of two adjacent palatine shelves just prior to contact. A thin epithelial microvillous process from one shelf is seen extending toward the opposite shelf. The epithelial surface of both shelves demonstrates a fine fuzz-like coating.
Epithelial borders of two adjacent palatine shelves just prior to contact. Note the thin epithelial microvillous process extending from one shelf toward the opposite shelf. Also note the fine coating (arrows) on both shelves. Uranyl acetate and lead citrate staining, × 72700.
DISCUSSION
After elevation and contact of the two developing palatine shelves, a tight adherence of one shelf to the other has been demonstrated (Zeiler et al. 1964; Farbman, 1968). This adherence is so strong that attempts to pull the shelves apart after initial contact produced artifactual cell tearing (Farbman, 1968; DeAngelis & Nalbandian, 1968).
The source of this firm adhesion remains unknown. The presence of desmosomes have been suggested by some to have the mechanical function of maintaining contact between the fusing epithelia of adjacent shelves until disruption of the epithelial seam and mesenchymal interpenetration occurs (DeAngelis & Nalbandian, 1968; Brusati, 1969; Chaudhry & Shah, 1973).
The occurrence of an extracellular surface layer, although controversial (Farbman, 1968; Matthiessen & Anderson, 1972), has been suggested (Mato et al. 1967 a; Hayward, 1969; Nanda & Kelly, 1973). Mato et al. (1967 a) have described what they term a ‘cell-reaction’, essential for fusion, that takes place at the contact sites between the nasal septum and the palatine shelves. Using the light microscope, they observe a thin dark cell surface layer on the surfaces which make contact. This layer could very likely be composed of the electron-dense surface cells described here (Fig. 7). Others have variously described an ‘ill-defined’ extracellular surface layer (Hayward, 1969) and the presence of electron-transparent areas (Nanda & Kelly, 1973) as perhaps playing a role in holding adjacent palatine shelves together. The nature of this surface coat, however, remains obscure.
Recently, the first positive evidence of a definitive surface coat on palatal shelf epithelium has emerged. Using concanavalin A, a carbohydrate-binding protein bound to horseradish peroxidase for electron microscopic study, Pratt, Gibson & Hassell (1973) have demonstrated reaction product at the surface of epithelia undergoing fusion. They concluded that a carbohydrate coat appears on the surface of palatal epithelial cells just prior to palatal fusion. Waterman, Ross & Meller (1973) utilizing the scanning electron microscope, have also exhibited the progressive accumulation of a filamentous material along the medial edge of shelves prior to contact.
Ruthenium red (ruthenium oxychloride) is a polyvalent cation which when used with osmium tetroxide results in an electron-opaque material easily observed ultrastructurally. Prior treatment with substances which bind to mucopolysaccharides (Martinez-Palomo, 1970) prevents ruthenium-red staining.
Although the mechanism by which this metallic dye acts can only be hypothesized (Luft, 1971 a), it has been reliably used as an ultrastructural indicator of acid mucopolysaccharides (Luft, 1971 b).
To confirm the presence of a surface coat, we have also attempted to detect aldehyde groups at the ultrastructural level by using metallic bismuth as a substitute for the Schiff reagent (Ainsworth, Ito & Karnovsky, 1972). This, we reasoned, would afford us another method to confirm the presence of a surface coat. This method, however, has not provided unequivocal results since they are difficult to reproduce.
The presence of a definitive carbohydrate surface coat may help to explain the phenomenon of an ‘acquired potential’ to fuse exhibited by palatine shelves of mice (Vargas, 1967) and rats (Pourtois, 1966) in vitro. Preparative changes have been reported to occur in the epithelium of organ-cultured palatal shelves prior to fusion (Morgan, 1969). Vargas (1967) has shown that palatal explants from early 12-day mouse fetuses did not fuse in vitro, while explants from older mouse fetuses fused normally. The potentiality to fuse in this species is acquired at least 40 h prior to actual closure of the secondary palate. Those explants that fused in vitro have been suggested as acquiring some properties in vivo which gave the explant the ability to fuse in vitro (Vargas, 1967). Since a carbohydrate surface coat first appeared on day 12 of gestation in our study, approximately 48 h prior to shelf elevation, it seems reasonable to speculate that it may be associated with the acquisition of potential to fuse which somehow does not occur if shelves are cultured from earlier embryos (Pourtois, 1972). Although fusion of palatal shelves in organ culture is not inhibited by the presence of proteolytic or saccharolytic enzymes (Pourtois, 1972), the time of acquisition by the shelves of a ‘potential to fuse’ in vitro has an intriguing parallel in the acquisition of a surface coat by the palatal epithelium.
Once the surface coat appeared, it was generally found at the inferior tip of vertically oriented shelves (Figs. 5, 6). A distinct lack of surface coat could be detected as one progressed superiorly on both the lateral and medial margins of vertically oriented shelves (Fig. 8). It is tempting to speculate with Waterman et al. (1973) that this inferior area might represent the future area of contact and fusion once the shelves have elevated.
The absence of surface coating on palatal epithelial cells may be of direct consequence to the production of cleft palate. Lack of normal adhesion between adjacent palatal shelves may prevent continued contact and result in separation of the shelves. Indeed, reopening of fused epithelial interfaces of palatal shelves has been suggested both in rodents (Buresh & Urban, 1967; Angelici, 1968; Greene & Kochhar, 1973) and man (Kitamura, 1966) as contributing to cleft palate.
ACKNOWLEDGEMENTS
This work was supported by NIH grant HD06550.