- Email: [email protected]
Desmosome structure in human foetal epithelia
Micron, 1972, 3:186-195
186
Desmosome structure in human foetal epithelia D. ADAMS and D. K. W H I T T A K E R Department of Oral Biology, Welsh National School of Medicine, Dental School, Heath, Cardiff CF4 4 X X , Wales, U.K.
Manuscript received September 17, 1971
The junctions between epithelial cells of skin and oral mucosa of human foetuses have been examined with the electron microscope. Densitometric tracings of desmosomes could be separated into three types based on the number of peaks between the attachment plaques. The most complex of these with three intermediate peaks corresponding to the classical desmosome comprised 20% of junctions in 10 week foetal skin and 90% in 22 week foetal skin. In the foetal lip at 10 weeks there were no desmosomes of this type and at 22 weeks only 16% of the junctions could be classified in this group. On the other hand the simple type with no intermediate peaks were found in greater numbers in oral epithelium than in skin. It is suggested that differences in desmosome structure are dependent on tissue differentiation and may precedefunctional differences. Les commissures entre les cellules gpith~liales de la peau et la muqueuse buccale de foetus humains ont gtg examinges par microscopie dlectronique. Des mesures densitomgtriques des desmosomes montr#ent que ceux-ci pouvaient ~tre sgparls en trois types distincts par le nombre de pics entre les plaques d' accolement. Les plus complexes de ceux-ci ont trois pics intermddiaires comprenaient 20% dejonctions daus la peau d'unfoetus de 10 semaines et 90% dans celle d'un foetus de 22 semaines. Dans les l~vres dufoetus de l Osemaines il n'y avait aucun desmosomede ce type et dans celui de 22 semaines seulement 16% desjonctions pouvaient ~tre classges dans cegroupe. D' autre part le cas le plus simple, ceux ne comportant pas de pics intermgdiairesfurent observgs en plus grand quantitg dans le tissu gpithdlial buccal que dans la peau. Une possibilitg que les diffdrences de structure desmosomiques soient dues aux diffgrences de constitution des tissus et ont gventuellement prdcgdence sur les diffgrences fonctionneUes est envisagge. Die Beriihrungsstellen zwischen Epithelzellen in der Haut und Mundschleimhaut menschlicher Feten wurden im Elektronenmikroskop untersucht. Anhand densitometrischer Aufzeichnungen konnten drei Arten yon Desmosomen unterschieden werden, und zwar auf Grund der Anzahl der Maxima zwischen den Ansatzplatten. Die klassischen Desmosomen sind die komplexesten und haben drie Kurvenmaxima. Sie bilden 20% der Beriihrungsstellen der HautzeUen in Feten yon 10 Wochen und 90% in Feten yon 20 Wochen. In der zehnten Woche befanden sich keine Desmosomen dieser Art in der fetalen Lippe und in der 22. Wochefielen nut 16% der Ber~hrungsstellen in diese Gruppe. Andererseits waren die einfachen Desmosomen ohne Zwischenmaxi m a i m Mundepithel zahlreicher als in der Haut. Es wird vermutet, dass die Unterschiede in der Desmosomenstruktur yon der Gewebsdifferenzierung abhdngen und funktionellen Unterschieden vorausgehen.
INTRODUCTION In 1954 Porter published the first description of the fine structure of the desmosome in vertebrate epidermis. The report was amplified in the following year by Selby (1955) who described the thickenings of areas of cell membranes in close contact with each other. This preliminary work was extended by Odland (1958) who used both electron micrographs and densitometric traces to characterise the desmosome in human epidermis. His description of the tonofilaments, attachment plaques, inter-
187
AI)AMS AND WHITTAKER
mediate dense layers and central intercellular contact layer has since been verified by many authors studying epidermis from various species both in health and disease (Farquhar and Palade, 1963; Kelly, 1966; Thilander and Bloom, 1968; Whitten, 1968; Whittaker and Adams, 1971). Since the original descriptions by Porter and Odland, Farquhar and Palade (1963) have classified junctions into three types, although Kelly and Luft (1966) suggested that there is a spectrum from the very simple to the very complex junctions. Desmosomes can thus exist in various degrees of complexity and the question arises as to whether their detailed structure is dependent upon the maturity of the desmosome, the maturity of the tissue in which they are found, upon other characteristics of the epidermal cells or upon combinations of these. The object of the following study was to investigate the desmosome in human foetal epithelium in an attempt to discover the relationships of the junctions to the maturity of the tissue and/or the properties of the epithelia. M A T E R I A L S AND M E T H O D S Specimens of scalp, lip and palate were excised from human foetuses ranging in age from 10 to 22 weeks as described previously (Adams and Whittaker, 1970; Whittaker and Adams, 1971). Normal adult human cheek mucosa and the epithelium surrounding a healing wound in hamster keratinising oral mucosa were obtained and all tissues were fixed immediately in 3% glutaraldehyde in phosphate buffer at 4°C (Karlsson and Schultz, 1965), post-fixed in osmium tetroxide and embedded in Araldite. The specimens were orientated so that ultra-thin sections could be cut at right angles to the epithelial surface, and sections were stained on the grid with uranyl acetate and lead citrate (Reynolds, 1963). The sections were a nominal 50-80nm in thickness. From these sections, desmosomes from the basal and spinous layers in each tissue were photographed and prints made at a standard magnification of × 200,000. Positive transparencies were made of each desmosome at the same magnification and these were scanned in a Joyce Loebl chromoscan at right angles to the desmosomal plaques and at two sites along the structure following a technique used by previous workers (Odland, 1958; Overton, 1962). Traces showing five major peaks were taken to represent the complex desmosome and it was possible to separate these from the traces with four or two peaks representing the simpler types. RESUI/I'S The desmosomes seen in the foetal epithelium could be grouped into three types. The most complex of these conformed to that described by Odland (1958) whose terminology is adopted here, and the densitometric tracing can be seen in Text Figure 1. The relationship of peaks to fine structure is illustrated diagrammatically. Types two and three did not possess the intercellular contact layer (ICL) but simply an amorphous or granular intercellular space. Type two had a well developed attachment plaque (AP) and tonofilaments associated with it whereas type three appeared to be an approximation of adjacent cell walls with thickening of cell membrane but no obvious intracellular specialisation (Text Figures 2 and 3). Since our grouping of the desmosomes was somewhat arbitrary an attempt was made to group them more
STRUCTURE OF DESMOSOMES
188
x '200,000 t
a = 1 6 . 0 nm
b=
6.5nm
c = 3 8 . 5 nm
X 400,000
d= a
8.0nm
:
c
x 800,000
IDL
Text Figure 1. Electron micrograph, diagram and densitometric tracing of complex desmosome from 22 week lip mucosa. The typical nine-layered structure can be seen in the micrograph and corresponds with the densitometric tracing. AP, attachment plaque; IDL, intermediate dense layer.
189
ADAMS AND WHITTAKER
a = 13.0nm
X 400,000
b=
5.7nm
c=
3.15 nm
C
x 800,000
Text Figure 2. Electron micrograph, diagram and densitometric tracing of second type of desmosome from 10 week skin epithelium. The intermediate dense layer is missing.
S T R U C T U R E OF DESMOSOMES
190
~ ~,~
x 400,000
a
I
a=15.0nm
1 '
! x 800,000
Text Figure 3. Electron micrograph, diagram and densitometric tracing of type three desmosome from 10 week lip mucosa. The simple nature of this junction is demonstrated.
191
ADAMS AND W H I T T A K E R
objectively by three independent observers segregating all the photomicrographs into three types using the above description of each type. The same observers then grouped the densitometric traces into three types on the basis of the number of peaks. In 3 cases out of 18 one observer differed from the others with respect to the photographs but in only 1 case out of 18 was there any variation from unanimity on the traces. Complex desmosomes made up the majority of junctions in foetal skin at 22 weeks comprising 90% of all those desmosomes studied. In contrast at 10 weeks in the lip i.e. in non-keratinising tissue, there were no desmosomes of the complex type whereas in 10 week skin 20% of desmosomes had the structure known as complex. The percentage increase in complex desmosomes seen in 22 week skin as compared to 10 week skin was not reflected in the non-keratinising epithelia since the increase went from 0% at 10 weeks to 17% at 22 weeks and eventually reached only 33% in the adult buccal mucosa. In the epithelium of the healing wound mature or complex desmosomes made up 20% of the total number of junctions examined. Our second type made up by far the majority of the remaining desmosomes and the very simple third type was only found in 10 week skin and 10 week lip, not being present either in the healing wound situation nor in any later foetal tissues examined. There did not appear to be any distribution pattern within any particular epithelium among the basal, spinous and granular layers. THOLE I
Percentageof Desmosome Types
10 week skin 22 week skin 10 week lip 22 week lip adult buccal mucosa healing wound
Type 1
Type 2
Type 3
20 90
70 10 50 68 67 80
10
0
16 33 20
0 50 16 0 0
The dimensions of the desmosomes are shown in the diagrams. These were measured on the densitometric tracings using a tracing of virus particles of known dimensions for calibration. The width of the space between cell membranes in type three was found to be 15nm which was of the same order as the distance between the intermediate dense layers (IDL) of adjacent cells in type two--13nm and type one--16nm. The distance between the middle parts of the density representing the attachment plaques was 38.5nm in the case of type three desmosomes compared with 31.5nm in the same measurement for type two desmosomes. DISCUSSION Glutaraldehyde fixed tissue stained on the grid with uranyl acetate and lead citrate was used since Komura and Ofuji (1967) showed that the desmosome as originally described by Odland was demonstrated most clearly in glutaraldehyde fixed and uranyl stained sections. This study confirms the reports of previous workers that
S T R U C T U R E OF DESMOSOMES
192
desmosomes can vary in structure (Kobayasi, 1966; DeAngelis and Nalbandian, 1968; Hoyes, 1968). The very simple type resembled the early developing desmosomes in reassociating cells of the chick blastoderm (Overton, 1962), whilst the intermediate type was similar to that seen in the forelimb bud of the chick (Jurand, 1965). Thilander and Bloom (1968) suggested that the appearance of desmosomes of the type which we have called type two may be due to the plane of section, producing apparent loss of the intercellular contact layer. In the present study enough desmosomes were examined to make this an unlikely explanation and since they were photographed at random in the various tissues, any defects in orientation or indeed in staining, would be common to all the grids examined. Sections of various thicknesses have been examined to find if the apparent structure was related to this characteristic. In no case was the grouping in doubt even when 120nm sections were studied. The use of photomicrographs and densitometric tracings in an attempt to provide a more objective assessment showed that examiner error was slightly greater in the interpretation of micrographs when compared with the interpretation of tracings. Our results, therefore, indicate that judgement depending on graphical representation is slightly more reliable than examination of electron micrographs alone. Kelly and Luft (1966) have described a range of cell junctions and state that they believe these to be part of a spectrum ranging from the very simple types which we have described here as existing in the 10 week old foetal material through other types to the most complex variety which we find in foetal skin at 22 weeks. They further state that these junctions are adapted to meet the varying functions that are put upon them although it is not clear whether they intend this to mean that individual desmosomes or junctions may change accordingly as their function changes. O u r results, which show differences in the percentages of the three types at a stage when functional differences are small, suggest that desmosome types appear to develop in advance of functional requirements. DeAngelis and Nalbandian (1968) describe 'young' desmosomes in the mouse and rat foetal palate, inferring that these structures 'mature' as the palate develops, and Breathnach (1971) demonstrates desmosomes similar to type three in six week foetal skin and shows two, one of which he believed to be at a more advanced stage of differentiation than the other which was similar to our type three. Modifications of desmosome structure have been previously described in relation to the epithelial layer with which they were associated. Horstmann and Knoop (1958) showed a loss of the I C L in the epithelium of the rat foot, whilst Listgarten (1964) found similar changes in the surface layers of oral epithelium and believed this to be a functional modification associated with eventual shedding of the cells. Thilander and Bloom (1968) stated that the exact structure of the desmosome in the surface layers was dependent upon the degree of keratinisation. In the present study most of the desmosomes examined were from the basal or spinous layers and we believe the difference to be due to more subtle factors than simple desquamation of the cells. The presence of complex desmosomes in keratinising tissues even in the basal layers suggests that in these tissues desmosomes are complex when formed and that this is not a property which develops as the cell moves to higher layers within the epithelium. O n the other hand, type two desmosomes which are found in great numbers in nonkeratinising epithelia do not appear in these particular tissues to achieve the com-
193
ADAMS A N D W H I T T A K E R
plexity of the first type. It appears therefore that the incidence of different desmosome types in foetal tissues anticipates the keratinisation potential of the tissue concerned. Nevertheless it must be borne in mind that there are differences other than keratinisation potential between the mucous membrane and the skin, e.g. some materials such as local anaesthetics penetrate much faster through the oral mucous membrane than skin, (Monash, 1957). Parmley and Seeds (1970) have shown that non-keratinised foetal skin at 14 weeks gave large permeability values similar to other non-keratinised foetal tissue and that these values were reduced when the skin became keratinised. Whilst this reduction may be due to a cornified layer on the skin it is possible that the structure of the desmosome might also be implicated in this change. The proportion of type one desmosome rises to 90% at 22 weeks which is close to the age at which Parmley and Seeds could get no detectable diffusion through the skin. Tight and septate junctions have been implicated in intercellular exchange by Farquhar and Palade (1965) and Satir and Gilula (1970) although this function has not, to our knowledge, been ascribed to desmosomes. Kelly and Luft (1966) believe that desmosome, tight and septate junctions belong to a continuous spectrum and thus it could be argued that desmosomes are capable of similar functions. It would appear that more work is required on this aspect of the nature of the desmosome. Since all epithelial cells probably possess a mucopolysaccharide coating (Rambourg and Leblond, 1967) the differences in desmosome structure are unlikely to be due simply to the presence or absence of extracellular coating material but may be due to variations in plasma membrane structure and/or mucopolysaccharide type (Frithiof, 1970; Sheffield, 1970; Kelly, 1971). Overton (1962) has suggested that desmosomes are permanent structures i.e., that the attachment sites remain throughout the life of the epithelial cell as it moves from the base to the surface. It is difficult to visualise how this could happen if the desmosome is a zone or area of attachment to other cells. It is much more likely that desmosomes are formed when the basal cell divides and that new attachments are formed between cells as they migrate towards the surface. We would suggest that if new desmosomes are formed then the differentiation of these desmosomes is a rapid process since we have found no evidence of 'maturation' as the successive layers from base to the surface are examined. The dimensions of the complex desmosome agree substantially with the figures of Odland (1958). He gave a figure of 20nm as the distance between the intermediate dense layers whereas our average measurement was 16nm. Odland found that the centres of attachment plaques were 30nm apart and our figure for the tracings studied was 38.5nm. Although Stern (1965) did not make this measurement it can be calculated from his figures to be 48-53nm. This is considerably greater than either Odland's figures and our own. Such a discrepancy could be caused by examining desmosomes with plaques of varying widths. Stern gave more detailed information than Odland for the dimensions of the layers within the desmosome and differences between Stern's figures and our own may be due to errors in measurement, to fixation shrinkage or change in dimension either in processing or under the electron beam. In spite of these reservations it can be seen that the width of the desmosome in type three is of the same order as the distances between intermediate dense lines. It thus appears
STRUCTURE OF DESMOSOMES
194
probable that the attachment plaque and the inner cell membrane has still to develop if these types do undergo differentiation. The close similarity of the intermediate dense layer dimensions in the first two types suggests that the intercellular contact layer m a y be a condensation of the intercellular material rather than a different structure interposed between the cells. In conclusion it is suggested that the relative numbers of complex and simple types of desmosome in stratified squamous epithelia are dependent firstly on the stage of development of the tissue and secondly on its keratinisation potential. The age of the individual desmosome does not appear to be of great importance in this respect since in healing wounds with presumed new desmosome formation the very simple types are not found. The possibility that permeability of the epithelium is related to the ratio of desmosome types requires further investigation. ACKNOWLEDGEMENTS O u r thanks are due to Professor B. E. D. Cooke in whose Department this work was carried out. We are also grateful to Mrs. D. Erasmus, Mr. R. Watkins, and Mrs. C. Winters for their technical assistance and to Miss C. A. Edwards for patiently typing the manuscript. REFERENCES ADAMS, D. and WHITTAKER,D. K., 1970. Observations on the periderm layer of human foetal oral mucosa, 07. Anat., 106:A411. BREATHNACH,A. S., 1971. An atlas of the ultrastructure of human skin. J. and A. Churchill, London. DEANoELIS, V. and NALBANDIAN,J., 1968. Ultrastructure of mouse and rat palatal processes prior to and during secondary palate formation, Arch. oral Biol., 13: 601-608. FARQUHAR, M. G. and PALADE, G. E., 1963. Junctional complexes in various epithelia, 07. Cell Biol., 17: 375-412. FARQUHAR, M. G. and PALADE,G. F., 1965. Cell junctions in amphibian skin. 07. Cell Biol., 26: 263-295. FRITHIOF, L., 1970. Ultrastructural changes in the plasma membrane in human oral epithelium. 07. Ultrastruct. Res., 32: 1-17. HORSTMANN,E. and KNOOP, A., 1958. Elektronenmikroskpische Studien an der Epidermis. 1. Rattenpfotte. Zellforsch. 47: 348-362. HoLEs, A. D., 1968. Electron microscopy of the surface layer (periderm) of human foetal skin. 07. Anat., 103: 321-336. JURAND, A. 1965. Ultrastructural aspects of early development of the forelimb buds in the chick and the mouse, Proc. roy. Soc., B., 162: 387-405. KARLSSON,U. and SGHULTZ,R., 1965. Fixation of the central nervous system for electron microscopy by aldehyde perfusion. 07. Ultrastruct. Res., 12: 160-186. KELLY, D. E., 1966. Fine structure of desmosomes, hemidesmosomes and an adepidermal globular layer in developing newt epidermis. 07. Cell Biol., 28: 51-72. KELLY, D. E. and LUFT, J. H., 1966. Fine structure, development and classification of desmosomes and related attachment mechanisms. In : Proc. 6th International Congress on Electron Microscopy, Kyoto, Ryozi Uyeda (ed.), Maruzen Co. Ltd., Tokyo,J apan, 2: 401-402. KOBAYASI, T., 1966. Development of fibrillar structures in human fetal skin. An electron microscope study. Acta. morph, dVeer. Scand., 6: 257-269. KO~URA, J. and OF~jI, S., 1967. Effect of electron stains and desmosome: an electron microscopic study. 07. Invest. Derm., 48: 304-308. LISTGARTEN, M. A., 1964. The ultrastructure of human gingival epithelium. Am. a7. Anat., 114: 49-69. MONASH, S., 1957. Location of the superficial epithelial barrier to skin penetration, 07. Invest. Derm., 29: 367-376.
195
ADAMS AND WHITTAKER
ODLAND,G. F., 1958. The fine structure of the interrelationship of cells in the human epidermis. ft. biophys, biochem. Cytol., 4: 529-538. OVERTON, J., 1962. Desmosome development in normal and reassociating cells in the early chick blastoderm. Devd. Biol., 4: 532-548. PARMLEY,T. H. and SEEDS, A. E., 1970. Fetal skin permeability to isotopic water ( T H O ) in early pregnancy. Am. 07. Obstet. Gynec., 108: 128-131. PORTER, K., 1954. Observations on the submicroscopic structure of animal epidermis. Anat. Rec., 118: A433. RA~OURG, A. and LEBLOND, C. P., 1967. Electron microscope observations on the carboydrate-rich cell coat present at the surface of cells in the rat. 07. Cell Biol., 32: 27-53. REYNOLDS, E. S., 1963. The use of lead citrate at high pH as an electron opaque stain in electron microscopy. 07. Cell Biol., 17: 208-212. SATIR, P. and GILULA, N. B., 1970. The cell junction in a lameUibranch gill ciliated epithelium. o7. Cell Biol., 4/: 468-487. SELBV, C. C., 1955. An electron microscope study of the epidermis of mammalian skin in thin sections. ,7. biophys, biochem. Cytol., 1 : 429-444. SHEFFIELD,J. B., 1970. Studies on aggregation of embryonic cells: Initial celt adhesions and the formation of intercellular junctions. 07. Morph., 132: 245-264. STERN, I. B., 1965. Electron microscopic observations of oral epithelium. 1. Basal cells and the basement membrane. Periodontics, 3: 224-238. TmLANDE~, H. and BLOOM, G. D., 1968. Cell contacts in oral epithelia. 07. periodont. Res., 3: 96-110. WmTTArmR, D. K. and ADAMS, D., 1971. The surface layer of human foetal skin and oral mucosa: A study by scanning and transmission electron microscopy. 07. Anat., 108: 453-464. WnIa~rEN, J. B., 1968. The fine structure of desquamative stomatitis. 07. Periodont., 39: 75-80.
186
Desmosome structure in human foetal epithelia D. ADAMS and D. K. W H I T T A K E R Department of Oral Biology, Welsh National School of Medicine, Dental School, Heath, Cardiff CF4 4 X X , Wales, U.K.
Manuscript received September 17, 1971
The junctions between epithelial cells of skin and oral mucosa of human foetuses have been examined with the electron microscope. Densitometric tracings of desmosomes could be separated into three types based on the number of peaks between the attachment plaques. The most complex of these with three intermediate peaks corresponding to the classical desmosome comprised 20% of junctions in 10 week foetal skin and 90% in 22 week foetal skin. In the foetal lip at 10 weeks there were no desmosomes of this type and at 22 weeks only 16% of the junctions could be classified in this group. On the other hand the simple type with no intermediate peaks were found in greater numbers in oral epithelium than in skin. It is suggested that differences in desmosome structure are dependent on tissue differentiation and may precedefunctional differences. Les commissures entre les cellules gpith~liales de la peau et la muqueuse buccale de foetus humains ont gtg examinges par microscopie dlectronique. Des mesures densitomgtriques des desmosomes montr#ent que ceux-ci pouvaient ~tre sgparls en trois types distincts par le nombre de pics entre les plaques d' accolement. Les plus complexes de ceux-ci ont trois pics intermddiaires comprenaient 20% dejonctions daus la peau d'unfoetus de 10 semaines et 90% dans celle d'un foetus de 22 semaines. Dans les l~vres dufoetus de l Osemaines il n'y avait aucun desmosomede ce type et dans celui de 22 semaines seulement 16% desjonctions pouvaient ~tre classges dans cegroupe. D' autre part le cas le plus simple, ceux ne comportant pas de pics intermgdiairesfurent observgs en plus grand quantitg dans le tissu gpithdlial buccal que dans la peau. Une possibilitg que les diffdrences de structure desmosomiques soient dues aux diffgrences de constitution des tissus et ont gventuellement prdcgdence sur les diffgrences fonctionneUes est envisagge. Die Beriihrungsstellen zwischen Epithelzellen in der Haut und Mundschleimhaut menschlicher Feten wurden im Elektronenmikroskop untersucht. Anhand densitometrischer Aufzeichnungen konnten drei Arten yon Desmosomen unterschieden werden, und zwar auf Grund der Anzahl der Maxima zwischen den Ansatzplatten. Die klassischen Desmosomen sind die komplexesten und haben drie Kurvenmaxima. Sie bilden 20% der Beriihrungsstellen der HautzeUen in Feten yon 10 Wochen und 90% in Feten yon 20 Wochen. In der zehnten Woche befanden sich keine Desmosomen dieser Art in der fetalen Lippe und in der 22. Wochefielen nut 16% der Ber~hrungsstellen in diese Gruppe. Andererseits waren die einfachen Desmosomen ohne Zwischenmaxi m a i m Mundepithel zahlreicher als in der Haut. Es wird vermutet, dass die Unterschiede in der Desmosomenstruktur yon der Gewebsdifferenzierung abhdngen und funktionellen Unterschieden vorausgehen.
INTRODUCTION In 1954 Porter published the first description of the fine structure of the desmosome in vertebrate epidermis. The report was amplified in the following year by Selby (1955) who described the thickenings of areas of cell membranes in close contact with each other. This preliminary work was extended by Odland (1958) who used both electron micrographs and densitometric traces to characterise the desmosome in human epidermis. His description of the tonofilaments, attachment plaques, inter-
187
AI)AMS AND WHITTAKER
mediate dense layers and central intercellular contact layer has since been verified by many authors studying epidermis from various species both in health and disease (Farquhar and Palade, 1963; Kelly, 1966; Thilander and Bloom, 1968; Whitten, 1968; Whittaker and Adams, 1971). Since the original descriptions by Porter and Odland, Farquhar and Palade (1963) have classified junctions into three types, although Kelly and Luft (1966) suggested that there is a spectrum from the very simple to the very complex junctions. Desmosomes can thus exist in various degrees of complexity and the question arises as to whether their detailed structure is dependent upon the maturity of the desmosome, the maturity of the tissue in which they are found, upon other characteristics of the epidermal cells or upon combinations of these. The object of the following study was to investigate the desmosome in human foetal epithelium in an attempt to discover the relationships of the junctions to the maturity of the tissue and/or the properties of the epithelia. M A T E R I A L S AND M E T H O D S Specimens of scalp, lip and palate were excised from human foetuses ranging in age from 10 to 22 weeks as described previously (Adams and Whittaker, 1970; Whittaker and Adams, 1971). Normal adult human cheek mucosa and the epithelium surrounding a healing wound in hamster keratinising oral mucosa were obtained and all tissues were fixed immediately in 3% glutaraldehyde in phosphate buffer at 4°C (Karlsson and Schultz, 1965), post-fixed in osmium tetroxide and embedded in Araldite. The specimens were orientated so that ultra-thin sections could be cut at right angles to the epithelial surface, and sections were stained on the grid with uranyl acetate and lead citrate (Reynolds, 1963). The sections were a nominal 50-80nm in thickness. From these sections, desmosomes from the basal and spinous layers in each tissue were photographed and prints made at a standard magnification of × 200,000. Positive transparencies were made of each desmosome at the same magnification and these were scanned in a Joyce Loebl chromoscan at right angles to the desmosomal plaques and at two sites along the structure following a technique used by previous workers (Odland, 1958; Overton, 1962). Traces showing five major peaks were taken to represent the complex desmosome and it was possible to separate these from the traces with four or two peaks representing the simpler types. RESUI/I'S The desmosomes seen in the foetal epithelium could be grouped into three types. The most complex of these conformed to that described by Odland (1958) whose terminology is adopted here, and the densitometric tracing can be seen in Text Figure 1. The relationship of peaks to fine structure is illustrated diagrammatically. Types two and three did not possess the intercellular contact layer (ICL) but simply an amorphous or granular intercellular space. Type two had a well developed attachment plaque (AP) and tonofilaments associated with it whereas type three appeared to be an approximation of adjacent cell walls with thickening of cell membrane but no obvious intracellular specialisation (Text Figures 2 and 3). Since our grouping of the desmosomes was somewhat arbitrary an attempt was made to group them more
STRUCTURE OF DESMOSOMES
188
x '200,000 t
a = 1 6 . 0 nm
b=
6.5nm
c = 3 8 . 5 nm
X 400,000
d= a
8.0nm
:
c
x 800,000
IDL
Text Figure 1. Electron micrograph, diagram and densitometric tracing of complex desmosome from 22 week lip mucosa. The typical nine-layered structure can be seen in the micrograph and corresponds with the densitometric tracing. AP, attachment plaque; IDL, intermediate dense layer.
189
ADAMS AND WHITTAKER
a = 13.0nm
X 400,000
b=
5.7nm
c=
3.15 nm
C
x 800,000
Text Figure 2. Electron micrograph, diagram and densitometric tracing of second type of desmosome from 10 week skin epithelium. The intermediate dense layer is missing.
S T R U C T U R E OF DESMOSOMES
190
~ ~,~
x 400,000
a
I
a=15.0nm
1 '
! x 800,000
Text Figure 3. Electron micrograph, diagram and densitometric tracing of type three desmosome from 10 week lip mucosa. The simple nature of this junction is demonstrated.
191
ADAMS AND W H I T T A K E R
objectively by three independent observers segregating all the photomicrographs into three types using the above description of each type. The same observers then grouped the densitometric traces into three types on the basis of the number of peaks. In 3 cases out of 18 one observer differed from the others with respect to the photographs but in only 1 case out of 18 was there any variation from unanimity on the traces. Complex desmosomes made up the majority of junctions in foetal skin at 22 weeks comprising 90% of all those desmosomes studied. In contrast at 10 weeks in the lip i.e. in non-keratinising tissue, there were no desmosomes of the complex type whereas in 10 week skin 20% of desmosomes had the structure known as complex. The percentage increase in complex desmosomes seen in 22 week skin as compared to 10 week skin was not reflected in the non-keratinising epithelia since the increase went from 0% at 10 weeks to 17% at 22 weeks and eventually reached only 33% in the adult buccal mucosa. In the epithelium of the healing wound mature or complex desmosomes made up 20% of the total number of junctions examined. Our second type made up by far the majority of the remaining desmosomes and the very simple third type was only found in 10 week skin and 10 week lip, not being present either in the healing wound situation nor in any later foetal tissues examined. There did not appear to be any distribution pattern within any particular epithelium among the basal, spinous and granular layers. THOLE I
Percentageof Desmosome Types
10 week skin 22 week skin 10 week lip 22 week lip adult buccal mucosa healing wound
Type 1
Type 2
Type 3
20 90
70 10 50 68 67 80
10
0
16 33 20
0 50 16 0 0
The dimensions of the desmosomes are shown in the diagrams. These were measured on the densitometric tracings using a tracing of virus particles of known dimensions for calibration. The width of the space between cell membranes in type three was found to be 15nm which was of the same order as the distance between the intermediate dense layers (IDL) of adjacent cells in type two--13nm and type one--16nm. The distance between the middle parts of the density representing the attachment plaques was 38.5nm in the case of type three desmosomes compared with 31.5nm in the same measurement for type two desmosomes. DISCUSSION Glutaraldehyde fixed tissue stained on the grid with uranyl acetate and lead citrate was used since Komura and Ofuji (1967) showed that the desmosome as originally described by Odland was demonstrated most clearly in glutaraldehyde fixed and uranyl stained sections. This study confirms the reports of previous workers that
S T R U C T U R E OF DESMOSOMES
192
desmosomes can vary in structure (Kobayasi, 1966; DeAngelis and Nalbandian, 1968; Hoyes, 1968). The very simple type resembled the early developing desmosomes in reassociating cells of the chick blastoderm (Overton, 1962), whilst the intermediate type was similar to that seen in the forelimb bud of the chick (Jurand, 1965). Thilander and Bloom (1968) suggested that the appearance of desmosomes of the type which we have called type two may be due to the plane of section, producing apparent loss of the intercellular contact layer. In the present study enough desmosomes were examined to make this an unlikely explanation and since they were photographed at random in the various tissues, any defects in orientation or indeed in staining, would be common to all the grids examined. Sections of various thicknesses have been examined to find if the apparent structure was related to this characteristic. In no case was the grouping in doubt even when 120nm sections were studied. The use of photomicrographs and densitometric tracings in an attempt to provide a more objective assessment showed that examiner error was slightly greater in the interpretation of micrographs when compared with the interpretation of tracings. Our results, therefore, indicate that judgement depending on graphical representation is slightly more reliable than examination of electron micrographs alone. Kelly and Luft (1966) have described a range of cell junctions and state that they believe these to be part of a spectrum ranging from the very simple types which we have described here as existing in the 10 week old foetal material through other types to the most complex variety which we find in foetal skin at 22 weeks. They further state that these junctions are adapted to meet the varying functions that are put upon them although it is not clear whether they intend this to mean that individual desmosomes or junctions may change accordingly as their function changes. O u r results, which show differences in the percentages of the three types at a stage when functional differences are small, suggest that desmosome types appear to develop in advance of functional requirements. DeAngelis and Nalbandian (1968) describe 'young' desmosomes in the mouse and rat foetal palate, inferring that these structures 'mature' as the palate develops, and Breathnach (1971) demonstrates desmosomes similar to type three in six week foetal skin and shows two, one of which he believed to be at a more advanced stage of differentiation than the other which was similar to our type three. Modifications of desmosome structure have been previously described in relation to the epithelial layer with which they were associated. Horstmann and Knoop (1958) showed a loss of the I C L in the epithelium of the rat foot, whilst Listgarten (1964) found similar changes in the surface layers of oral epithelium and believed this to be a functional modification associated with eventual shedding of the cells. Thilander and Bloom (1968) stated that the exact structure of the desmosome in the surface layers was dependent upon the degree of keratinisation. In the present study most of the desmosomes examined were from the basal or spinous layers and we believe the difference to be due to more subtle factors than simple desquamation of the cells. The presence of complex desmosomes in keratinising tissues even in the basal layers suggests that in these tissues desmosomes are complex when formed and that this is not a property which develops as the cell moves to higher layers within the epithelium. O n the other hand, type two desmosomes which are found in great numbers in nonkeratinising epithelia do not appear in these particular tissues to achieve the com-
193
ADAMS A N D W H I T T A K E R
plexity of the first type. It appears therefore that the incidence of different desmosome types in foetal tissues anticipates the keratinisation potential of the tissue concerned. Nevertheless it must be borne in mind that there are differences other than keratinisation potential between the mucous membrane and the skin, e.g. some materials such as local anaesthetics penetrate much faster through the oral mucous membrane than skin, (Monash, 1957). Parmley and Seeds (1970) have shown that non-keratinised foetal skin at 14 weeks gave large permeability values similar to other non-keratinised foetal tissue and that these values were reduced when the skin became keratinised. Whilst this reduction may be due to a cornified layer on the skin it is possible that the structure of the desmosome might also be implicated in this change. The proportion of type one desmosome rises to 90% at 22 weeks which is close to the age at which Parmley and Seeds could get no detectable diffusion through the skin. Tight and septate junctions have been implicated in intercellular exchange by Farquhar and Palade (1965) and Satir and Gilula (1970) although this function has not, to our knowledge, been ascribed to desmosomes. Kelly and Luft (1966) believe that desmosome, tight and septate junctions belong to a continuous spectrum and thus it could be argued that desmosomes are capable of similar functions. It would appear that more work is required on this aspect of the nature of the desmosome. Since all epithelial cells probably possess a mucopolysaccharide coating (Rambourg and Leblond, 1967) the differences in desmosome structure are unlikely to be due simply to the presence or absence of extracellular coating material but may be due to variations in plasma membrane structure and/or mucopolysaccharide type (Frithiof, 1970; Sheffield, 1970; Kelly, 1971). Overton (1962) has suggested that desmosomes are permanent structures i.e., that the attachment sites remain throughout the life of the epithelial cell as it moves from the base to the surface. It is difficult to visualise how this could happen if the desmosome is a zone or area of attachment to other cells. It is much more likely that desmosomes are formed when the basal cell divides and that new attachments are formed between cells as they migrate towards the surface. We would suggest that if new desmosomes are formed then the differentiation of these desmosomes is a rapid process since we have found no evidence of 'maturation' as the successive layers from base to the surface are examined. The dimensions of the complex desmosome agree substantially with the figures of Odland (1958). He gave a figure of 20nm as the distance between the intermediate dense layers whereas our average measurement was 16nm. Odland found that the centres of attachment plaques were 30nm apart and our figure for the tracings studied was 38.5nm. Although Stern (1965) did not make this measurement it can be calculated from his figures to be 48-53nm. This is considerably greater than either Odland's figures and our own. Such a discrepancy could be caused by examining desmosomes with plaques of varying widths. Stern gave more detailed information than Odland for the dimensions of the layers within the desmosome and differences between Stern's figures and our own may be due to errors in measurement, to fixation shrinkage or change in dimension either in processing or under the electron beam. In spite of these reservations it can be seen that the width of the desmosome in type three is of the same order as the distances between intermediate dense lines. It thus appears
STRUCTURE OF DESMOSOMES
194
probable that the attachment plaque and the inner cell membrane has still to develop if these types do undergo differentiation. The close similarity of the intermediate dense layer dimensions in the first two types suggests that the intercellular contact layer m a y be a condensation of the intercellular material rather than a different structure interposed between the cells. In conclusion it is suggested that the relative numbers of complex and simple types of desmosome in stratified squamous epithelia are dependent firstly on the stage of development of the tissue and secondly on its keratinisation potential. The age of the individual desmosome does not appear to be of great importance in this respect since in healing wounds with presumed new desmosome formation the very simple types are not found. The possibility that permeability of the epithelium is related to the ratio of desmosome types requires further investigation. ACKNOWLEDGEMENTS O u r thanks are due to Professor B. E. D. Cooke in whose Department this work was carried out. We are also grateful to Mrs. D. Erasmus, Mr. R. Watkins, and Mrs. C. Winters for their technical assistance and to Miss C. A. Edwards for patiently typing the manuscript. REFERENCES ADAMS, D. and WHITTAKER,D. K., 1970. Observations on the periderm layer of human foetal oral mucosa, 07. Anat., 106:A411. BREATHNACH,A. S., 1971. An atlas of the ultrastructure of human skin. J. and A. Churchill, London. DEANoELIS, V. and NALBANDIAN,J., 1968. Ultrastructure of mouse and rat palatal processes prior to and during secondary palate formation, Arch. oral Biol., 13: 601-608. FARQUHAR, M. G. and PALADE, G. E., 1963. Junctional complexes in various epithelia, 07. Cell Biol., 17: 375-412. FARQUHAR, M. G. and PALADE,G. F., 1965. Cell junctions in amphibian skin. 07. Cell Biol., 26: 263-295. FRITHIOF, L., 1970. Ultrastructural changes in the plasma membrane in human oral epithelium. 07. Ultrastruct. Res., 32: 1-17. HORSTMANN,E. and KNOOP, A., 1958. Elektronenmikroskpische Studien an der Epidermis. 1. Rattenpfotte. Zellforsch. 47: 348-362. HoLEs, A. D., 1968. Electron microscopy of the surface layer (periderm) of human foetal skin. 07. Anat., 103: 321-336. JURAND, A. 1965. Ultrastructural aspects of early development of the forelimb buds in the chick and the mouse, Proc. roy. Soc., B., 162: 387-405. KARLSSON,U. and SGHULTZ,R., 1965. Fixation of the central nervous system for electron microscopy by aldehyde perfusion. 07. Ultrastruct. Res., 12: 160-186. KELLY, D. E., 1966. Fine structure of desmosomes, hemidesmosomes and an adepidermal globular layer in developing newt epidermis. 07. Cell Biol., 28: 51-72. KELLY, D. E. and LUFT, J. H., 1966. Fine structure, development and classification of desmosomes and related attachment mechanisms. In : Proc. 6th International Congress on Electron Microscopy, Kyoto, Ryozi Uyeda (ed.), Maruzen Co. Ltd., Tokyo,J apan, 2: 401-402. KOBAYASI, T., 1966. Development of fibrillar structures in human fetal skin. An electron microscope study. Acta. morph, dVeer. Scand., 6: 257-269. KO~URA, J. and OF~jI, S., 1967. Effect of electron stains and desmosome: an electron microscopic study. 07. Invest. Derm., 48: 304-308. LISTGARTEN, M. A., 1964. The ultrastructure of human gingival epithelium. Am. a7. Anat., 114: 49-69. MONASH, S., 1957. Location of the superficial epithelial barrier to skin penetration, 07. Invest. Derm., 29: 367-376.
195
ADAMS AND WHITTAKER
ODLAND,G. F., 1958. The fine structure of the interrelationship of cells in the human epidermis. ft. biophys, biochem. Cytol., 4: 529-538. OVERTON, J., 1962. Desmosome development in normal and reassociating cells in the early chick blastoderm. Devd. Biol., 4: 532-548. PARMLEY,T. H. and SEEDS, A. E., 1970. Fetal skin permeability to isotopic water ( T H O ) in early pregnancy. Am. 07. Obstet. Gynec., 108: 128-131. PORTER, K., 1954. Observations on the submicroscopic structure of animal epidermis. Anat. Rec., 118: A433. RA~OURG, A. and LEBLOND, C. P., 1967. Electron microscope observations on the carboydrate-rich cell coat present at the surface of cells in the rat. 07. Cell Biol., 32: 27-53. REYNOLDS, E. S., 1963. The use of lead citrate at high pH as an electron opaque stain in electron microscopy. 07. Cell Biol., 17: 208-212. SATIR, P. and GILULA, N. B., 1970. The cell junction in a lameUibranch gill ciliated epithelium. o7. Cell Biol., 4/: 468-487. SELBV, C. C., 1955. An electron microscope study of the epidermis of mammalian skin in thin sections. ,7. biophys, biochem. Cytol., 1 : 429-444. SHEFFIELD,J. B., 1970. Studies on aggregation of embryonic cells: Initial celt adhesions and the formation of intercellular junctions. 07. Morph., 132: 245-264. STERN, I. B., 1965. Electron microscopic observations of oral epithelium. 1. Basal cells and the basement membrane. Periodontics, 3: 224-238. TmLANDE~, H. and BLOOM, G. D., 1968. Cell contacts in oral epithelia. 07. periodont. Res., 3: 96-110. WmTTArmR, D. K. and ADAMS, D., 1971. The surface layer of human foetal skin and oral mucosa: A study by scanning and transmission electron microscopy. 07. Anat., 108: 453-464. WnIa~rEN, J. B., 1968. The fine structure of desquamative stomatitis. 07. Periodont., 39: 75-80.