AN ELECTRON MICROSCOPIC STUDY OF AND DYSKERATOSIS IN FOLIACEUS

WITH A SPECIAL NOTE ON PECULIAR INTRACYTOPLASMIC BODIES* GEORGE F. WILGRAM, M.D., JAMES B. CAULFIELD, MD. AND EVE B. MADGIC Our interest in pemphigus foliaeeus stemszation could not proceed (1). Dyskeratosis in from previous electron microscopic studies ofpemphigus vulgaris is very rare and was ob- ancantholysis in (1) and ofserved only in cases that had been treated with dyskeratosis and acantholysis in Hailey-Hailcy'scorticosteroids. In Hailey-Hailey's disease and disease (2) and in Darier's disease (3). It wasin Darier's disease, the destructive changes in felt that a detailed study of the behavior of thethe epiderinal cells are much less severe and the dcsmosomes and the tonofilaments in pcmphigustonofilament-dcsmosome complex seems to be foliaceus might strengthen the concept that theselectively altered—presumably by a genetic dcsmosomcs serve an important function in thedefect (2, 3). Whether in these two genetic dis- process of keratinization (4). orders the tonofilaments arc separated from This concept is based on the following obser-their corresponding desmosomes, or fail to con- vation: Dcsmosomes are present in all strata ofnect with them during their initial synthesis or the epidermis, including the horny layer, al-whether both mechanisms prevail is not estab- though their ultrastructurc varies slightly inlished. Whatever the etiology of the tonofila- different layers of the epidermis (Fig 1, lbment-dcsmosome alteration, the tonofilaments and le). It was suggested that this persistenceare deprived of proper orientation by their of the desmosomes up to the horny layer wasdesmosomes and alterations in keratinization not incidental, but served a special purpose. Itensue. was also postulated that the desmosomes not It seemed not unlikely that the morphologic only serve as points of cohesion between cpi-changes in the cpidermal cells in pemphigus dcrmal cells but also function as points offoliaeeus should lie in between the severe, gener- anchorage for the tonofilamcnts which areally destructive changes in pemphigus vulgaris widely accepted to be the precursors of keratinand the much milder selective lesions in Hailey- (5). Thus the dcsmosomes may provide orienta-Hailey's disease and Darier's disease: this indeed tion for the tonofilaments in the cpidermal cellsis the case, as will be shown in this publication. during their passage from the basal to the In the course of the investigations designed horny layer. And it is felt that proper spatialto explore these possibilities, peculiar intra- orientation of the tonofilaments is necessary forcytoplasmic bodies were observed in epidermal normal keratinization to take place. This con-cells. They will be described in detail and cept formulated on observations on normal skinpossible interpretations will be given. was then strengthened by investigations on the so-called acantholytic conditions, in which a MATERIAL AND METHODS loss of desmosomcs and thereby a loss of their Seven patients with generalized pemphigus fo- orienting function for the tonofilamcnts lead toliaceus were examined. Three of them were mid- dle-aged females, one was an elderly woman, two errors in keratinization. In pcmphigus vulgaris,were young men in their thirties, and one patient a severe necrotizing injury of unknown etiologywas a male teenager. No patients with the Senear- leads to complete destruction of the tonofila-Usher type of were ex- ments with ensuing loss of desmosomes. Deathamined in this series since it was deemed advisable to investigate them separately. Biopsies were per- of epidcrmal cells follows. Therefore, keratini-formed on all these patients during the active state of their eruption. Only small blisters meas- *Fromthe Departments of anduring not more than 2 mm in diameter were taken Pathology, Massachusetts General Hospital, Har- vard Medical School, Boston, Massachusetts. by punch biopsy from several different locations, This work was supported by Grants GM 9726under minimal novocaine anesthesia. An attempt and C-4955 from the National Institutes of Health.was made to select lesions which with reasonable 288 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY 4 1. 4 A" -Q 'I' - "- :0.1 '5, -41

FTc. l.A. Three epidermal cells in the lower Malpighian layer with several desmosomes (ID) are shown. Tonofilaments (Tfil) "sprout" from these desmosomes at the attachment plate (AP). The desmosomes are seen in continuity with the cell membrane (CM). RNP particles are in close association with tonofilaments. Epon >< 45000.

had been placed on therapy just one or two daysthat varied from specimen to specimen. Some- prior to the removal of biopsy specimens. times there was mild crenation in the nucleus, The biopsy specimens were fixed in buffered 1% osmium-tetroxide with sucrose for one hour andperinuclear retraction of the endoplasmic retic- a half. Half the biopsy material was embedded inulum and swelling of the mitoehondria (Fig. 2). prepolymerized n-butyl-methacrylate with 1%Frequently the tonofilaments were separated Luperco. The other half of each specimen wasfrom the attachment plate of their correspond- processed in Epon as recommended by Luft (6)ing desmosomcs (Figs. 3 and 4). The desmo- or Parsons (7). Sections were cut on a Porter-Blum microtome with diamond knives and after mount-somes, however, were usually present in the ing on carbon-coated grids were stained withbasal cell region, and were also distincly visible either uranyl acetate or lead citrate. Pictures werein the lower Malpighian area although the taken on a Siemens elmiscope. epidermal cell changes described above seemed more pronounced (Fig. 5). These mild changes RESULTS in the basal cells and the lower Malpighian When early lesions of pemphigus foliaceuscells became more pronounced in the upper were examined by electron microscopy, theMalpighian stratum and in the sub-corneal layer. dermis appeared unchanged. The basement In the upper Malpighian region of involved membrane was intact (Fig. 2). The basal cellsareas, one could frequently observe on many with tbe aid of the so-called junction granulescell surfaces a diminished number of desmo- adhered to the basement membrane in a normalsomes and in some instances complete loss of fashion. The nucleus and the cytoplasm of thedesmosomes producing acantholytic cells which basal cells, however, often showed early changeseither floated freely or formed rows of detached AN ELECTRON MICROSCOPIC STUDY 289 OH ' ..: I 1, t.

Fio.lB. Several desmosomes can be seen in the upper epidermal strata in this high power electron micrograph. Odland bodies in the upper portion are indicative of the lower granular layer. Tonofilaments are seen streaming from the attachment plate (AP) of the desmosomes. Two densely osmophilic lines and three fine interdesmosomal lines can clearly be distinguished in several of these desmosomes. The desmosome surrounded by a rectangular drawing is illustrated under still higher power in Figure 10. Epon >< 30000. epidermal cells (Fig. 6). This loss of desmo-grounds to the ones seen in normal skin (5, somes appears to have been preceded by re-[Fig. 9]). The nuclei showed crenation with pen- traction of the tonofilaments from their attach-nuclear halo formation due to retraction of the ment plate (Fig. 7). The tonofilaments whichendoplasmie reticulum. The mitoehondria were had separated from their corresponding desmo-either swollen, contracted or partially dissolved. somes at the basal layer were discernible higherComplete epidermal cell degeneration, was how- up in the Malpighian layer in packets (tono-ever, only rarely observed. fibrils) without their usual linear arrangement, Another notable finding in all the cases of and without their proper orientation: Conse- which were examined was quently they appeared as an irregular networkthe presence of round bodies about 2000 Ang- of frequently interwoven packets, representingstroms in diameter with a distinct external dyskeratosis (Fig. 5). Keratohyaline granulesmembrane and a definite fine sub-vesicular in- were present in many of these dyskeratotieternal structure (Figs. 2, 3, 5 and 7). These areas. Tonofilaments were seen to be associatedbodies were seen in the cytoplasm but never with them and to he coursing through them inwithin the nuclei. They could not be found with an apparently normal fashion. Odland bodiescertainty outside the epidermal cells. They were also present in the upper granular areacould be seen most frequently odj scent to and appeared to be similar on morphologicalscantholytic areas but were observed on oces- 290 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY .r ...j,

DM1 DM2

FIG.1 .C.The area in the rectangle in Figure lB is shown here under highest m agniflea- tion. Three fine interdesmosomal lines (IDL) can clearly be distinguished in the inter- desmosomal substance. On both sides of these three fine lines is a heavily osmophilie dense double-layered membrane (DM) which forms the outer limitation of the attach- ment plate (AP). Epon )< 100,000. sion in epidermal cells close to the basementthe epidermis. The earliest visible changes could membrane. They were distinctly different inbe seen in basal cells. Marked changes appeared their structure from mitochondria (Figs. 2 andwhen the epidermal cells had reached the upper 5), and could also be distinguished from Od-Malpighian or the sub-corneal strata. In the land bodies (8, [Fig. I) because they werelower layers of the epidermis, cellular injury larger and had a different internal structure.was mild and most desmosomes persisted. De- In this series of eases there was no unusualspite the continued presence of the desmosomes, evidence of increased pinoeytosis or transfer ofthe arrangement of the tonofilaments was ab- any extracellular material to the interior of thenormal. They had begun to detach themselves cytoplasm in the form of lysosomes. On mor-from the attachment plates of the correspond- phologic grounds, these peculiar bodies wereing desmosomes. This finding is of importance comparable to vaccinia virus; tbe latter, how-because it bears out our previous observations ever, tend to occur in much larger numbers (9). that the first change in acantholysis appears to be detachment of the tonofilaments from the DIscussIoN attachment plate of their eorrespondthg des- Striking pathological changes in this seriesmosomes with subsequent disappearance of the of cases of peinphigus foliaceus were confined todesmosomes. The tonofilaments which had be- loss ofdesmosomes. Thisledtoaeantholysis region there wasusually retraction ofthepen- of theepidermis. Intheepidermalcellsin this and theformation oflaeunaeintheupperstrata pighian areawasastriking decreaseortotal companying thesechanges in theupperMal- seemed tohepresentinnormal amounts.Ac- ets wereassociatedwiththem. Odlandbodies the disorderlyarrangedtonofilamentous pack- tohyaline granuleswerepresent,andsomeof hian region.Inthedyskeratotiematerial,kera- of dyskeratotiematerialintheupperMalpig- were arrangedasirregularlyinterwovenpackets gun toretractinthelowerMalpighianstrata corner. EponX20000. tron mierograph.Apeculiarintracytoplasmicbody(B)canbe seenintheupperrighthand mitochondria (MIT)showsignsofswellingparticularlyinthe leftcellshownonthiselec- with theaidofso-calledjunctiongranules.Thedesmosomes havelargelydisappeared membrane (BM)isintactandthebasalcellsareseenadhering tothebasementmembrane in thisparticularareashownhereandthetonofilamentshave losttheirorientation.The Fie. 2.Earlyacantholysisatthebasallayerinpemphigus foliaceus.Thebasement AN ELECTRONMICROSCOPICSTUDY is lessstriking thanthat seenineither Hailey- tion ofthetonofilaments fromtheirdesmosomes dyskeratosis which followstheinitialsepara- Hailey's disease (2)andDarier'sdisease(3). The the relativelymildselectivelesions ofHailey— pemphigus vulgaris(1)and moreseverethan be lessseverethanthedestructive changesof foliaeeus seenbyeleetronmieroscopy seemto indicated mildcellularinjury. tial dissolution.Thesemorphologicalterations either swellingorcontractingevenbypar- The mitochondriashowedmoderatechangesby nuclear cytoplasmresultinginhaloformation. The epidermalabnormalities ofpemphigus

• d•_ 291 292 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY 7•. i SI CAbs C -I > ;

Fm. 3. Early acantholysis in the lower Malpighian layer. Two desmosomes (D) are seen persisting in this area. However, the tonofilaments (Tfil) have separated from their attach- ment plates and are seen as irregular aggregates of tonofilaments into tonofibrils. A peculiar intracytoplasmic body (B) is shown in the right lower corner of this photomicrograph. Epon X 20000.

Hailey's disease or Darier's disease, but never-called acantholytic diseases (pemphigus vul- theless resembles it in certain ways, i.e., thegaris, Hailey-Hailey's disease, Darier's disease, tonofilaments after separation from the desmo-and pemphigus foliaccus) are most likely dif- somes are no longer visible in an orderly ori-ferent. The pathologic alterations in the cyto- ented fashion in linear, discrete bundles, butplasm of the epidernial cells vary greatly in appear as aggregates of packets in a disorderlyextent and severity. The changes in the tono- arranged network. filament-desmosome complex, however, arc In all four "acantholytie" conditions, acan-morphologically closely related, and appear as tholysis is due to loss of desmosomes which isthe earliest epidermal cell alteration. This is apparently preceded by the above describedactually quite predictable, since the tonofila- alteration in the tonofilaments. The associationment-desmosome complex is a highly organized of acantholysis and dyskeratosis can now bestructure and is thus readily susceptible to a readily explained, because both phenomena aregreat variety of injurious stimuli. And it ap- due to injury to the same subcellular structure,pears that different injuries produce morpholog- e.g., the tonofilament-desmosome complex. Aically similar end results in this specialized possible explanation for the detachment of thestructure (4), which lead to acantholysis and tonofilamcnts from their desmosomes and thedyskeratosis. It is only in pemphigus vulgaris subsequent loss of dcsmosomes will be giventhat acantholysis is seldom accompanied by below. dyskeratosis. The severe necrotizing injury in The etiolcgy and pathogencsis in all four so-this disease apparently abolishes the viability AN ELECTRON MICROSCOPIC STTJDT 293 tr' V 'SI ¼.

.4 e,I r Äô 4:: /tT

FIG.4. Acantholysis in the lower Malpighian layer in pemphigus foliaceus. Two desmo- somes (D) are still persisting. The tonofilaments, however, have entirely retracted from the cell surface and are shown in an irregular pattern throughout the cytoplasm of the epidermal cells. The mitochondria (MIT) show signs of mild injury. Epon ><16000. of the epidcrmnl cells which is necessary for kera-Interpretation of the peculiar intracytoplasmic tinization to proceed. By contrnst, in pcmphigus bodies observed in this series of eases of foliaceus, the injury is not severe enough to pemphigus foliaceus destroy the viability of the epidermal cells nnd The peculiar bodies described under Results thus abolish keratinization altogether. Epider-were seen in all seven patients examined in this mal cell injury, however, seems to be severeseries. One of the possible interpretations of enough to disrupt the tonofilament-desmosomethese peculiar bodies is that they represent complex to interfere with the normal orientationvirus-structures. If these bodies were an infect- of the tonofilaments and to hinder their transfor-ing virus, they could easily cause sufficient in- mation into keratin. As a consequence of thesejury to the epidermal cells to disrupt the alterations, acantholysis and dyskeratosis ensues.highly specialized desmosome-tonofilamcnt com- The orienting function of the desmosomes forplex, but still permit the cell to survive. With tonofilaments has also recently been shown bythe breakdown of the desmosome-tonofilament electronmicroscope studies on embryologicalcomplex, acantholysis as well as dyskeratosis epidermal cells (10, 11). First desmosomcs ap-would ensue. It is obvious that electronmicros- pear at points of contact between epidermalcopy can not demonstrate whether these pecul- cells and then tonofilaments "sprout" from theiar bodies are in fact viruses; or if viruses, attachment plates of these newly formed des-whether they are the cause of the disease. Proof mosomes. of this sort can only be obtained by other types 294 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY

_j#* IC' 4

? '? ?\

3

FIG. 5. This electron micrograph shows the earliest change in pemphigus foliaceus in an area that was more heavily involved than those shown in Figures 2—4. Desmosomes (D) are still seen persisting but tonofilaments do not connect with their attachment plates anymore. The tonofilaments are seen instead as irregularly oriented strands throughout the cytoplasm. This particular picture shows that desmosomes do persist while tonofilaments have retracted from them. Thus this picture tends to prove that tonofilamentous separa- tion preceded disappearance of the desmosomes. In the right middle portion of this picture some tonofilaments are still seen inserting into an apparently intact desmosome. Mito- chundria (MIT) show signs of definite swelling indicative of cellular injury. A peculiar intracytoplasmic body (B) is seen in the upper portion of this electron micrograph. Epon X 20000. of tecbnics. The arguments concerning the virnspatients with and with etiology of pemphigus vulgaris have been re-erythema multiforme, thus weakening their viewed by Nasemann and Marchionini in 1957argument. Yet the possibility that pemphigus (12, 13). In 1962, Crosti and co-workers re-foliaceus might be caused by a virus was ported that they had cultured a virus obtainedmade as early as 1931 by Tlrbach and Reiss from the skin of "pemphigus" (14). These(15). The finding of the present study should authors did not differentiate between pemphi-at least renew interest in the concept of the gus vulgaris and pemphigus foliaceus. Further-virus etiology, since these peculiar bodies were more, they reported finding a similar virus inobserved in all biopsy specimens taken from AN ELECTRON MICROSCOPIC STUDY 295 • . vs in. - ••

Ac

FIG.6. An advanced state of acantholysis near the granular layer in pemphigus foliaceus. The desmosomes have all disappeared on the three cells shown here. The tonofilamerits have all retracted, into an irregular network of tonofibrils around the nucleus, forming dyskeratotic material (DY). Methacrylate x20000. seven patients with pemphigus foliaceus. Onbeen examined so far. However, such bodies morphologic grounds these bodies compare wellhave as yet not been found in a large and with vaccinia virus (9), but they have not beenvaried amount of epidermal pathology that has observed in clusters of large numbers tobeen subjected to electronmicroscopy. A defi- justify a morphologic designation of virus withnite interpretation of the true nature of these certainty. peculiar bodies awaits further work with tissue There arc other possible interpretations ofculture technics, as well as increased experience these bodies: a) They might represent alteredin the electron microscopy of diseases of the mitocbondria, although normal mitochondriaepidermis. Finally, it should be noted that simi- arc usually much larger than the bodies seenlar bodies were not seen by us in a series of here; b) These bodies could be a product offive cases of pemphigus vulgaris (1) although cellular injury peculiar to pemphigus foliaceus,virus bodies were looked for. (This point is but of a non-viral nature; c) The possibilityunder reinvestigation.) The present finding of has been considered that these pecuhnr bodiesthese peculiar bodies, however, might suffice to represent lysosomes, but their distinct externalrevive an interest in the possibility of a viral membrane and their consistent size do notetiology of pemphigus foliaceus. match the lysosomal structures seen under a variety of other circumstances; d) One should SUMMARY consider that similar bodies might also be ob- Seven cases with generalized pcmphigus served in other skin conditions that have notfoliaceus were examined by electron micros- 296 THEJOURNAL OF INVESTIGATIVE DERMATOLOGY

I'

,rP p1 41

Ac4

4. ii FIG. 7. Advanced acantholysis near the granular layer in pemphigus foliaceus. No des- mosomes can be seen on the two cells shown here. The tonefilaments have lest their orientation and are illustrated as an irregular network of fibrils. A peculiar intracytoplasmic body (B) is shown in the right portion of this electron micrograph. It is illustrated under higher magnification in the insert in the left lower corner of this picture. Epon X 17000. Insert Epon >< 50000. AN ELECTRON MICROSCOPIC STUDY 297

4.

Ac

• — Ii;_: — rId'

FIG.8. Advanced acantholysis with complete loss of desmosomes in the sub-granular area in pemphigus foliaceus. The tonofilaments form an irregular network of interwoven fibrils after their retratlon from the desmosomes. This network of fibrils represents dyskera- tosis (DY). Epon X 18000. 298 TUEJOURNAL OF INVESTIGATIVE DERMATOLOGY

Ac

Fn. 9. Acantholysis (AC) in a sub-corneal lacuna in pemphigus foliaceus. The area of the lacunae is represented by the space AC. The granular cells are characterized by numer- ous Odland bodies (OB) and keratohyaline granules (KIIG). Tonofilaments (Tf) are seen to be associated with keratohyaline granules. Methacrylate X 15000. copy. No case with pcmphigus erythematosusrelationship in pemphigus foliaceus seem to was included in this series. The dermis and thesupport the concept that the desmosomes have basement membrane appeared normal. Strikingan orienting function for the tonofilaments and pathological changes were confined to the epi-thus serve a purpose in the process of keratini- dermis. The earliest change appeared to be azation. separation of the tonofilamcnts from the attach- Peculiar bodies were observed in the cyto- ment plate of their corresponding dcsmosomesplasm of epidermal cells in all seven cases of in the basal region. Acantholysis and dyskerato-pemphigus foliaceus. Different interpretations sis in the Malpighian and granular layer wereare given for these peculiar bodies. The op- due to injury of the same subcellular structure,portunity is taken to revive interest in the pos- e.g.,thetonofilament-desmosome complex. Be-sibility of a virus etiology for pemphigus cause epidermal cell injury was mild cellularfoliaceus. viability and attempts at keratinization were retained to some degree. Epidermal cell injury, REFERENCES however, was sufficiently severe to cause a de- 1. WILGaAM, G. F., CAULFIELD, J. B. AND LEVEE, W. F.:An electronmicroscopic study of tachment of the tonofilaments from their des- acantholysis in pemphigus vulgaris. J. mosomes. This was followed by: 1) a loss of Invest. Derm., 36: 373, 1961. 2. WiLGEAM, G. F., CAULFiELD, J. B. AND LEVEE, desmosomes leading to acantholysis, and 2) loss W. F.: An electron microscopic study of of proper orientation of the tonofilaments lead- acantholysis and dyskeratosis in Hailey- ing to dyskeratosis. Hailey's disease. J. Invest. Derm., 39: 373, 1962. These findings of the tonofilament-desmosome 3. CAULFiELD, J. B. AND WiLGEAM, G. F.: An AN ELECTRON MICROSCOPIC STUDY 299

electron microscopic study of dyskeratosis scope. Vaccinia and fowl pox viruses. J. and acantholysis in Darier's disease. J. Exp. Med., 100: 301, 1954. Invest. Derm., 41: 57, 1963. 10. OVEETON, J.: Desmosome development in 4. WILGRAM, G. F., CAULFIELD, J. B. AND MAD- normal and reassociating cells in the early ole, E.: A possible role of the desmosomes in chick blastoderm. Develop. Biol., 4: 532, the process of keratinization. In: The 1962. Epidermis. New York, Academic Press,11. HASHIMOTO, K.: Personal communication, 1964. March, 1964, Boston, Massachusetts. 5. Mntcim, E. H.: Keratin and Keratinization.12. MARCRIONINI, A. AND NASEMANN, T.: On the International series of monographs on Pure virus etiology of pemphigus and dermatitis and Applied Biology. New York, Pergamon herpetiformis Duhring. J. Invest. Derm., Press, 1961. 24: 267, 1955. 6. LUFT, J. H.: Improvements in epoxy resin 13. MARCHIONINI, A. UND NASEMANN, T.: tTher embedding methods. J. Biophys. Biochem. die virus aetiologie des pemphigus und der Cytol., 9:409, 1961. dermatitis herpetiformis Duhring. Derma- 7. PARSONs, D. F.: A simple method for obtaining tologica, 115: 320, 1957. increased contrast in Araldite sections by 14. CROSTI, A., GANOTTI, F., HAHN, E., BUBOLA, using post-fixation staining of tissues with D. AND BONANI, V.: Ulteriore ricerche viro- potassium permanganate. J. Biophys. Bio- chem. Cytol., 11: 492, 1961. logiehe nelle dermatosi pemfigose. G. Ital. 8. ODLAND, G. F.: A submicroscopic granular Derm., 2: 81, 1962. component in human epidermis. J. Invest. 15. URBACR, E. AND REISS, F.: Tier-experimen- Derm., 34: 11, 1960. telle untersuehungen zur frage der infektios- 9. MORGAN, C., ELLISON, S. A., RosE, H. M. toxischen genese des pemphigus vulgaris AND MoORE, D. H.: Structure and develop- und der dermatitis herpetiformis Duhring. ment of virus observed in the electron micro- Arch. Derm. (Chicago), 162: 713, 1931.