66 KOBA Y ASHI AND HASHIMOTO

Dermatol 105(suppl):1l-1 2, 1981 37. Chu A, Kung P, Edelson R: Dermal Langerhans cells in cutaneous

lymphoma. An in situ study using monoclona l ant ibod ies (abstr) . J Invest Dermatol 76:324, 1981

38. Thomas JA, Janossy G, Graham-Brown RAC, Kung PC, Goldstein G: The relationsh ip between T-Iymphocyte subsets and la- like ant igen posit ive nonlymphoid cells in early s tages of cutaneous T cell ly mphoma. J Invest Dcrmatol 78:1 69-176, 1982

39. McMillan EM, Wasik R, Everett MA: HLA-DR positive ce lls in large plaque (atrophic) parapsoriasis. J Am Acad Dermatol 5:444-449, 1981

0022-202X/ 83/ 800 1-0066$02.00/ 0 THE JOURNAL OF INVESTIGATIVE DERMATOLOGY , 80:66-72, 1983 CopYl'ight © 1983 by The Williams & Wilkins Co.

Vol. 80, No.1

40. ' I'jerlund U: Beta-2-microglobulin a nd HLA-DR (la) antigens in normal and pathological skin. Acta Derm Venereol [Suppl] (Stockh) 94, 1981

41. McMillan EM, Martin D, Wasik R, Everett MA: Demonstration in s itu of T-ce lls and T-cell subsets in lichen planus using monoclo­nal antibodies. J Cutan Pathol 8:228-234, 1981

42. MacKie RM: Initial event in mycosis fungo ides of the s kin is viral infection of epidermal Langerhans cells. Lancet 2:283-285, 1981

43. Norris DA: The pathogenesis of mycosis fun goides. Clin Exp Der­matol 6:77-87, 1981

Vol. 80, No.1 Printed in U.S.A .

Amyloidogenesis in Organ-Limited Cutaneous Amyloidosis: An Antigenic Identity Between Epidermal Keratin and Skin Amyloid

HITOSHI KOBAYASHI, M.D. AND KEN HASHIMOTO, M.D.

DepartmeJit of Dermatology an.d Syphilology, Wayn.e State University School of Medicine, Detroit, Michigan, U.S.A.

Epidermal keratin was extracted and antibody against this protein was produced in rabbits. Various forms of organ-limited cutaneous amyloidosis (lichenoid, macu­lar, and nodular amyloidosis, and basal cell epithelioma) and primary systemic amyloidosis were immunohisto­chemically examined to test the identity between epider­mal keratin and skin amyloid.

Amyloids in lichenoid and macular amyloidoses, and in basal cell epithelioma had an identical antigenicity with epidermal keratin, whereas amyloids in nodular amyloidosis and systemic amyloidosis did not have this identity. In addition, amyloid in lichen amyloidosis con­tained disulfide bonds as in keratin. Connective tissue components including filaments of fibroblasts a~d vas­cular endothelial cells did not react with this antikeratin antibody. It was concluded that at least some of the amyloid substance in organ-limited cutaneous amyloi­dosis is derived from degenerated epidermal keratino­cytes through filamentous degeneration or apoptosis.

Among various forms of amyloidosis occurring in the skin, such ,as primary systemic amyloidosis, secondary systemic amy­loidosis, familial systemic amyloidosis, and organ-limited cuta­neous amyloidosis, the organ-limited cutaneous · amyloidosis appears to be a distinct entity. This entity distinguishes itself because (1) it is not related to any systemic diseases [1,2], (2) amyloid deposition is always limited in subepidermal area, and (3) sometimes it is found even within the epidermis [3]. It has

Manuscript rece ived Dece'mber 4,1981; accepted for publication May 26, 1982.

Supported in pa rt by a research grant from the Veterans Adminis-tration.

Reprint req uests to: Dr. Ken Hashimoto, Department of Dermatol­ogy and Syphilology, Wayne State University School of Medicine, 540 E. Canfield Street, Detroit, Michigan 48201.

Abbreviations: DAGM: N-(7-d imethylamino-4-methylcoumarinyl) -malemide PBS: phosphate-buffered saline SDS-PAGE: sodium dodecyl ulfate-polyacrylam ide gel electro­

phoresis

been suggested by several investigators that the epidermis may participate in the pathogenesis of organ-limited cutaneous amy­loidosis [4,5]. Recently, we reported that in various forms of organ-limited cutaneous amyloidosis degenerated epidermal cells drop off into the dermis instead of normally migrating upward to the skin smface, and we postulated that those degenerated epidermal cells contribute to the formation of skin amyloid [6-9]. Ultrastructurally, tonoftlaments in degenerating epidermal cells sequentially change their featmes ("ftlamentous degeneration") and become morphologically identical to amy­loid filaments [6-9].

Recent biochemical studies on keratin have clarified that epidermal keratin consists of a number of polypeptide subunits which are referred to epidermal fibrous protein or keratinous protein [10,11). Furthermore, it has been shown that keratin ftIaments can be assembled in vitro from this group of polypep­tides or form a mixtme of purified monomers [12,13]. Recent progress in pmification procedure of keratin made it possible to produce specific antibody and to demonstrate that it is t he tonofilaments of epidermal cells that express the antigenicity of pmified epidermal keratin [14-16).

In the present study we pmified this epidermal keratin and produced antiserum in rabbits. Amyloid tissues from organ­limited cutaneous amyloidoses (including lichenoid, macular, and nodular amyloidoses, and amyloid associated with basal cell epithelioma) and primary systemic amyloidosis (AL amy­loid) were immunohistoch emically tested using the antiserum thus produced to determine whether any antigenic relationship exists between skin amyloid and epidermal keratin.

MATERIALS AND METHODS

Amyloid Tissues

Six biopsies were obtained from 3 patients with lichen amyloidosis, 2 biopsies from 2 patients with macular a myloidosis. In addition, 1 biopsy from a patient with basa l cell epithelioma, 1 biopsy each from 2 patients with nodular a myloidosis, and skin and heart tissues from an autopsied patient with primary systemic amyloidosis were obtained. Either freshly frozen sections or formalin-fixed paraffin sections were cut at 4 lim and used for the following immunohistochemical stainings. The presence of amyloid was always confirmed by Congo red and crystal v iolet stainings on serially cut sections. For immunoelectron microscopic procedures, specimens were fixed in 4% paraformaldehyde

Jan. 1983 KERATIN AND SKIN AMYLOID 67

or 1% glutaraldehyde in phosphate buffer and frozen sections were cut at 10-20 11m . 1 a Preparation of Antigen and Antibody

The extraction and pmification of epidermal keratin from human plan tar callus were performed by Steiner t's method [11]; the keratin material was suspended against Tris-HCI buffer, containing 8 M urea and 25 M 2-mercaptoethanol, and resulting Ul'ea and mercaptoethanol sol.uble fract ion was purified by the assembly and disassembly system 3 times. SDS-PAGE was done using the discontinuous system described by Neville [1 7]. Assembled filaments in vitro were examined under an electron microscope. Antibod ies against purified epidermal keratin were raised in 3 White New Zealand rabbits by fTeq uent injections of antigen. AntI~erum was obtained from those that had the highest titer (5,120x passive hemagglu t ination test, 640X by indirect immunoflu orescence). In addition to our ant iserum, antikeratin rabbit serum produced in a sim ilar fashion was kindly offered by Dr. Tung-Tien Sun.

[mmunohistochemi~try

Immunofluorescent a nd immunoperoxidase techniques were per­formed by the methods described by Beutner et al [18] and Taylor [19]. Before the primary incubation, endogenous peroxidase activity of the skin lesion was blocked with methanol containing 0.3% hydrogen peroxide [19]. Primru'y incubation was done with 1:40 diluted antiserum against epidermal keratin (anti keratin serum)' for control studies preimmune rabbit serum and antiseru.m absorbed by antigen (2 mg/ ml epidermal keratin) were used. Secondary incubation was done with 1:20 diluted FITC-conjugated a ntirabbit IgG serum produced in goat (Cap­pel Laboratories, Cochranville, Pennsylvania) for immunofluorescence ? nd 1:40 diluted peroxidase-conjug!\ted a ntirabbit IgG serum produced 111 goat (Cappel Laboratories) for immunoperoxidase reaction. The method of Graham and Karnovsky was used for demonstration of peroxidase [2?J. P rolonged incubation in primary and secondru'y a nt i­sera and rU1s ll1g were calTied out at 4°C for immunoelectron micro­scopic sludy according to Na kane's method [21]. After diaminobenzi­dine reaction by Graha m a nd Kru'novsky's method, sections were fixed in 1% osmium tetl'Oxide, dehydrated, and embedded in AJ·aldite. Ultra­thin sections were observed wi th 0 1' without coun terstainings. For contro l stud ies, specimens were incubated with preimmune rabbit serum II1stead of specific antiserum and were processed in the same manner .

. In addition to immunohistochemical studies by antikeratin serum, ' duect Immunofluorescence was performed on frozen sections of lichen a myloidosis and primal'y systemic amyloidosis with FITC-conjugated ?ntIhuman, IgG, IgA, IgM, C3, fibrinogen, and albumin sera produced U1 goat (Cappel Laboratories).

Histochemical Del/wnstration of Sulfhydryl Groups (H) and Disulfide Bonds (S-S)

The distribu tion of sulfhydr yl groups and disulfide bonds was histo­chemically examined in the lesions of lichen amyloidosis a nd primary system.ic amyloidosis, using N-(7-dimethylamino-4-methylcoumru'inyl) ­mal em Ide (DACM) (Teika Seiyaku Co. Ltd ., Toyama, Japan) accord ing to Ogawa's method [22].

RESULTS

Antigen Preparation and Antiserum

By SDS-PAGE, purified epiderm al keratin was resolved into 4 major polypeptides ranging from 40,000 to 65,000 daltons (Fig 1a). In Fig 1b are sh own assembled keratin filaments that are 8 nm wide and up to several micra long. These findings conform to Steinel't's original data [ll]. Antibody production was con­firmed by Ouchterlony gel diffus ion test: precipitation lines were produced between antigen a nd a ntiserum o f 3 rabbits. N o precipitation line was formed w he n preimmune control serum was used (Fig Ie).

Immunohistochemical Stains with Antiheratin Serum

Immunohistoc h emical findings are summarized in Table 1. With both direct immunofluorescence a nd immunoper oxidase techniques, whole epidermis, ha ir folli cles, sweat ducts, a nd

• Th is 2-step procedure will be referred to as indirect immunofluo­rescence 0 1' indirect immunoperoxidase stain.

2€

lC

FIG 1. Characterization of ant igen and a ntibody. a, SDS-PAGE of purified epiderma l fibrou s protein (lefliube).1'he sa mple was dissolved in lhe presence of 4 M m ea, 5% 2-mercaptoethanol , and 2% SDS and was run in 10% acrylam ide gel at pH 8.8. Epidermal fibrous prot~ i n is composed of 4 major polypeptides (arrowheads). Markel' proteins were run in the right tu be (phosphorylase B, 92,500; bovine serum a lbumin, 66,200; ova l-bumin, 45,000; soybean trypsin inhibi tor, 21 ,500; lysozyme, 14,400). b, In vitro assembled keratin fila ments from purified epidermal fibrous protein . Individ ual filament.s average 8 nm wide and up to several micra long. Bar = 0.] 11m . c, Ouchte rlony gel diffusion test. Cenier well conta ins 1.0 mg/ ml ep iclermal fibrou s protein. P eripheral wells conta in antisera or 3 rabbits (/ ,2,3) and preimmune control sera 0(' each rabbi t (/C,2C,3C) .

some of th e sweat gla nd cells combined with a n tikeratin serum and showed strongly positive staining. T h e reaction was ob­served up to 1:640 dilution of antiserum. Dermal components, such as fibroblasts , vessels, smooth muscles, and collagen did not reveal any positive r eaction with either immunofluores­cence or immunoperoxidase tech niques. There were no differ­e n ces in t h ese staining patterns of t h e skin nor following amy­loid tissues between a n tiserum donated by Dr. Sun [15,16,23] and ours. Dermal elastic fibers showed yellowish autofluores­cence w hich was easily differentiated from greenish FITC flu­orescence.

In lic h en amyloidosis, amyloid depositions, which were stained pinkish by Congo red, were presen t in papillary dermis adjacent to t h e epidermis (Fig 2a). They sh owed greenish birefringence under a polarized light t hat was distinctly differ­e n t from white birefringence of collagen 01' yellowish orange birefringence of stratum corneum of the epidermis (Fig 2b). Indirect immunofluorescence demonstrated a strong positive fluorescence in t h e epidermis and th e al'ea where amyloid depositions were confirmed by Congo red. The fluorescence of amyloid masses was as str ong as t h e epidermis and it was adj acent a nd continuous to t h e epidermal cells (Fig 2c,d). Amyloid presen t in macular a myloidosis showed similar posi­tive reaction to the antiseJ'llI11. In t h e basal cell epithelioma , amylo id depositions were present in the stroma among tumor nests (Fig 3a). By indirect immunoperoxidase technique, amy­loid depositions were s hown positive to a n tikeratin serum ' also some tumor cells revealed a weakly positive reaction (Fig 3b): The intensity of this positive stain was similar to that of the epidermal basal cells in t h e same section. Fibrous components of t h e dermis such as collagen and e lastic fibers were negative (Fig 2c,d). F ilamentous structures of fibroblasts, vascular en­dothelial cells, smooth muscles (arrector muscles and perivas-

68 KOBAYASHI AND HASHIMOTO Vol. 80, No.1

TABLE I. Immunohistochemical findings

Antihuman Anti-ep iderma l Fibrous protein Controls IgG IgM IgA C;J Fibrinogen Albumin

Lichen amyloidosis

Macular amyloidosis

Basal cell epithelioma Nodular amyloidosis Primary

Systemic amyloidosis Skin lesion Heart lesion

1 2 3 1 2

++ ++ ++ ++ ++ ++

cular and periglandular myoepithelial cells) were also ~egative (Fig 2c,d). In nodular amyloidosis, however, no reaction was demonstrated in amyloid deposit ions by antikeratin serum. In primary systemic amyloidosis, neither cutaneous amyloid dep­ositions (Fig 4) nor a myloid depositions in heart combined with antiserum. The reaction of antiserum with both epidermal cells a nd amyloid was completely blocked by ant igen absorption (Fig 5). Preimmune control serum did not show a posit ive reaction by eit her indirect immunofluorescence or immunoperoxidase techniques.

Immunoelectron Microscopy

lmmunoperoxidase technique demonstrated densely stained amyloid islands. Under a high magnification, positively reacted peroxidase depositions were localized on amyloid filaments. Surrounding collagen bundles revealed no reaction (Fig 6a). No peroxidase depositions were seen on control sections that were reacted with preimmune 'con trol serum. Rigid, straight, and B-10 nm wide typical amyloid filaments were well observed in control sections (Fig 6b) .

Direct Immunofluorescence with Antihuman Serum Proteins

All antisera against human serum proteins tested here showed positive flu orescence on amyloid depositions of both lichen amyloidosis and primary systemic amyloidosis (Fig 7). The depositions of serum proteins were rather uneveri in each amyloid mass of lichen a myloidosis (Fig 7a), whil e they were

+ +

+

+ +

+

+ +

+

+ +

+

+ +

+

+ +

+

FIe: 2. Lichen amyloidosis. 0 , Congo red staining demonstrates amyloid dep­ositions (A) in papi llary dermis adjacent to the epidermis. Countersta ined by he­matoxylin . X 120. b, Under a polarized light, amyloid depositions (A) show greenish birefringence. Birefringence in dennal collagen is whi tish and that of the ~tratum corneum is yellowish orange. X 120. c, lmmuno(luorescence with an­tikeratin serum on a serial section to Fig 20. Positive reaction is seen in the epi­dermis and amyloid depositions (A). No reaction in the dermis. Asterisli indicates the area enlarged in Fig 2d. X 120. d , High magnifica tion of area marked by aster isk in Fig 2c. Strongly reacted ho­mogeneous amyloid masses (A) are present adjacent to the epidermis (E). x 480.

homogeneous in amyloid depositions of primary systemic amy­loidosis (Fig 7 b) . Positive depositions of immunoglobulins were not diminished by extensive r insing in phosphate-buffered sa­line (PBS) (pH 7.4) or incubation in acidified buffer (glycine­HCI buffer, pH 2.0) before the incubation in FITC-conjugated antisel·a.

Sulfhydryl Groups and Disulfide Bonds

Sulfhydryl groups, which are abundantly possessed by non­keratinized, living keratinocytes, were not detected in amyloid depositions of lichen amyloidosis (Fig Ba) . Disulfide bond, which are not present in living keratinocytes bu t present in fully keratinized cells, were positively demonstrated in amyloid depositions of lichen amyloidosis (Fig 8b). Amyloid depositions of primary systemic amyloidosis contained neither sulfhydryl groups (Fig 9a) nor disulfide bonds (Fig 9b).

DISCUSSION

A study similar to ours has been reported by Masu et al [24] in which they demonstrated the binding of ant i keratin anti­bodies to the amyloid of lichenoid and macular amyloidoses by immunofluorescence. No detailed studies of ultrastructw'allo­calization of antikeratin an tibodies on amyloid have been in­cluded. The present immunohistochemical study confirmed their results a nd further demonstrated that basal cell epithe­lioma-associated amyloid has an antigenic iden tity with epider­mal keratin purified from stratum corneum of the epidermis. A

Jan. 1983

FIG 3. Basal ce ll epi thelioma. a, Co ngo red s ta ining. Amyloid dep­osit ions (arrowh ead.,) which show greenish birefringence under a po­la rized light are present among tumor nests. x 190. b, Immunoperoxi­dase reaction with a nLikeratin serum . Amyloid depositions (arrow· heads) a.re strongly posit ive. Some tumor cells, presumably those undergo ing e ith er filam entous degeneration or advanced keratinization are a lso stained moderately strong. Coun Le rstained by methyl green. x 190.

positive reaction with DACM, which becomes fluorescent only when combined with sulfhydryl groups [22], indicated that antikeratin antibody-posit ive amyloid indeed contains sulfur­rich proteins such as keratin . We also demonstrated that anti­keratin antibodies are ultrastructw·ally localized on amyloid filaments. Interestingly, amyloid in nodular amyloidosis of t he skin and, expectedJy, amyloid in primary systemic amyloidosis do not crossreact with epidermal keratin. These results support our hypothesis obtained by ul trastructw-al observations that some of skin amyloid is at least partially derived from degen­erated epidermal cells th.rough "fil amentous degeneration" of tonofilaments and subsequent dropping off into t he dermis [6-9].

Immunologic techniques have become a useful tool widely used in diagnostic and research studies. Several investigators have u tilized these techniques in the detection or quantitation of SAA protein by antisera against amyloid AA fibril protein [25,26] or immunohistochemical typing of amyloid fibril by an tisera against various amyloid fibril proteins [27]. On the other hand , it was pointed out by Glenner [28] that utilization of antisera in diagnostic studies of amyloidosis and chemical analysis of amyloid require caution because of the problem of pw-ity of antigens, and hence the specificity of antisera. It is also necessary to consider the tendency of amyloid to absorb serum proteins nonspecifically.

In the present study, there seem to be no possibilities that antigen contains any substances other than epidermal fibrous protein: other cellular components, such as cell membranes, cellular organelles, nuclear proteins, actins, etc., should have been removed dUTing the pw-ification procedure of the keratin

I(ERATIN AND SKIN AMYLOID 69

material because t he final precipitate was free from these contaminan ts under the electron microscope. It is also known that repeated purification removed any traces of nonkeratin proteins [12). OUT antiserum, therefore, should have been highly specific to epidermal fibrous protein.

Depositions of serum proteins in skin amyloid have ab·eady been reported by several investigators [29-31). Piamphongsant [30] considered t hat skin amyloid was an accumulation of fragmented y-globulins leaked from capillaries damaged by scratching. MacDonald et al [29] considered that amyloid might be acting as a filamentous sponge to which in1l11unoglobulins

FIG 4. Primary systemic a myloidosis, skin lesion. a, Congo red stain­ing. Amyloid depositions (A) which show greenish biTefringence under a polarized light are present beneath the epidermis x 190. b, Immuno­flu orescence with ant ikera Un serum . No positive reaction in subepider­ma l aTea. Epidermis (E) a nd sweat dueLs (S) are sLrongly posiLive. Dermal elastic fibers show ye llowish autoflu orescence. x 190.

·FIG 5. Lichen amyloidosis. Immunofluorescence with antigen-ab­sorbed a ntiserum (2 mg/ ml epidermal fibrous prote in). No reaction is produced in eith er epidermis or amyloid depositions. x 120.

70 KOBAYASHI AND HASHIMOTO

and complements were absorbed, because immunoglobulin dep­ositions were more intense toward the periphery of amyloid depositions. Similar serum protein depositions were reported in colloid bodies of lichen planus [32] which also derive from tonofilaments of epidermal keratinocytes [33]. An extensive absorption of serum proteins by amyloid was autoradiographi­cally observed in chemically induced early amyloid deposit ions

6b c D!lpm

F IG 6. Lichen a myloidosis. Ct, l mmunoelectl'on microgra ph of' the spec!men reacted with antikeratin serum. Positive ly reacted dense pe roxidase depositions are observed on a mylo id fi la ments (A). S ur­rounding collagen fibers (C) reveal no reaction. Counterstained by uranyl acetate and lead citrate. b, Control specime;l r eacted with pre immune cont rol serum. R igid , straigh t, 8-10 nm wide a myloid fLl a­Illents a re seen without peroxidase depositions. Coun te rstained by uranyl acetate and lead citrate.

Vol. 80, No.1

of mice [34]. The removal of serum proteins from amyloid by in vivo perfusion with salines, however, did not alter the tinc­torial natUl'e of amyloid [34]. In th e present study we fa iled to remove immunoglobulins by an extensive rinsing with PBS and a low pH acidic buffer. It seems to be difficult to remove serum proteins in vitro. We believe that various serum proteins were nonspecifically absorbed by amyloid as a filamentous sponge in lichen amyloidosis, despite the failure to remove serum proteins in vitro: in contrast to a uniform deposition of serum proteins in primary systemic amyloidosis (AL protein) , there was a tendency of irregular depositions of serum proteins in lich e n amyloidosis. This fmding may suggest that serum proteins in the latter were not an inherent component of amyloid .

Demonstration of disulfide bonds in amyloid deposit ions of lichen amyloidosis, but not in those of primary systemic amy­loidosis, indicates that th eir biochemical natUl'e is different.

Sulfhydryl groups abundantly present in living keratinocytes are converted to disulfide bonds in fully keratinized cells in th e stratum corneum. Amyloid depositions of lichen amyloidosis contain the latter rather than the former. This conversion is well demonstrated in F ig 9a, b. It was speculated that tonofil­aments become mature by the formation of intermolecular disulfide bonds dUl'ing keratinization [15]. In dyskeratotic cells of squamous cell carcinoma [35] and in sunburn cells [36], this disulfide bond formation was histochemically observed. Those cells were ultrastructurally found to .be composed of keratin­like filaments [33,37] and indistingu ishable from colloid bodies of lichen planus or degenerated epidermal cells of lichen amy­loidosis [33]. While those cells are out of the normally keratin­izing process, fil ament matUl'ation by disulfide bond formation seems to progress. Filaments in those cells, however , are dis­tinctly different from those of normally keratinized cells in which 10 nm thick, orderly al'l'anged keratin fil aments are produced. Keratin filaments in degenerating epidermal cells of lich en amyloidosis are thinner and less electron dense than those in fully keratinized cells [7,33]. Furthermore, neither keratohyaline granules which are considered to supply a matrix protein of keratinized cells, nor thickening of cellular envelope were found in those degenerating cells [33]. From the above, the preCUl'sor cells of amyloid, i.e., cells undergoing filamentous degeneration seem to be more keratinized than cells in th e viable layer of the epidermis (basal through granular layers) but not completely keratinized. It is speculated that these incompletely keratinized precursor cells are easily degraded after (hopping into the dermis because of t heir immatUl'ity.

In previous work from our labora,tories, the role of connective tissue cells in the production of skin amyloid had been empha­sized by the ultrastructUl'a1 findings, mainly because connective tissue cells surrounding amyloid contained amyloid fil aments in their vacuoles and endoplasmic reticulums [38,39]. Consider­ing the results obtained in the present study, connective t issue -cells may act as phagocytes of degenerated epidermal cells dropped into the dermis, and may further process pre-amyloid filaments from the material contained in degenerated epidermal cells.

The absence of amyloid depositions in lichen planus, in which degeneration of the epidermal cells and their dropping-off phe-

FIG 7. DirecL immunofluorescence re­acted with FITC-conjugated an tihuma n IgM goat serum_ Ct, Lichen amyloidosis . Th e IgM deposit ion is rather irregular a nd tends to be intense toward the pe­riphe ry of' amyloid masses, in con trast to homogeneous staining of anti keratin (Fig 2c,d). x 190. b, P rimar y systemic a my­loidosis, skin lesion_ IgM is rathe r ho­mogeneous_ x 190.

Jan. 1983

FIG 8. Lichen amyloidosis. Histochem­ical demonstration of sulfhydryl groups (a) and disulfide bonds ( b ), using OACM. a, Sulfhydryl groups a re present only in li ving layers of the epidermis (E ), not in fully keratinized horny layer (K) . S ulf11ydryl groups are absent in amyloid deposit ions (A). b, Disul fide bonds which appea l' in fully keratinized horny layer ( K ) are a lso detected in a myloid deposit ion (A). Note the a bsence of di­sulfide bonds in living layers 0 [' the epi­dermis (E ). a a nd b, X 120.

KE RATIN AND SKIN AMYLOID 71

FIG 9. P rimary systemic a myloidosis, skin lesion. Histochemical demonstra­tion of sulfhyclryl groups (a) a nd disu l­fid e bonds (b), using OACM. Neither sulfhyclryl groups nor disulfide bonds aJ'e detected in amyloid depositions (A). a, S ulf11ydryl groups exist abundantly in living layers of the epide rmis (E). b, Disu lfide bonds appeal' in keratinized horny layer (K) . a and b, X 190.

Epidermis or Tumor Nest

F IG 10. P roposed pathway of a myloid formation from keratinocy tes in skin­limited amyloidoses. This does not apply to the nodular 0 1' t umefactive amyloid of the sk in in which ant i keratin a nt ibody stain is negat.ive.

Dermis

nom en on w'e simila r to lich e noid a nd m acula r a m yloidoses [33] , m ay suggest t he impo rtance of T -Iy mphocyte accumula ­t ion in t h e upper dermis [40,41]' D e layed h y per sens itivity r e ­actio n in liche n pla nus m ay eliminate t h e degene rated ker atin­ocytes, i.e., fore ign substance to the dermis [8). On t h e ot h er h a nd, in organ -limi ted cu taneous a m y loidos is s uc h r eaction is a bsen t. It is co ns ide red that degen e rated e pide rma l cells are comple te ly phagocytized a nd digested after dropping off in to t he d ermis by massively infil tra ted cells in lic he n pla nus, while degener a ted e pidermal cells in or gan-limited cu taneous a m y ­loidosis preserve th eir fila m e n t ous features because o f a p OOl' for eign -body r ecognition by t h e hos t.

A s peculative pathway of a m yloidogenesis as discussed a bove is summarized in F ig 10.

We wish to acknowledge the help provided by 01'. Tung-Tien Sun (ant iserum) , 01' . George Ioannides (specimens), and Or. Luis CaJ'vajl (specimens).

( " Apoptosfs " ) -. Phagocytos i s by

or

1 Amyloid

REFEREN CES

I. Brownstein MH, Helwig EG: The cutaneous amyloidoses. I. Local­ized forms. Ar ch Oermatol 102:8-19, 1970

2. Brown tein MH, Helwig EG: T he cutaneous a myloidoses. II . Sys­temic forms. Arch Dermatol 102:20-28, 1970

3. Westmark P: Amyloidosis of the skin: a comparison between local­ized and systemic amyloidosis. Acta Derm Venereol (Stockh) 59:34 1-345, 1979

4. Black MM , Wilson-Jones E: Macular amyloidosis. A study of 21 cases wi th special reference to the role of the epidermis in its histogenesis. Br J Oermatol 84:199-209, 1971

5. Nagao S, Iijima S: Light and electron microscopic study of Riehl 's melanosis. Possible mode of its pigmentary incont inence. J Cut:an Pathol 1:165-175, 1974

6. Hashimoto K, Kumakiri M: Colloid-amyloid bodies in PUV A­treated human psoriatic pat.ients. J Invest Oermatol 72:70-80, 1979

7. Kuma kiri M, Hashimoto K: Histogenesis of primary localized cu­taneous amyloidosis: sequent ia l change of epidermal keratino­cytes Lo a myloid via fi lamentous degeneration. J Invest Dermatol 73: 150-162, 1979

72 KOBA Y ASH] AND HASHIMOTO

8. Hashimoto K, Kobayashi H: Histogenesis of amy loid in the skin. Am J Dermatopathol 2: 165-171, 1980

9. Hashimoto K, Kobayashi H : Amyloidogenesis in primary skin amyloidosis, Amyloid and Amyloidosis. Proceedings of the Third International Symposium on Amyloidosis. Edited by GG Glen­ner, P P ihho e Costa, A Fa lcao de Freitas. Amsterdam/ Oxford/ Princeton, Excerpta M edica, 1980, pp 426-435

10. T ezuka T , Freedberg 1M: Epidermal structural proteins. II . Isola­tion and purification of tonofilaments of the newborn rat. Biochim Biophys Acta 263:382-396, ] 972

11. Steinert PM: The extraction and characterization of bovine epider­mal a-keratin. Biochem J 149:39-48, 1975

12. Steinert PM, Gu iLlino MI: Bovine epidermal keratin filam ent as­sembly in vitro. B iochem Biophys Res Commun 70:221-227, 1976

13. Lee LD, Baden HP: Organization of the polypeptide cha ins in ma mma lian kera t in. NatUl'e 264:377-379, 1977

14. Lee LD, Baden HP, Kubilus J , Fleming BF: Immunology of epi­dermal fibrous proteins. J Invest Dermatol 67:521-525, 1976

15. S un T-T, Green H: Keratin filaments of cultured human epiderma l ce lls. Formation of intermolecular disulfide bonds during termi­na l differentia tion. J BioI Chern 253:2053- 2060, 1978

16. S un T -T, Green H: Immunolluorescent staining of keratin fib ers in cultu.red ce lls . Cell 14:469-476, 1978

17. Neville DM Jr: Molecular weight determination of protein-dodecyl sulfate complexes by gel e lectrophoresis in a discontinuous buffe r system. J BioI Chern 246:6328-6334, 1971

18. Beutner E H, Ha le WL, Nisengard RJ , Chorzelski TP, Holubar K: In, Laboratory S tudies in Immunopathology of the S kin: Labeled Antibody Stud ies, Pru·t II. Ed ited by EH Beutner, TP Chorzelski, SF Bean, RE Jordan. Stroudsberg, Pa, Dowden, Hutchinson, and Ross, 1973, pp 197-246

19. Taylor CIt Immunoperoxidase techniques. Practical and theoreti­cal aspects. Arch Pathol Lab Med 102:113-121, 1978

20. Graham RC Jr, Karnovsky MJ : The early stages of absorption of injected horseradish peroxidase in the proximal tubules of mouse kidney: ultrastructural cytochemistry by a new technique. J Histochem Cytochem 14:29] -302, 1966

21. Nakane PK: Recent progress in the peroxidase- labeled a ntibody method. Ann NY Acad Sci 254 :203-210, 1975

22. Ogawa H , Taneda A, Kanaoka Y, Sekine T: The histochemical distr ibution of protein bound sulfllydl'yl groups in human epider­mis by the new sta ining method. J Histochem Cytochem 27:942-946, 1979

23. Sun 1'-1', S hih C, Green H: Keratin cytoskeletons in epithe lia l ce lls of in ternal organs. Proc Natl Acad Sci USA 76:2813-28] 7, 1979

24. M asu S, Hosokawa M , Seiji M: Studies on cutaneous amyloidosis with anti -keratin ant ibodies. Jap J Dermatol 90:623-626, 1980

25. Husby G, S letten K , Michaelson TE, Natvig JB: Amyloid fibril protein subuni t, " prote in AS": distribution in t issue and serum in

Vol. 80, NO.1

different clinical types of amy loidosis including tha t a sociated with myelomatosis a nd Wa ldenstrom's macrog lobulinemia. Scand J Immunol 2:395-404 , 1973

26. Linke RP, Sipe JD, Pollock PS, Ignaczak TF, Glenner GG: Isolation of a low-moleculru'-weight serum component a ntigenically re lated to a n amyloid fibr il prote in of unknown origin. Proc Natl Acad Sci USA 72:1473-1476, 1975

27. Cornwell GG III , Husby G, Westermark 1', Natv ig JB, Michaelsen TE, Skogen B: Identification and characterization of differen t amyloid fibril prote ins in tissue sections. Scand J Immunol 6:1071-1080, 1977

28. Glenner, GG: Amyloid deposits and amyloidosis. N Engl J Me d 302:1333-1343, 1980.

29. M acDonald DM, Black MM , Ramnara in N : Immunofluorescence s tudies in primary loca lized cutaneous amyloidosis. Br J D er­matol 96:635-641, 1977

30. Piamphongsant T: An immunolluorescent study of cutaneous a my­loidosis. Australas J Dermatol 18:29-35, 1977

31. Habermann MC, Montenegro MIt Primary cutaneous amyloidosis: clinical, laboratoria l and histopathological study of 25 cases. Identification of gammaglobulins and C;J in th e les ions by im­munofluorescence. Dermatologica 160:240-248, 1980

32. Faille-Kuyper EH de la, Faille H de la: An immunofluorescence study of lichen planus, Br J Dermatol 90:365-371, 1974

33. Hashimoto K: Apoptosis in lichen planus a nd several other der­matoses. Intra-epiderma l cell death with fil amentous degenera­tion . Acta Derm Venereol (Stockh). 56:187-210, 1976

34. Schultz RT: Role of al te red vascu la r permeability in amyloid for­mation. Am J Pathol 86:321-342, 1977

35. Ogawa H, T aneda A: Characteristic distribu tion of sul!hydryl groups in huma n epidermal cancer by the new histochem ical s ta ining method. Arch Dermatol Res 264 :77- 81, 1979

36. Danno K, Horio T : His tochemical sta ining of sunburn cells for sulphydryl and disulphide groups: a time course study. Br J Dermatol 102:535-539, 1980

37. Wilgram GE, Kid RL, Krawczyk WS, Cole PL: Su nburn effect on keratinosomes. A report with specia l note on ul trav iole t-induced dyskeratos is. Arch Dermatol 101 :505- 519, 1970

38. Hashimoto K, Yoong Onn LL: Lichen a myloidosus. Electron mi­croscopic study of a typical case and a review. Arch Dermatol 104:648-667, 1971

39. Hashimoto K , Brownstein MH: Amyloidogenesis in healing wound. E lectron microscopic studies of biopsied wounds in macular amyloidosis. Am J Pathol 68:371-380, 1972

40. A1ario A, Ortonne JP, Schmi tt D , Thivole t J: Lichen planus: study with a nti -huma n T lymphocyte antigen (an ti-HTLA) serum on frozen tissue section~. Br J Dermato l 98:601-604, 1978

41. Bjerke JH, Krogh HK: ldent ifica tion of mononucleru' ce lls in s itu in skin les ions of lichen planus. Br J Dermatol 98:605- 610, 1978

An International Symposium on Hair and Aesthetic M edicine will be h eld in Rome on March 24-26, 1983. For

further informatio n: Dr. L. Celieno, c/o Foundation for R esearch in Dermatology, Largo A. Gemelii 8, 00168 Rome, Italy.

The Fifth Annual W estwood Western Conference on Clinical Dermatology will take place March 24-27, 1983 at the S heraton-Harbor Island in San Diego, California. The program featmes Pediatric and Adolescent Dermatology.

For information, wri te to D ermatology Educational Services, P.O. Box 4207, Kenmore, New York 14217, or 716/887-3702. .

"Practical Dermatology, " sponsored by Stanford University M e dical School, to be held March 30, 31, and April 1, 1983 at Fairchild Auditorium on the Stanford campus. For further information: Paul H. Jacobs, M.D., Department of

Dermatology, Stanford University M e dical School, Stanford, California 94305 (415/497-6101).