<<

Monoclonal Antibodies to Human Epidermal , Some Not Recognizing Profilaggrin

Michel Simon, Mireille Sebbag, Marek Haftek,* Christian Vincent, Elisabeth GiTbal-Neuhauser, Joel Rakotoarivony, Gerard Somrne;l" Daniel Schmitt,* and Guy Serre Department ofBioiOb')' and of the . Toulouse-Purpan School of Medi cin e. University of Toulouse III . Toulouse; *INSER.M U 346/CNRS, E. J-lerriot Hospital, Lyon; and "j"C io natec-Uiosoft S.A., P:u·is. France

To improve understanding of human profilaggrin AHFS and AHF6 with K14 was revealed processing to filaggrin, we produced seven monoclo­ on imtnunoblots. It is interesting that AHF7 recog­ nal antibodies against epidermal filaggrin (AHFl-7). nized filaggrin, but not profilaggrin, and labeled only They were characterized on human epidermis by the corneocyte matr~ and not the keratohyalin gran­ indirect immunofluorescence, immunogold labeling, ules. This indicates that filaggrin and and immunoblotting and found to be directed against share several antigenic determinants and that filag­ seven different epitopes of (pro)filaggrin. AHFl-5 grin bears at least one epitope absent from its pre­ labeled the k ;ratohyalin granules and the fibrous ma­ cursor. The original series of tnonoclonal antibodies trix of the lower corneocytes, and recognized filag­ described here appears to be a powerful tool for grin and profilaggrin. AHF6 also labeled the kerato­ studying human profilaggrin processing in normal hyalin granules and the corneocyte matrix, but only conditions and in the keratinization disorders in recognized filaggrin. In addition to this reactivity which processing is altered. Key words: dijfereutiatio11f within the upper epidermis, AHF4-6 stained the l111ma11 epiderm is/keratol1ya.liu gra r111les/stra tum co meum. of the basal cells, and cross-reactivity of J I1west Dermatol105:432-437, 1995

uman fila ggrin, a cationic of37,000 appar­ After fi .l ament aggregation, basic arginine residues of filaggri.n are ent molecul ar weight, is synthesized in the gran­ converted to citmllines by a pcptidylarginine deiminasc; this results ular cells of the epidermis as a large, highly in a lowe•· affinity of the molecule fo r cytokeratins (3]. Finally, phosphorylated precursor, profilaggrin , which ac­ fila ggrin degradation into free amino acids in the upper stratum cumulates in the keratohya lin granules (1.-6]. comeum seems to be necessary for the proper hydration of tlus HC haracterization of the gen e encoding human profilaggri.n [7-1 0] protective layer [16,17). In addition, it has been suggested tbat has shown that the protein consists ofl 0 to 12 filaggrin repeats, 317 fi laggrin might be cross-linked to, or at least associated with, the amino acids long, separated b y seven-amino-acid linker p eptides. cornified cell envelopes [1 8-20). N eighboring repeats display up to 30% variation in amino acid A number of keratinization di sorders (1,5,21] are associated with sequences, induciJ1 g a large heterogeneity in their isoelectric points. dys regulation in the formation of keratohyalin granules and in the T he fil aggrin repeats are flanked by truncated units fused to processing o f profilaggrin. In lamellar ichthyosis, for example, a nonfilaggrin . In particular, the amino terminal peptide is large accumula tion of profilaggrin and li ttle processing to fil aggrin homologous to calcium-binding proteins (9 ,1 0]. During the tran­ arc observed [1]. Immunologic probes able to differentiate filaggrin sition from granular to cornified cell , profilaggrin is dephosphory­ from its precursor directly on skin biopsy specimens would there­ lated, the linker peptides are protcolytically removed, and basic fore be useful diagnostic tools. Moreover, we recently reported that filaggrin units arc produced [11-13). Many questions concernn1g human fi laggrin is the target of rheumatoid-arthritis-specific the m olecular steps of profilaggrin processing into filaggrin remain autoantibodies (22]. Monoclonal antibodies (MoAbs) specific for unsolv ed, e.g., What are the e nzymes involved in the process? filaggrin would also be useful to determine the epitopes recognized Does human profilaggrin processing begin before or after the by these autoantibodies. dispersion of keratohyalin granules? T he basic f.Jagg•:in units ap­ A number of antisera and a few MoAbs have been raised against parently interact with and aggregate cytokcratin fi laments, fa cili­ human f.ila ggrin [4,23 ,24]. However, all of them r eact with both tating fonnation of intermolecular disulfide bo nds to yield the the protein and its precursor, and th e epitopcs they recognize have intracellular fibrous matri..x of the cornified ce ll s (2 ,3,6, 14 15). not been identif1cd. W e report on the characterization of a series of M.oAbs directed against purified human epidermal f.i la ggrin, includ­ ing two that recognize fi laggrin but not proftlaggrin. Manuscript received August 19, 1994: final revisio n received May 17, 1995; accepted for publi cation june 12, 1995. MATER.JALS AND METHODS Reprint requests to: Dr. Guy Serre, Laboratoire de Biologic CcUulaire et Cytologic , CHU Purpan, Pl ace du Dr Daylac, 31059 Toulo use Ccdcx, Protein Extraction aud J> urificatiou I-ILnnan fi la ggrin and profilaggrin France. were extr:l cted fTom breast skli1 (obtained from p:l ti cnts undergoing plastic Abbreviations: AI-IF. anti-hun1an fi laggrin; IIF. indirect in1Jntmo Au orcs­ surge•·y) and partia ll y purified by io n-exchange chromatography according ccncc. to Lynl ey and Dale 12]. as previo usly modifi ed 122 1.

0022-202X/95/S09.50 • SSDI0022-202X(95)00277-R • Copyright © 1995 by T he Sociery for In vestigative Dermatology, In c.

432 VOL. ·105, NO. 3 SEPTEMllER. ·1995 ANTIUODIES TO FILAGGRJN BUT NOT PROFILAGGRJN 433

Table I. Summary of the Characteristics of Anti- Filaggrin MoAbs

1 I ntmunoRuorcsccn cc ' lnlnlunoclcctron Microscopyh Immunoblotting'

N ame sc SG ss SB Hair Corneocyte Matrix KHG Fil Pro K14

AKHl + ++ ++ ++ ++ ++ AHF1 + ++ ++ , ++ ++ ++ ++ AHF2'. + ++ ++ ++ ++ ++ AHFJ' + ++ ++ ++ ++ ++ AHF4 + ++ + I - + +on...s ++ ++ ++ ++ 93-109 AHF5 + ++ + +OilS ++ ++ ++ ++ +I- AHF6 ++ + ++ ++oils ++ + ++ ++ AHF7 ++ + ++

" 11 F reactivity to the various la yers of epidermis. i.e., stratum corn eum (SC), stratum granulosum (SG), stratum spinosum (SS) , and stratum basalc (SD). and to the components of the hair fo llicle (hair), i.e. , outer root sheath (ORS) and papilla (P) . Strong reactivity is indicated by(++). moderate reactivity by(+ ). weak reactivity by(+/-). and(- ) indic;:ltes no dctcct:lblc reactivity. 6 lmmunoclcctron mi croscopy reactivity to corncoc)'tC fibrous matrix and to keratohya lin granules (KHG). r. lmmunobl ottin g reacti vity ro human epidermal filaggrin (fi l) . pro fi laggrin (pro), and cytokcratin K1 4. d Positio n of the rcacrjvc pcptidcs o n the " consensus" sequence of huma11 filaggri11 18 !. e T heir d ifferent migrations in nondcnaturing gels clca_rly differentiated AHF2 and AH F3 .

Production of MaAhs Balb/c mice were immunized by three injections at 37°C with the pep tides solubilized in 0.01 M acetate buffer, pH 5.6 (0-40 of partiall y purified human fi laggrin at 2-weck intervals. Before fusion, J.Lg/ml). After rinsing with 50 mM Tris-H C l, pH 8.0, 150 mM NaC l, and performed as originall y described b y Kohler and Milstein [25], the presence 0.05% Tween 20 (TBST), the plates were incubated w ith 1% gelatin in of anti-filaggrin antibo dies in the serum was confim1ed by immunoblotting. TBST for 0.5 h at 37°C aJ1d washed, and antibodies diluted in TBST were Hybridoma supernatants w ere screened fo r antibody production by indirect added for 2 h at 3 7°C. After further rinsing, appropriate anti-lgG conjugated immunoAu o rescence (l!F) o n cryostat sectio ns of hum an skin and by with peroxidase and the s ubstrate o rthophenylene diamine dihydrochloride immunoblotting. Seven independent hybrido ma isolates were cloned twice were added successively, as described previously (29). T he absorbance at and then u sed to produce ascites Auids in Pristane-primed mice. Immuno­ 492 tun was measured with a Un.iskan li photometer (Labsystems France, globulins (Igs) were purified by ammo nium sulfate precipitation and Les U lis, France). A U assays were performed in duplicate and the resul ts chromatography o n diethylamiuoethyl T risacryl LS (IBF, Vill eneu ve Ia averaged. Garenne, France). T he MoAbs w ere all of the lgG1 isotype, as determined by MonoAb-ID (Tago, B urlingame, CA) enzyme-linked immunosorbent RESULTS assay (ELISA). To produce anti-ftlaggrin MoAbs, we use d basic filagg ri.n of human Other Immunologic Reagents AKJ-11 , a MoAb specific for human e pidennis as a.n immunogen . T h e seven independe nt clones w e profil aggrin and fila ggrin, was purchased from Biomedical Technologies, d escribe h ere were d e rived lnc. (Stoughton, MA). Aoti-h.K1 4, a rabbit antiserum specific for human from a single fusion and called AHFl to cytokeratin 10 4 (26]. was a generous gift from E. Fuchs (H oward Hughes AHF7 for th eir nnti-/ruman .fi laggri.n reactivity (initially desig n ated Medical Institute, University of C hicago, Chicago, lL). BL6 MoAb (Immu­ lOSe, 621b, 11 05b, 812a, 1117a, 134j, and 711c, respectively) . ootech, M arseill e, France), specifi c fo r a C I 1 a antigen expressed b y T h e ir reactivities with huma n e pide rmis, e pidermal protein s, a nd epidcrn1al Lan- gerhans cells (27], w as used as a negative control in synth etic p e ptides are d escribed in d etail b elow and summarized in i.Jnrn unoclectron microscopy. Table I. IIF IlF was perfom1cd on 4-f.L111-tltick unfixed cryostat sections and C m10y's­ AHF7 Does Not Label the Keratohyalin Granules ofHuman fixed, paraffin-embedded sections of hw11aJ1 sk.-in, as described pre \~o u s l y (28]. Control experiment• omitting the p1imary :mtibody or u sing IgG from a nonim­ Epidermis T h e reactivity of the AHF MoAbs was first tested by mune 1nouse were always negative. IIF on cryosection s ofhw11 an skin. T h eir p attems of reactivity were compared with th at produced by AKHl, a well-c haracterized and Inununoelectron Microscopy Normal human skin was analyzed with widely use d anti-(pro)filaggrin MoAb (Fig 1). T hree of the AHFs post- embedding indirect immunogold labeling, as described earli er (20,28]. (AHFl-3) showe d strong granu lar cytoplasmic s taining Fragments of abdominal skin w ere .fixed in 3'!1,, paraformaldehyde, deh y­ of th e drated at low temperature iJ1 ethanol, cryosubstituted, and embedded at stratum granulosum a nd diffuse staining of th e lower stratum -35°C in Lowic.ryl K4M (Chemischc Werke Lowi GMBH , Waldkmiburg, corneum. The p atte rns of reactiv ity of th ese three MoAbs were Germany). U ltrathin sections were preincubated for 20 min with 2.5% fetal indisting uish able from th at of AKl-Il. AHF4 a nd AHFS also showed bovine serum in phosphate-buffered saline, then incubated for 1 h with stron g g ra nular staining of the stratum gra.nulosum and diffuse MoAb-containing ascites Auids diluted to 1:50 in 0.1 % fetal bovine serum. staining of the lower stratum corne um, but they al so weakly stained Mter washes, the sections were incubated for 1 h with goat lg to mo use IgG the cytoplasm of th e basal cells. AH F4 was also observed to lab el G-1 0 coll oidal gold conjugate, diluted to 1 :1 0 (Amcrsham international, the cytoplasm of the spinous cells irregularly from one skin frag­ Aylesbury, UK), washed again, and counterstained w ith uranyl acetate in m e n t to a nothe r. AHF6 produced weaker granul a r staining of the methanol for 5 min. stratum g ra nuJosum but intense staining throughout the e n tire Protein Analysis Proteins were an alyzed b y sodium dodecylsulfate­ stratum corne um up to the d esquamating com eocytes. It also polyacrylamide gel electrophoresis (SDS-PAGE) on 12.5% acrylamide gel stron g ly labe .l ed the cytoplasm of th e basal cell s. T h e staining of the or by two-dimensio nal el ectrophoresis in the presence of urea, using the stratum gra.nulosum was consistent with the distribution o f k e rato­ PhastSystem, as described by Pharmacia LJCB (Uppsala, Sweden) . Protein h yalin g ranules in every case. T h e seventh MoAb, AHF7, did not markers from Bio-Rad La boratori es (Richmond, CA) were used a s molec­ stain the viable cells, includin ular weight references, and the isoelectri c point gradient profiles were g the g ranular cells, but lab eled all th e obtained b y the Broad pi Calibration Kit (Pharmacia LK.B) rnn in parallel. cornifie d cell layers. Mter electrophoresis, the proteins we re elcctrotransferred to nitrocellulose On sections of Camoy's-fixed paraffin-embedded huma n skin, membranes and probed, as reported previo usly [20], with MoAbs diluted to th e AHF MoAbs sh owed th e p attern of reactivity observed o n 5 J.Lg/ml. T he M oAb-reactive proteins were visualized with peroxidase­ cryosections except that a U of th e antibo dies labele d only the lowe r conjugated sheep antibodies to mouse lgG (Zymed, San Francisco, CA). part of the stratum con1eum, a nd AHF4 and AHF5 no lo n ger ELISA Peptides were syn thesized by Neosystem Laboratories (Stras­ la b eled the cytop lasm of the b asal cell s. T h e absence of reactivity of bou.rg, France) using an o ptimized sol.i d-phase procedure, and their purity AHF7 on th e stra tum granulosum was confirmed (Fig 1). w as checked by reverse phase ch romatography. W e lls of polystyrene No differences were observed in th e · pattems of reactiv ity microtitration pl ates (Nunc, C openhagen, Denmark) were coated overnight between the inte rfollic ula r a nd intrato llic ular e pide nnis on c ryo- 434 SIMON ET AL T H E JOUil.NAL OF INVESTI GATIVE DERMATOLOGY

AHF6 and AHF-7 React With Filaggrin But Not With Profilaggrin on Immunoblots Analyzed by immunoblotting o n epidermal proteins, aU seven of the AHF MoAbs reacted with a doublet of approximate molecular weight 37,000 (Fig 3). All but two (AHF6 and AHF7) also reacted with a high-molecular-weight protein. Because the same bands were also detected by AK.Hl, these proteins were identified as fila ggrin and profilaggrin, respec­ tively. In addition, AHF6 recognized a protein of 48 kD that was also w eakly stained by AHFS and comigrated with cytokeratin K14. T he binding to the 48- k.D protein was due to cross-reactivity, as it disappeared after preabsorption of the MoAbs on purified filaggrin together with the anti- fi.laggrin reactivity (not shown). To con.finn these results, w e analyzed filaggrin and profilaggrin by immune­ blotting after partial purification by a nion exchange chromato­ graphy. All MoAbs reacted with the bas ic fi laggrin (flow-through fractions), and aU but AHF6 and AHF7 also reacted with profilag­ grin and its proteolyzed fragments (data not shown). T he identity of the recognized proteins was confirmed by two-dimensional gel analysis (Fig 4). Once again, the seven MoAbs reacted with the cationic filaggrin. Five of them (AHF1-5) recog­ nized the anionic profilaggrin, whereas AHF6 and AHF7 did not (or very faintly for AHF6). Al-TF6 ;1l so strongly stained the 48-k.D protein that was identified as cytokerati.t1 K14, as it reacted with an anti- human K14-specific antiserum. T he cytokeratin K14 was also weakly stained by AHFS.

AHF4 Reacts With the Filaggrin Region 93-109 When Analyzed by ELISA To characterize further the AHF MoAbs as well as AK.Hl, we tested their reactivity by ELISA with 32 synthetic peptides 14 to 16 amino acids long, de rived fi·om the published " consensus" sequence of human filaggrin [8]. These peptides presented four- to five-amino-acid overlaps, and their combined sequence covered the entire sequence of one filaggrin unit without its seven-amino-acid linker. In a preliminary experiment, tl1e antibodies have been shown to react in ELISA with purified human epidermal filaggrin. Howe ver , when the synthetic peptides were used, only AHF4 presented a sig>1ifi cant reactivity to one of them Figure 1. AHFl-6 label the stratum granulosum and stratum (GTSGSRSASRQTRNQEQ), corresponding to the filaggrin re­ corneum of normal human epidermis, whereas AHF7 labels only the stratum cornetnn. AKHl and the AI-I F MoAbs (as shown on the gion 93-109. plates) were ;o nalyzed by II F on cryoscctions of hum;on ski n and on sections ofCarnoy's-fixed slcin (a sterisk). Note that AHF7 docs not label the stratum DISCUSSION granulosum but strongly labels the lower stratum corneum when tested on We have characterized seven AHF MoAbs produced against basic as well as on sections of Carney's-fixed ski n un fixed cryoscctiom (AHF7) filaggrin of human epidermis. Various patterns of IIF staining, (AH F7*). An·o•vs indicate some of the uppe r nucleated cells of the stratum immunoblotting, and ELISA reactivities (summarized i.t1 Table I) gra mLl osum; mTOIIJheads show the surface of the epidern1is. The s tratum corneum (SC) and the stratum granulosum (SG) are indicated (brackets). indicate that they reco gnize diffe rent epitopes on human fila ggrin. Dashed fi11 es denote the derm oepid crmal junction. Bar, 50 p.m. Even if AHF2 and AHF3 did not differ w ith respect to th ese characteri stics and therefore might recognize the same or a closely related epitope, their migration in nondenaturi.t1g gels clearly difFerentiated them and confirmed that they do not originate from sections of scalp skin. T he three MoAbs (AHF4-6) that recognized the sam e antibody- producing cell. Moreover, on.ly AHF2 recog­ the basa l cells of the epidermis stain ed the cel.l s of the outer root nized a recombinant human filaggrin unit on immunoblots (pre­ sheath of the hair fo ll icles, and, curiously, AHF1 gave clear liminary result). cytoplas mic labeling of cell s of the hair follicle papilla (data not Six M oAbs (AHF1-6) labeled the keratohyalin granules of the shown). granular cell s and the fibrous matrix of the cornified cel.l s. When To detem1ine the ultrastructural location of the anti gen(s) rec­ analyzed by immunoblotting, all of these MoAbs except AHF6 ognized in the upper layers of human epidermis, we analyzed the recognized filaggri.t1 and profilagtp·in. Surprisingly, AHF6 recog­ AHF M oAbs by post-embedding immunoelectron microscopy after nized fi.la ggrin but not (or very faintly) pro fiJaggrin. It may therefore paraformaldehyde fi xation (Fig 2). In the stratum granulosum, discri.minate discrete steps in the maturation of profiJaggrin to fil aggrin keratohyalin granules were clearly visible as polygonaJ structures that occur already in the keratohyalin granuJes. Alternatively, the showing a microgranular texture typical of the embedding and profilaggrin epitope recognized by AHF6 may have been denatured counterstaining methods used . T hey showed typical reactivity with after SDS treatment. In tlus event, because AHF6 labels basic filaggrin AK.Hl (Fig 2a) and were also strongly labe led b y five of the AHF w1der the sa me immunoblotting conditions, the epitope it recognizes MoAbs, namely AHF1-5 (Fig 2/J,c). With AHF6, the labeling must be modified i.t1 some way and therefore less resistant to SDS intensity of the granules was weaker than that of the lower stratum denaturation in the precursor molecule. Consistent with this asswnp­ corneum (F ig 2d). In contrast, the keratohyalin g ranules were tion is the fact that AHF6 reacted less strongly than the other fiv e unJabeled by AHF7 (Fig 2e). ln the stratum corneum, immunogold MoAbs with the keratohyalin granules and thus may show a lesser labeling obtained w ith the seven AHF M oAbs was o bserved affm.ity for tl1e precursor than for the mature protein. AHF7, which throughout the i.n tracel.lular m atrix of three to seven of the lower selectively recognized fi.la ggri.n but not prof!laggrin on U1m1w1oblots, co rnified cel.l layers; the label.in g in tensity decreased toward the stained onJy the com eocyte fibrous matrix and did not label the surface and became undetectable in the upper com eocytes (Fig 2c). keratohyaJin granules. The epitope with which AHF7 reacts nught be VOL. 105. NO. 3 SEPTEMBER. 1995 ANTIBODIES TO FILAGGilJN BUT NOT PR.OFILAGGR.IN 435

Figure 2. Immunoelectron microscopy anal ysis reveals t he absence of reactivity of AHF7 with keratohyalin granules. NonnaJ h uman skin was analyzed by post-embedding in direct immunogold labeling. as described in Mnterin ls nud Methods. \ XI hen compared with the reactivity of AKI-1 l (n), five of the A I-I F MoAbs (AHF1-5) showed a similar labeling pattcm of the keratohyalin granules (s 111 nll nmll/ls) and of the fibrous intracellular matrix of the lower cornified cells. T his typica l pattem is shown for AHF2 (/>,c). Note the d isappearance oflabeling in the upper com ificd cell layers 5 and 6 (r). In contrast. AHF6 shows weaker reactivity o n the keratohyalin granules than o n the matrix of the lower corneocytes (d) and AHF7 proves unreactive on these granules, w hereas it labels the coml!ocytc matrix (c). SC. stratum corneum; SG. stratum granulosum. Some dl!smosom cs arc in dicated (opeu fliTOJOs). Bnrs. I IJ.IIl. 436 SIMON ET AL THE JOURNAL OF INVESTIGAT IVE DER.MATOLOGY

absent on profilaggrin because of the high degree of phosphorylation or because of a different conformation of this protein. Alternatively, AHF7 may react with a citrulline-containing epitope only present on the post-translationally modified filaggrin. Indeed, such a modification 200- that results in subsequent degradation of filaggrin is known to occur after the aggregation of cytokeratin filaments. Similarly, the rheuma­ , 16_ toid-arthritis-specifi.c anti-filaggrin autoantibodies stain d1e comeocyte 97- matrix but not the keratohyalin granules of human epidem1al kerati­ filaggrin but not proftlaggrin [22,30]. Tlus 66- nocytes, and recognize result confirms the existence of filaggrin-specific epitopes absent from 43- profilaggrin. In addition to the staining of keratohyalin granules and com eo­ cyte matrix, AHF4-6 reacted with the cytoplasm of the basal cells. In agreement, AHFS and AHF6 cross-reacted with cytokeratin K14 on immunoblots. Although K14 is present throughout the living layers of epidennis [26,31]. it was detected by AHFS and AHF6 only in the basal cells. SuGh a result suggests that d1e antigenic sites recognized by these MoAbs on K14 are masked i11 sit11 in the suprabasal cells. This was previously shown to be the case for Figure 3. AHFl-5 recognize human epidermal filaggrin and pro­ the monoclonal anti-type I cytokeratin antibody AEl [31]. On the filaggrin, whereas AHF6 and AHF7 only recognize filaggrin. Pro­ other hand, it is known that human filaggrin and the suprabasal teins of human epidermis extracted with 6 M urea were separated by cytokeratins Kl, K2, and KlO share an epitope that is defined b y SDS-PAGE and stained with Coomassie blue (CB) or analyzed by immu­ AE2 MoAb. This antibody (raised against epidermal cytokeratins) noblotting with AKHl , with the AHF MoAbs, and with a control MoAb 3 labels the suprabasal cells of the epidermis a nd cross-reacts in positions of molecular weight standards (X 1 ) (CO), as indicated. The o- immunoblotting with fuaggrin.:j: Our results show that human are indicated (lift), and proteins of interest are shown (right). P, profilaggrin; filaggrin and K14, although very different from a biochemical point F, filaggrin; K, cytokeratim. of view, nevertheless share two antigenic determinants. Whether this immunologic relationship is related to the filaggrin-cytokeratin NEPHGE interaction is not known. Although AHF4 did not detect cytokera­ tin on immunoblots, it is tempting to speculate about its labew1g of the basal cell s by IIF; it appears to cross-react with a conformational e or discontinuous epitope of cytokeratin K14 (or KS) that is lost after AHF1 AHF2 p I L l p SDS denaturation. 200- s Among the eight MoAbs used in this study, AHF4 was the only 116- in ELISA to bind a synthetic peptide (93-109) derived "­ D one shown .. _ from the "consensus" sequence of filaggrin. Tlus result suggests s that most of the epitopes become inaccessible after interaction of F the relevant peptide with polystyrene, or, more likely, that the unreactive MoAbs recognize discontinuous or conformation-de­ pendent epitopes, as is commonly observed [3 2]. Longer peptides or recombinant molecules must be tested to defme the filaggrin region recognized by the anti-filaggrin MoAbs. Nevertheless, it . :' p AHF4 p AHF7 should be noted that a rabbit anti-filaggrin antiserum we produced 200- ,,. - as a positive control also bound to peptides (27-46 and 62-76) "- corresponding to the amino terminal tlurd of the filaggrin. More experiments will be necessary to test whether this part of the .. - molecule is immu11odominant. 43- F F Some of the MoAbs described here immunoprecipitate native filaggrin (M. Simon, unpublished data) and react with the dena­ tured protein on immunoblots. They will therefore be useful tools to probe the structure and function of filaggrin. Moreover, four of the AHF MoAbs show no cross-reactivity with cytokeratins, label Camoy's-fixed, paraffin-embedded tissues, and can be used in p K14 post-embedding immunoelectron microscopy studies. These prop­ erties make them potentially useful for lustopathologic and diag­ nostic purposes, especially AHF7, which does not recognize kera­ tohyalin granules and profilaggrin. F

1-Ve th nuk Professor M. Cos fnglioln (Scnlice de C hir111gie Plnstique, CHU Rnugueil, Figure 4. lmmunoblotting analysis after two-dimensional gel elec­ Toulouse) )in· prol'idiug us ll'ith /uu1wu ski11. The tccl111icn l nssislnuce of M. -F. /soia , trophoresis confirms d1e specificity of the AHF MoAbs. Proteins of M. -P. Rue, aud M.-H. Teillou is grn t ~fully nckuoll'lcdged. We ore iudcbted to Dr. separated by nonequilib­ human epidermis extracted with 6 M urea were E. Fucl1s (Holl'nrd Hughes Medical lustitute, Uui iJers if)' of C l1icngo, Chica,~o , rium pH gel electrophoresis (NEPHGE) in the first dimension and by Jlliu ois) forller geuero us gift

supported by grn llfs j im 11 rit e INSER f\11 (C. R. E. 93 0205), ji-o 11t rit e "Associnrio11 16. Scott IR. Ho1rd ing R . Darrerr JG: H istidine-rich protein of the keratohyalin pour Ia R cclt erclr c s11r Ia Po/ym1l11ire," a11d j im11 rit e "R egio11 Midi-Pyre11 ees." granules. Source of Lh e free amino acids, urocani c acid and pyrro l.i donc M . S imoJI fl'ns supported b)' n.fdlmllshipJi·o m the 11A l·socialion de Secours des A mis carboxyli c acid i.n the stratum corneum. Bior.ltiw Biot'h )ls II cut 7 19: 110-11 7. des Scie 11 ccs." 1982 17. R awlings AV. Scott lA. H ard in g C R. . Bowser PA: Straturn con1etnn moistur­ ization ar the mo lecular level. J 1111/l!St Dcn11 ntol 103:73.1-740. 191)4 18. Richards S. Scott IR . H arding C it . LiddciiJE. Powe R.EF.EREN C.ES ll GM. C urtis CG: Evidence fo r fi laggrin as a component of rhc cell envelope of the newbon1 rat. 13iorll em 1.. Dale .B A. Rcsin g KA. Haydock PV : Fil aggrius. In: Goldman RD, S te inert PM J 253: 153-1 60. 1988 (eds:.). 1\1/cJit!C ulnr mtd Ccllult~ r BiohJ,{! )' •!f lutcrmeciinte Filruu cuts. Plenum Press , New 19. Ala.s dair CS. Steinert PM: Protein composition of cornified cell envelopes of York, 1990, pp 393-41 2 epidennal keratinocytcs. J Cc:/1 S.:i I 07:693-700. 1994 2. LynJ cy AM. Dal e BA: T he charactcri zarion of humau epidermal fiJaggrin, a 20. Haftc k M , Serre G, Mils V. T h.i volet J: Immunocytochemical evidence fo r a histidine-rich, filament-aggregating protein. JJioclliw Biopltys Acta 744 : possible ro le o f cross-linked k cratinocyte e n velopes in stranun corneum 28- 35. 1983 cohesio n. J Histod11·m C)'tochcm 39:1 53 1- 1538. 1991 3. Hardi ng C H... , Scott IR: Histidine-rich proteins (fi_l aggrins): structural and func­ 2 1. Syberl V. D:ll c l3 A, H olbrook K: Ichthyosis v ulgari s: identification of a defect in tional heterogcneiry during epidermal differentiation. J Mol Bioi J 70:65 1-673, synthesis of filaggrin correlated with an absence of kerato hya lin granu_les. J 1983 !tt l!t!SI Dt!I'I/Ulto/ 84:191-194. 1985 4. Flcck:m:m P, Dale BA. H o lbrook KA : Pro fil aggrin, a high-mo lecul ar-weigh t 22. Simon M , Girbal E. Sebbag M. GomCs- Daudrix V. Vincent C. S01l anw G. Serre precursor o f fal :~ggrin in human epidermis and cultured k cratinocytcs. J lmn~s t G: The cytokerati n fi lamcllt.-ilgg:rcgating protein fal ::aggrin is the target of the Derm al"/ 85:507-5 12. 1985 so-ca ll ed ··antikcra ti n an ti bodies,'' autoantibodies specific fo r rheumatoid 5. Man abe M , Sanchez M. Sun ~IT. Dale BA: Interaction of fi laggrin w ith keratin arthritis. ) C /i11 lu vcsr 92:'1387-1393. 1993 fi laments during : 1dvanccd stages of normal human cpidcn11 al differen ti ation 23. Dale BA, Gown AM . Fleckman P. Ki mball JR. R esing KA: Characterization of and in ichthyosis vul D gari s. i.fft'rt'lllintiou 48:43- 50. 199 1 two monoclonal antibodies to human epidcrm:1l kerato hya lin: rc:-~crivi ty w ith 6. Giin zcl S. Wcidcnthnlcr B. H auBe r I, Anton- Lamprccht J: Keratohyalin g ranules fi la ggrin and rela ted pro teins. ) ltllllW Denuatol 88:306-3'13. 1987 arc heterogene ous in ridged and no n-ridged human skin: evidence from 24. Kubilus j , Scott I, Harding C R . Ycndle J, K vedar J , Baden HI': T he occurrence anti-filnggrin itnmunogold labeling of no rmal skin and skin of auto soma.l o f pro lilaggrin and irs processing in cultured kcr:1tinocyres. J ltlllt'St Dcmwtol dominant ichthyosis vulg:1ris patients. llrclt Dcrmntol R es 283:421-432, 1 991 85:5 13-5'17 . 1985 7. M cKjnley-Gr:mt LJ , Idler Bernstein lA, Parry D AD , C an nizzaro L, C roce w·w. 25. Kohler GT, Milstein C: Cominuous cultures of fu sed cells secreting antibody of C M, Hue bner K . Lessin SR., Steinert PM: C h:m1 cterizatio n of a eDNA clo ne predefined specificity. Nllwrc (J..mul) 256:459-497. 1975 encoding human fi lo1ggrin and locali zati on of the gene to c hromosome region 26. Stoler A. Kopan R . Duvic M. F uchs E: Usc of monospecific antisera :md cRNA 1q2 1. l'roc Nat/ Awd Sci USA 86:4848- 4852, 1989 probes to locali ze the maj or changes in keratin expression duri ng normal and 8 . Gan S-Q . McBride OW. Idler WW, Markova N, Steinert PM: O rganization, abnormal e pidermal difle rcn tiation. J Cell Bioi 107:427-446. 1988 structure and polymorphisms o f the human profila ggrin gene. Biochcmistr)' 27. Yo n_ish-Ro uach E. Schmitt D. Viae KJ1 0 wlcs R . Cordier G. 29:9432-944 0, 1990 J. Brochier J: Monoclonal anti-thymic cell antibodies detectin g cpidennal cell s. Tlt)11111 lS 9. Prcslnnd RJ3. Haydock PV. Flcckman P. Nirunsukrisi W , Dale l3 A: C haracter­ 6:67-78. '1984 izari on o f the human epidermal pro fihtggri11 gene. J Bioi C lwm 267:23772- 28 2378 1. 1992 . Serre G. Mils V. Haftek M. Vincent C. C roure F . Rcan o A. O uhayoun J- P, llettin g:er S. SolciJh avoup .J - P: Identificati I 0. Markova NG, Marekov LN . C hipev CC. Gan S-Q. Idler W W . Steinert PM: on of larc d iRCrcntiation a_ntigcns of human cornified epithelia, expresse Profilaggri11 is a majo r cpidcnnal calcium-binding protein. !Hoi Cell Bioi d in re-organized d esomosorn cs :md bound 13:613- 625, 1993 10 cross-linked envelope. ) lu ~>csr Dmuato/ 97:'106 1- 1072. 199 1 11. Kashima M, Fukuyam:1 K, Kikuchi M, E pstein W L: Limited pro teolysis of high 29. Serre G. Vincent C , Viraben R. Soleilhavoup J-1': Natura l lgM and lgG mo lecular weigh t 11i stidinc-ricll protein o f ra t epidermis by epidermal pro tcin­ autoantibod ies to epidermal kc r<~tin s iu normal human ser:L 1: ELI SA titration. ascs. J i11 11rst Dmtwtol 90:829-833, 1988 irnmunoftu oresccncc study. ) lm11'S f Dcnunto/ 88:2 1-27. 1987 12. H.esing KA. \Valsh KA. 1-laugen-Scofi eldj. Dale BA: Identificatio n of pro teolytic 30. Simon M, Vincent C. 1-l aftek M. G irbal E. Sebbag M. Gomi:s- Daudrix V. Serre cleavage sites in the conversion o f profib ggrin to filaggrin in mammalian G: The r heumatoid arthriris-;usociated autoanti bodies to filaggrin label the epidermis. ) Bioi /11"" 264:1837-1845, 1989 fibro us matri..x of the cornified cells b ur not the profilaggrin-containing 13. Dale BA, Presland IUJ. Fleckman P. Kam E. Resing KA : Phenotypic expression kerato hya lin gr:u1ulcs in l1um;1n epidermis. Cli11 Exp lmmuuol I 00:90-98. 1995 and processin g: o f fila ggrin in epidermal d i ll C rc nti:~ti o n . In : D :.t nno n M , 31. W oodcock- Mirc hcliJ, Eichnc.r R , Nelson WG. Sun T-T: lnununolocalization of Blumenberg M ( eds. ). A-lolecular Biolo.. ~:: y of til e Skill: Tile Kt•rnliuocyte. Academic keratin polypeptides in human epidermis using mo nod o nal antibodies. J Cell Press. San Diego. CA. 1993, pp 79-1() (o Bioi 95:580-588. '1982 14. Dale BA, H o lbrook KA. Steincn PM: Assembly o f stratum corneum basic protein 32. 13 cnj amiu DC, B cr zof.~ k y J A. East IJ , Gurd 1:: R N. H annum C. Leach SJ, and keratin fil aments in macro fibrils. Norurl' (Lcmd) 276:729-73 1. 1978 Margoli as h E. Michael J C. Miller A. Prager E, R.eichlin M, Scrcarz EE, 15. M ack JW. Steven AC, Stein ert PM: T he mechanism of interacti on of fil aggrin Smith-Gill SJ . Todd PE, W il son AC: T he a n tigenic structure of proteins: a with i ntermediate fi laments. J A1ol Bioi 232:50- 66, '1993 reappraisal. A 111111 R cJJ JumllltWI 2:67-10 1. 1984