Epiplakin, a Novel Member of the Plakin Family Originally Identified as a 450-kDa Human Epidermal Autoantigen

Abstract

A 450-kDa human epidermal autoantigen was originally identified as a protein that reacted with the serum from an individual with a subepidermal blistering disease. Molecular cloning of this protein has now shown that it contains 5065 amino acids and has a molecular mass of 552 kDa. As reported previously this protein, which we call epiplakin, belongs to the plakin family, but it has some very unusual features. Epiplakin has 13 domains that are homologous to the B domain in the COOH-terminal region of desmoplakin. The last five of these B domains, together with their associated linker regions, are particularly strongly conserved. However, epiplakin lacks a coiled-coil rod domain and an amino-terminal domain, both of which are found in all other known members of the plakin family. Furthermore, no dimerization motif was found in the sequence. Thus, it is likely that epiplakin exists in vivo as a single-chain structure. Epitope mapping experiments showed that the original patient's serum recognized a sequence unique to epiplakin, which was not found in plectin. Immunofluorescence staining revealed the presence of epiplakin in whole sheets of epidermis and esophagus, in glandular cells of eccrine sweat and parotid glands and in mucous epithelial cells in the stomach and colon. A 450-kDa human epidermal autoantigen was originally identified as a protein that reacted with the serum from an individual with a subepidermal blistering disease. Molecular cloning of this protein has now shown that it contains 5065 amino acids and has a molecular mass of 552 kDa. As reported previously this protein, which we call epiplakin, belongs to the plakin family, but it has some very unusual features. Epiplakin has 13 domains that are homologous to the B domain in the COOH-terminal region of desmoplakin. The last five of these B domains, together with their associated linker regions, are particularly strongly conserved. However, epiplakin lacks a coiled-coil rod domain and an amino-terminal domain, both of which are found in all other known members of the plakin family. Furthermore, no dimerization motif was found in the sequence. Thus, it is likely that epiplakin exists in vivo as a single-chain structure. Epitope mapping experiments showed that the original patient's serum recognized a sequence unique to epiplakin, which was not found in plectin. Immunofluorescence staining revealed the presence of epiplakin in whole sheets of epidermis and esophagus, in glandular cells of eccrine sweat and parotid glands and in mucous epithelial cells in the stomach and colon. Clarification of the basic structure of desmoplakin has been followed by the identification of many related proteins, such as BPAG1,1 plectin, envoplakin, and periplakin (1Green K.J. Parry D.A.D. Steinert P.M. Virata M.L.A. Wagner R.M. Angst B.D. Nilles L.A. J. Biol. Chem... 1990; 265: 2603-2612Google Scholar, 2Tanaka T. Parry D.A.D. Klaus-Kovtun V. Steinert P.M. Stanley J. J. Biol. Chem... 1991; 266: 12555-12559Google Scholar, 3Sawamura D. Li K. Chu M.L. Uitto J. J. Biol. Chem... 1991; 266: 17784-17790Google Scholar, 4Wiche G. Becker B. Luber K. Weitzer G. Castanon M.J. Hauptmann R. Stratowa C. Stewart M. J. Cell Biol... 1991; 114: 83-99Google Scholar, 5McLean W.H.I. Pulkkinen L. Smith F.J.D. Rugg E.L. Lane E.B. Bullrich F. Burgeson R.E. Amano S. Hudson D.L. Owaribe K. McGrath J.A. McMillan J.R. Eady R.A.J. Leigh I.M. Christiano A.M. Uitto J. Genes Dev... 1996; 10: 1724-1735Google Scholar, 6Ruhrberg C. Hajibagheri M.A.N. Simon M. Dooly T.P. Watt F.M. J. Cell Biol... 1996; 134: 715-729Google Scholar, 7Ruhrberg C. Hajibagheri M.A.N. Parry D.A.D. Watt F.M. J. Cell Biol... 1997; 139: 1835-1849Google Scholar). These proteins form a family known as the “plakin family” (8Uitto J. Pulkkinen L. Smith F.J.D. McLean W.H.I. Exp. Dermatol... 1996; 5: 237-246Google Scholar). Almost all members of this family have a common structure, with predicted globular amino-terminal and COOH-terminal domains that are separated by a central rod domain. Some homologous domain structures have been identified in both globular domains of many plakins, while the central domain is rich in heptad repeats and is believed to form a parallel α-helical coiled-coil structure with a dimerization partner (9Green K.J. Virata M.L.A. Elgart G.W. Stanley J. Parry D.A.D. Int. J. Biol. Macromol... 1992; 14: 145-153Google Scholar). As suggested by this model, it has been demonstrated that desmoplakin I can form homodimersin vitro (10O'Keefe E.J. Erickson H.P. Bennett V. J. Biol. Chem... 1989; 264: 8310-8318Google Scholar). Early investigations revealed that the COOH-terminal domains of plakins are involved in binding to intermediate filaments (11Stappenbeck T.S. Bornslaeger E.A. Corcoran C.M. Luu H.H. Virata M.L.A. Green K.J. J. Cell Biol... 1993; 123: 691-705Google Scholar, 12Wiche G. Gromov D. Donovan A. Castanon M.J. Fuchs E. J. Cell Biol... 1993; 121: 607-619Google Scholar, 13Nikolic B. Nulty E.M. Mir B. Wiche G. J. Cell Biol... 1996; 134: 1455-1467Google Scholar). The amino-terminal domains of desmoplakin and of BPAG1 are believed to bind to desmosomes or hemidesmosomes. Furthermore, some splicing variants of plectin and BPAG1 have actin- or microtubule-binding domains at their amino termini, and it has been proposed that these domains form cross-links between microfilaments and/or microtubules and intermediate filaments (14Yang Y. Dowling J., Yu, Q.C. Kouklis P. Cleveland D.W. Fuchs E. Cell.. 1996; 86: 655-665Google Scholar, 15Yang Y. Bauer C. Strasser G. Wollman R. Julien J.P. Fuchs E. Cell.. 1998; 98: 229-238Google Scholar, 16Fuchs P. Zoerer M. Rezniczek G.A. Spazierer D. Oehler S. Castanon M.J. Hauptmann R. Wiche G. Hum. Mol. Genet... 1999; 8: 2461-2472Google Scholar). Studies of a few inheritable diseases that appear to involve plectin or desmoplakin and of a BPAG1 null mouse have shown that each plakin plays a critical role in the tissue integrity in specific tissues (5McLean W.H.I. Pulkkinen L. Smith F.J.D. Rugg E.L. Lane E.B. Bullrich F. Burgeson R.E. Amano S. Hudson D.L. Owaribe K. McGrath J.A. McMillan J.R. Eady R.A.J. Leigh I.M. Christiano A.M. Uitto J. Genes Dev... 1996; 10: 1724-1735Google Scholar, 17Gache Y. Chavanas S. Lacour J.P. Wiche G. Owaribe K. Meneguzzi G. Ortonne J.P. J. Clin. Invest... 1996; 97: 2289-2298Google Scholar, 18Armstrong D.K.B. Mckenna K.E. Purkis P.E. Green K.J. Eady R.A.J. Leigh I.M. Hughes A.E. Hum. Mol. Genet... 1999; 8: 143-148Google Scholar, 19Guo L. Degenstein L. Dowling J., Yu, Q.C. Wollmann R. Perman B. Fuchs E. Cell.. 1995; 81: 233-243Google Scholar). Moreover, it seems likely that, in many autoimmune blistering diseases, plakins, located in the epidermis, might be target antigens, and these plakins are used for markers of specific diseases (2Tanaka T. Parry D.A.D. Klaus-Kovtun V. Steinert P.M. Stanley J. J. Biol. Chem... 1991; 266: 12555-12559Google Scholar, 20Oursler J.R. Labib R.S. Ariss-Abdo L. Burke T. O'Keefe E.J. Anhalt G.J. J. Clin. Invest... 1992; 89: 1775-1782Google Scholar, 21Kim S.C. Kwon Y.D. Lee I.J. Chang S.N. Lee T.G. J. Invest. Dermatol... 1997; 109: 365-369Google Scholar, 22Proby C. Fujii Y. Owaribe K. Nishikawa T. Amagai M. J. Invest. Dermatol... 1999; 112: 153-156Google Scholar). However, their pathological roles remain to be clarified. Several years ago we described an individual with a subepidermal blistering disease that resembled bullous pemphigoid both clinically and pathologically (23Fujiwara S. Shinkai H. Takayasu S. Owaribe K. Tsukita S. Kageshita T. J. Dermatol... 1992; 19: 610-613Google Scholar). Immunoblot analysis revealed that the patient's serum did not react with the 230-kDa (BPAG1) and 180-kDa bullous pemphigoid antigens, whereas it did recognize a 450-kDa epidermal polypeptide. This polypeptide was expressed in human keratinocytes and in some transformed cell lines that included HeLa, KB, and A431 cells (24Fujiwara S. Kohno K. Iwamatsu A. Shinkai H. Dermatology.. 1994; 189: 120-122Google Scholar). In a preliminary study, to determine the molecular structure of this antigen, we screened a human keratinocyte cDNA library with the patient's serum. We isolated two kinds of cDNA: one encoding a protein that was strongly homologous to rat plectin and another encoding a protein, with partial homology to plectin, which appeared to be a novel and previously unidentified protein (25Fujiwara S. Kohno K. Iwamatsu A. Naito I. Shinkai H. J. Invest. Dermatol... 1996; 106: 1125-1130Google Scholar). We report here the cloning of the cDNA and determination of the entire structure of the novel protein, which we compare with other members of the plakin family. We also demonstrate the tissue localization of this protein, as determined by Northern blotting and immunostaining. In our previous study, it was unclear whether the epitope that was recognized by our patient's serum was a sequence that is present within plectin itself. In this study, therefore, we also identified the major epitope of the protein. A cDNA library prepared from HeLa cells (CLONTECH) was screened with the inserts of clones pE450-C and pE450-D, which had been isolated by extension cloning, using the same library, after the immunoscreening described in a previous report (25Fujiwara S. Kohno K. Iwamatsu A. Naito I. Shinkai H. J. Invest. Dermatol... 1996; 106: 1125-1130Google Scholar). For isolation of the 3′ end repeats, we constructed a cDNA library from the poly(A)+ RNA of HeLa cells. cDNA was synthesized with a cDNA synthesis kit (Life Technologies, Inc.). We modified the procedure provided with the kit by adding trehalose to the reaction for first-strand synthesis (26Carninci P. Nishiyama Y. Westover A. Itoh M. Nagaoka S. Sasaki N. Okazaki Y. Muramatsu M. Hayashizaki Y. Proc. Natl. Acad. Sci. U. S. A... 1998; 95: 520-524Google Scholar) and, as specific primers for first-strand synthesis, we used 5′-CCAGACACAACAAGTATGCC-3′ for clone Ep115 and 5′-TAGCGCTTGACCGAGTCCATC-3′ for clone Ep 4. The cDNAs were ligated to an EcoRI adapter and then inserted into the EcoRI site of pUC 18 (Amersham Pharmacia Biotech) for construction of plasmid libraries. Each library was screened with the specific probe. To confirm the size of 3′ end repetitive region and cover a of the reaction was using primers and To some cDNA we of cDNA using the (Life and RNA from HeLa cells or the cDNAs with cDNA prepared from poly(A)+ RNA from HeLa cells The of were into the Each cDNA clone and clone of was by a with an primers were constructed for analysis of We protein and using the from the for Furthermore, we also using the and from the of We also used the to of sequence homology of structure were using and are here in a preliminary of heptad were by by analysis the of in heptad between at and were as a of and Stewart M. J. Mol. Biol... 98: Scholar, D.A.D. T.P. J. Mol. Biol... Scholar). The that the at of the heptad of with that in of the heptad of were synthesized with amino acids to the from the of the R. H. Mol. Biol... 1996; Scholar, J.R. C. H. J. 1990; Scholar). The was used to the amino of and the for each of of an amino amino of epiplakin were synthesized as a of with an of amino This region one of five COOH-terminal The cDNA clone that we by immunoscreening a of this domain. To the of the epitope within the sequence we synthesized a of with an of amino acids were synthesized of the the were with our patient's human and and were from The target A B and were synthesized by a D. J. Chem... Scholar). were using an to the were by a with a were with a from to in The molecular mass and amino sequence of each were by of the of mass and a protein Each was to serum to the The was with and and C. of were with of that had been to in with at with that for at to of serum or in that were to the a with at the were with and of a of human that had been to in were for with at the were with and with a with the at was determined with a The with each was by of the to reaction of the with by itself. were as described previously (23Fujiwara S. Shinkai H. Takayasu S. Owaribe K. Tsukita S. Kageshita T. J. Dermatol... 1992; 19: 610-613Google Scholar). of at a of the proteins were separated by an and to an which was with the original patient's serum or the or after with the The cDNA that a specific region of epiplakin was by with specific primers and in as an EcoRI The of the protein in was by at in with for with cells were and by in with at for were with (Amersham Pharmacia Biotech) for at The were with that to the The amino sequence of the protein was with the after blotting of proteins a protein were in The protein was with an of or of protein each of the same of protein were and after the The were after the The were for of A was with using and were with the were by a with in and with at RNA was from and Northern blotting was to the The of with and as primers and as was with and used as the specific for epiplakin The of with and as primers and as was used as the specific for plectin (25Fujiwara S. Kohno K. Iwamatsu A. Naito I. Shinkai H. J. Invest. Dermatol... 1996; 106: 1125-1130Google Scholar). cDNA was used as a probe. tissues were at or into and in were and in for and then for at in that specific or The were in and then with rat or for at with in were in and a We screened the HeLa cell library of cDNA using the end of clone as The clone was and primers from the end of clone were used for to five clones and The original HeLa cell cDNA library was then screened and clones and were In clone the was by an For cloning, we screened the HeLa cell cDNA library using the 3′ end of clone We isolated clones and a clone that included these we that the 3′ of the had homologous repeats of and both with of the repetitive However, the not be identified in this region of the presence of five strongly homologous We constructed a cDNA library using a specific from the region of clone The was and it was isolated together with and We constructed another cDNA library using a specific from the repetitive region of the and isolated a Ep 4. We to the repetitive region and a clone that was in and no The of in the that had in a in the cDNA by sequence. appeared that an at the which was by an the was not by a sequence that the of M. Scholar). The of the was and the predicted amino sequence amino is shown in The molecular mass was to be and was in with the molecular mass of determined previously by The of the protein was the presence of 13 B domains of the identified in the COOH-terminal domain of desmoplakin (1Green K.J. Parry D.A.D. Steinert P.M. Virata M.L.A. Wagner R.M. Angst B.D. Nilles L.A. J. Biol. Chem... 1990; 265: 2603-2612Google Scholar). This identified epiplakin as a of the plakin family, which and plectin. However, unusual were also as The B domains be into two one with to the B domain in desmoplakin and by B in and another with to the B domain in desmoplakin and by in However, all 13 repeats were to the B domain to the A or domain that were also recognized in the COOH-terminal domain of desmoplakin and K.J. Parry D.A.D. Steinert P.M. Virata M.L.A. Wagner R.M. Angst B.D. Nilles L.A. J. Biol. Chem... 1990; 265: 2603-2612Google Scholar). We that the linker in that domains and followed these five B domains were In we identified homologous within each of five strongly linker For in the linker region two of the and were with particularly the last a a homologous region which was the other showed homology the same We predicted that these two or that are separated from one another by and/or The of was and we that these be in two or that a motif found in the structures of many other globular We to a coiled-coil rod domain and an amino-terminal domain, both of which are of all other known members of the plakin family. Furthermore, no dimerization motif was found in the that epiplakin exists in vivo as a single-chain structure. A for other protein did not or or domains B. Nulty E.M. Mir B. Wiche G. J. Cell Biol... 1996; 134: 1455-1467Google Scholar, 16Fuchs P. Zoerer M. Rezniczek G.A. Spazierer D. Oehler S. Castanon M.J. Hauptmann R. Wiche G. Hum. Mol. Genet... 1999; 8: 2461-2472Google between of epiplakin and the COOH-terminal domains B and of 13 domains, with the of domains and of amino These other domains of and B domains of epiplakin that are homologous to the B domain in desmoplakin. in a structure of the COOH-terminal domains of of the of five strongly homologous Each of the linker in followed by B domains and 13 in In amino acids were acids that are in five repeats are the amino of the of homologous within one of five strongly linker The of homologous are with each of these and are with particularly homology the last The homologous region is the two the homology the same acids that are in are 13 domains, with the of domains and of amino These other domains of and B domains of epiplakin that are homologous to the B domain in desmoplakin. To the epitope in epiplakin we the original patient's which had been used for cDNA for of amino acids in with an in the region between amino acids and of epiplakin This region one of five strongly COOH-terminal repeats and the sequence that was recognized by the patient's serum (25Fujiwara S. Kohno K. Iwamatsu A. Naito I. Shinkai H. J. Invest. Dermatol... 1996; 106: 1125-1130Google Scholar). The patient's serum reacted strongly with the amino acids sequence and with amino acids The sequence is unique to epiplakin, but the is also found in plectin and is with the of at the last of to a in desmoplakin. To confirm the of the we an and found that the patient's serum reacted with the To the identification of the we a of with amino acids from the sequence to the patient's serum in another We strongly to from to and, the major epitope recognized by the patient's serum was determined to be blotting showed that the patient's after with the no reacted with the 450-kDa epidermal autoantigen and These that in the patient's serum recognized the unique epitope in blotting using the 450-kDa epidermal of human epidermis was to analysis with the patient's serum and after with the and with the and after with the same and the protein, but not the and recognized the 450-kDa epidermal in all are to of desmoplakin I and in the epidermal the of the 450-kDa epidermal The of molecular size are shown the We used a specific cDNA that linker for Northern the RNA The are shown in A. The that epiplakin was in a of tissues as was plectin but the tissues were The to plectin was in the of and the epiplakin reacted strongly with RNA from and and strongly with RNA from the and the reacted with the 450-kDa epidermal but not after with the and and the protein of epiplakin and also reacted with the 450-kDa epidermal This sequence is within one of five linker that is not homologous to human plectin We staining of human using the the protein of The reacted with the entire epidermis and was keratinocytes we used the which is the same epitope recognized by the patient's we the the the protein of epiplakin, the of eccrine sweat glands and were also for and In the esophagus, the same reacted with all the sheets of the the Immunofluorescence was in cells in the parotid and in mucous epithelial cells of the stomach and In this we have described the cloning and of cDNA clones and a clone for a region that together to the entire for a novel of the plakin family. Epiplakin to be a 450-kDa human epidermal autoantigen that is homologous to plectin and other members of the plakin family (1Green K.J. Parry D.A.D. Steinert P.M. Virata M.L.A. Wagner R.M. Angst B.D. Nilles L.A. J. Biol. Chem... 1990; 265: 2603-2612Google Scholar, 2Tanaka T. Parry D.A.D. Klaus-Kovtun V. Steinert P.M. Stanley J. J. Biol. Chem... 1991; 266: 12555-12559Google Scholar, 3Sawamura D. Li K. Chu M.L. Uitto J. J. Biol. Chem... 1991; 266: 17784-17790Google Scholar, 4Wiche G. Becker B. Luber K. Weitzer G. Castanon M.J. Hauptmann R. Stratowa C. Stewart M. J. Cell Biol... 1991; 114: 83-99Google Scholar, 5McLean W.H.I. Pulkkinen L. Smith F.J.D. Rugg E.L. Lane E.B. Bullrich F. Burgeson R.E. Amano S. Hudson D.L. Owaribe K. McGrath J.A. McMillan J.R. Eady R.A.J. Leigh I.M. Christiano A.M. Uitto J. Genes Dev... 1996; 10: 1724-1735Google Scholar, 6Ruhrberg C. Hajibagheri M.A.N. Simon M. Dooly T.P. Watt F.M. J. Cell Biol... 1996; 134: 715-729Google Scholar, 7Ruhrberg C. Hajibagheri M.A.N. Parry D.A.D. Watt F.M. J. Cell Biol... 1997; 139: 1835-1849Google Scholar, S. Kohno K. Iwamatsu A. Naito I. Shinkai H. J. Invest. Dermatol... 1996; 106: 1125-1130Google Scholar). The COOH-terminal region of plectin contains five homologous known as the B domain and one other homologous known as the domain. the B and domains a which is believed to form a of that is followed by (9Green K.J. Virata M.L.A. Elgart G.W. Stanley J. Parry D.A.D. Int. J. Biol. Macromol... 1992; 14: 145-153Google Scholar). by contains B domains in and these domains are the sequence with Furthermore, epiplakin not appear to a coiled-coil rod domain or an amino-terminal domain. These that epiplakin belongs to a of plakins C. Watt F.M. Dev... 1997; Scholar, Cell Biol... 1999; Scholar). The COOH-terminal domain of epiplakin five strongly repeats of amino each of which linker and a B domain. Moreover, each linker domain homologous We identified a of in the linker regions, but it is likely that form of α-helical an coiled-coil The for this are as coiled-coil structures between in heptad and such structures of in at and (1Green K.J. Parry D.A.D. Steinert P.M. Virata M.L.A. Wagner R.M. Angst B.D. Nilles L.A. J. Biol. Chem... 1990; 265: 2603-2612Google Scholar). that form of a are or and the and are by or In the amino sequence of epiplakin, the of the in the are to of a of of the in globular In the same these in epiplakin not have the of a dimerization motif and, the single-chain of epiplakin another unusual of this protein. The unique of the structures in epiplakin to the in has been suggested that the COOH-terminal domains of plectin, and BPAG1 all with intermediate filaments (11Stappenbeck T.S. Bornslaeger E.A. Corcoran C.M. Luu H.H. Virata M.L.A. Green K.J. J. Cell Biol... 1993; 123: 691-705Google Scholar, 12Wiche G. Gromov D. Donovan A. Castanon M.J. Fuchs E. J. Cell Biol... 1993; 121: 607-619Google Scholar, 13Nikolic B. Nulty E.M. Mir B. Wiche G. J. Cell Biol... 1996; 134: 1455-1467Google Scholar, Y. Dowling J., Yu, Q.C. Kouklis P. Cleveland D.W. Fuchs E. Cell.. 1996; 86: 655-665Google Scholar). an novel human epidermal might also be to bind to such However, no domain, of the found in rat plectin B. Nulty E.M. Mir B. Wiche G. J. Cell Biol... 1996; 134: 1455-1467Google was within the of intermediate that might as might be Northern revealed that epiplakin is expressed at in the in in the and plectin was in and to a in other Thus, epiplakin might with intermediate filaments other In a previous we that epiplakin might be in other in to the epidermis (25Fujiwara S. Kohno K. Iwamatsu A. Naito I. Shinkai H. J. Invest. Dermatol... 1996; 106: 1125-1130Google Scholar). Northern blotting and staining revealed that the or of the and of glandular that of sweat and expressed this protein. In our of it appeared that the protein was expressed in the entire and, in some strongly in the of the These are with in our previous study, in which we in keratinocytes (25Fujiwara S. Kohno K. Iwamatsu A. Naito I. Shinkai H. J. Invest. Dermatol... 1996; 106: 1125-1130Google Scholar). In the previous study, we used a protein from a clone and by a of We to the epitope recognized by these recognized the sequence which is by the region of of the clone and of the The protein by the cDNA had no at amino and no it seems likely that it is in the The of our this As is the for the pathological role of the the epiplakin is not The presence of envoplakin, periplakin and plectin has been reported in the serum of with an autoimmune blistering disease J.R. Labib R.S. Ariss-Abdo L. Burke T. O'Keefe E.J. Anhalt G.J. J. Clin. Invest... 1992; 89: 1775-1782Google Scholar, 21Kim S.C. Kwon Y.D. Lee I.J. Chang S.N. Lee T.G. J. Invest. Dermatol... 1997; 109: 365-369Google Scholar, 22Proby C. Fujii Y. Owaribe K. Nishikawa T. Amagai M. J. Invest. Dermatol... 1999; 112: 153-156Google Scholar). The serum of the we in a previous report recognized a sequence in epiplakin that is not found in plectin. In we have the structure and tissue of a novel human epidermal which is an unusual of the plakin family that was originally found as an autoantigen in a with a subepidermal blistering disease. The of this protein and the diseases in which it is involved be in We for the study, and we and for this bullous pemphigoid reaction of cDNA serum