Epidermal-Specific Defect of GPI Anchor in Pig-a Null Mice Results in Harlequin Ichthyosis-Like Features

Transcription

1 Epidermal-Specific Defect of GPI Anchor in Pig-a Null Mice Results in Harlequin Ichthyosis-Like Features Mariko Hara-Chikuma, 1 Junji Takeda,w Masahito Tarutani,z Yoshikazu Uchida,y Walter M. Holleran,y Yoko Endo, Peter M. Elias,y and Shintaro Inoue Basic Research Laboratory, Kanebo Ltd, Odawara, Japan; wcollaborative Research Center for Advanced Science and Technology, and Department of Social and Environmental Medicine, Osaka University, Osaka, Japan; zdepartment of Dermatology, Graduate School of Medicine, Osaka University, Osaka, Japan; ydepartment of Dermatology, School of Medicine, University of California, and Dermatology Service and Research Unit Department of Veterans Affairs Medical Center, San Francisco, California, USA We previously demonstrated that the epidermal-specific glycosylphosphatidylinositol (GPI)-anchor-deficient mice, generated by Pig-a gene disruption (Pig-a null mice), exhibited wrinkled and dry skin with hyperkeratosis and abnormal differentiation, and they died within a few days after birth. Here, we investigated the basis for the early demise of these animals, and the potential role of epidermal structural and biochemical abnormalities. The rapid demise of these animals was associated with both diminished epidermal permeability barrier function and decreased stratum corneum (SC) water content. The barrier abnormality could be attributed abnormal internal contents of lamellar bodies, with a downstream failure to generate normal extracellular lamellar bilayers in the SC. Moreover, processing profilaggrin to its monomeric form was impaired in Pig-a null mouse epidermis, while levels of the differentiation-specific proteins, involucrin, loricrin and profilaggrin were normal. Failure of filaggrin processing was accompanied by decreased activity of protein phosphatase 2A, an enzyme involved in profilaggrin to filaggrin processing. Thus, these studies demonstrate a critical role for GPI anchor and GPI-anchored proteins in divergent arms of epidermal terminal differentiation. While the permeability barrier abnormality can be attributed to defects in the lamellar body secretory system, the hydration abnormality is, in part, due to lack of availability of filaggrin-derived proteolytic products. Finally, since the dual abnormalities in the lamellar body secretory system and filaggrin processing resemble two key features of human Harlequin ichthyosis, Pig-a null mice could provide an appropriate analog for further studies of this disease. Key words: epidermal barrier function/filaggrin/gpi-anchored proteins/harlequin ichthyosis/protein phosphatase 2A J Invest Dermatol 123: , 2004 A large number of eukaryotic membrane proteins are modified covalently by a glycosylphosphatidylinositol (GPI) moiety that serves as a membrane anchor. These GPI-anchored proteins exhibit diverse functions, including enzymes, cell adhesion molecules, cell surface antigens, and membranebound receptors (Kinoshita et al, 1995). In humans, a somatic mutation in hematopoietic stem cells of the X-linked Pig-a gene causes paroxysmal nocturnal hemoglobinuria, a hemolytic disease, due to deficiency of the GPI-anchored protein, CD 59 (Kinoshita et al, 1995). GPI anchor synthesis is a multistep process involving a series of gene products. All of the genes involved in GPI anchor biosynthesis are autosomal except Pig-a, which instead is located on the X chromosome (Kinoshita et al, 1997). Lack of the Pig-a gene, which encodes a class A Abbreviations: EDTA, ethylenediamine-n,n,n 0,N 0 -tetraacetic acid; GPI, glycosylphosphatidylinositol; HI, harlequin ichthyosis; LB, lamellar body; OA, okadaic acid; PBS, phosphate-buffered saline; PIG-A, phosphatidyl-inositolglycan class A; PMSF, phenylmethylsulfonylfluoride; PP2A, protein phosphatase 2A; SC, stratum corneum; SG; stratum granulosum; TEWL, transepidermal water loss 1 Present address: Departments of Medicine and Physiology, University of California, San Francisco, California, USA. phosphatidylinositol glycan, causes a variety of downstream deficiencies in the GPI anchor and in GPI-anchored proteins (Watanabe et al, 1998). To investigate whether GPI anchor and GPI-anchored proteins are important in the epidermis, we generated a strain of epidermal-specific, Piga-gene-deficient mice (Pig-a null mice), utilizing a Cre-loxP recombinant system (Tarutani et al, 1997; Gao et al, 2002). The skin of affected Pig-a null mice was wrinkled and dry, exhibiting prominent hyperkeratosis but no epidermal hyperplasia, and they died within a few days after birth. Although these studies demonstrated that GPI anchor and GPI-anchored proteins must be critical for epidermal structure and function, the basis for the early demise of the Pig-a-deficient animals remained unknown. The permeability barrier, which is required for life in a terrestrial environment, relies largely upon a lipid-enriched matrix localized within the extracellular domains of the outer layers of the epidermis, the stratum corneum (SC). A competent barrier requires organization of this hydrophobic matrix into mature lamellar membranes, which are formed from lipid contents secreted from epidermal lamellar bodies (LB) (Elias and Menon, 1991). LB appear first in the stratum spinosum layer, and their numbers increase in the stratum granulosum (SG) layer of the epidermis. In the outermost SG Copyright r 2004 by The Society for Investigative Dermatology, Inc. 464

2 123 : 3 SEPTEMBER 2004 layer, LB migrate to the apical cytosol, where they fuse with the plasma membrane, releasing their contents into the extracellular domains at the SG SC interface, immediately prior to cornification (Elias et al, 1998). Cornified envelopes (CE), which consist of several cross-linked proteins; e.g., involucrin and loricrin, impart mechanical and chemical resistance to the corneocyte, as well as a scaffold for the organization of secreted LB contents into extracellular lamellar membranes. By linking adjacent keratin filaments, processed filaggrin also becomes a CE constituent, at least in the lower SC (Steven and Steinert, 1994). Several proteases and phosphatases, including a protein phosphatase of the phosphatase 2A family, are believed to partake in profilaggrin to profilaggrin processing (Kam et al, 1993; Pearton et al, 2001). Filaggrin is a hydrophilic and cationic protein that aids both in the initial aggregation of keratin filaments within the corneocyte cytosol, and in attachment to keratin filaments to the CE (Ishida-Yamamoto et al, 1999). Higher in the SC, filaggrin is hydrolyzed to its constituent amino acids by an aspartate proteinase (Horikoshi et al, 1999), and further deiminated to products that together comprise much of the natural humectants of the SC (Rawlings et al, 1994). Harlequin ichthyosis (HI) is a severe dermatological, autosomal recessive disorder (Dale et al, 1990; Stewart et al, 2001), with affected infants often dying within days-toweeks of birth, apparently due to massive hyperkeratosis and/or a severe barrier abnormality (Dale and Kam, 1993). HI epidermis shows (1) absence or paucity of LB, (2) abnormal (decreased-to-absent) extracellular lamellar membranes structures, and (3) massive hyperkeratosis of the SC. Although the gene(s) that is (are) responsible for HI has (have) not yet been identified, HI has been divided into three subtypes (I III), based in part on the status of profilaggrin and filaggrin processing (Dale et al, 1990). A defect of conversion from profilaggrin to filaggrin has reported in HI types I and II (Dale and Kam, 1993), while profilaggrin is lacking in HI type III. In addition, decreased protein phosphatase 2A (PP2A) activity is associated with HI type II (Kam et al, 1997). A relationship between HI pathogenesis and GPIanchored protein(s) has not yet been reported. To investigate the role of GPI anchor and GPI-anchored proteins in epidermal function, we further analyzed epidermal-specific Pig-a null mice. We demonstrate: first, severe abnormalities of epidermal permeability barrier function; second, abnormal structure of both LB internal contents and their resultant extracellular membrane lamellae; and third, defects in the processing of profilaggrin to filaggrin, associated with a reduced PP2A activity. These studies show that GPI anchor and GPI-anchored proteins influence key epidermal functions, including epidermal permeability barrier, as well as SC hydration. Since the abnormalities in REQUIREMENT OF GPI ANCHOR FOR EPIDERMIS 465 Pig-a null mice resemble key features of human HI, this model might allow new insights into the pathogenic mechanisms operative in HI. Results Impaired barrier function and reduced SC hydration in Pig-a null mice To determine whether GPI anchor and GPIanchored proteins affect one or more critical epidermal functions, we first examined the epidermal permeability barrier function(s) and SC hydration. Transepidermal water loss (TEWL) across Pig-a null mouse skin explants were compared to rates for comparable samples from wild-type littermates (control) using a gravimetric technique (Nolte et al, 1993). As shown in Fig 1A, TEWL values were markedly higher in the Pig-a null mice than in wild-type control (5.0-fold increase; po0.01). Despite increased rates of water loss, high-frequency, skin-surface conductance, representative of SC water content, was significantly lower in Pig-a null mice than in control mice (po0.01) (Fig 1B). Thus, two key epidermal functions: (1) epidermal permeability barrier function and (2) SC hydration, are compromised in epidermal-specific Pig-a null mice. Abnormalities in lamellar body secretory system account for barrier abnormality To investigate the basis for the increased TEWL in the Pig-a null mice, we next performed an ultrastructural analysis of the outer epidermis by electron microscopy, utilizing both osmium tetroxide (OsO 4 ) and ruthenium tetroxide (RuO 4 ) post-fixed tissue samples (Fig 2). While normal extracellular lamellar structures were present in the SC of wild-type mice (Fig 2B), lamellar membrane structures in Pig-a null mouse SC were shortened, distorted, and replaced by incompletely processed, LB-derived materials (Fig 2A). Moreover, most of the LBs in the SG layer of Pig-a null epidermis also displayed abnormalities in size and/or internal contents, indicating an abnormality in LB formation, although the overall numbers of LB ( ¼ denensity) appeared to be normal in Pig-a null mice (Fig 2C). In contrast, the structure of corneocytes, including the cornified envelope (CE), appeared normal in Pig-a null mouse SC (Fig 2A). Together, these findings suggest that the permeability barrier abnormality in Pig-a null mice results from defects in LB structure/contents, which in turn result in extracellular abnormalities in lamellar membrane structures. Defective profilaggrin-to-filaggrin processing accounts for abnormal SC hydration Since our prior study demonstrated hyperkeratosis in the Pig-a null mice (Tarutani et al, 1997), we next investigated whether alterations in either the levels of keratinocyte differentiation protein markers or their Figure 1 Increased transepidermal water loss (TEWL) and reduced skin conductance in Pig-a null mice. (A) TEWL in dorsal skin of wild-type and Pig-a null mice at 6 8 h after birth (mean SE, four mice per group); po0.01. (B) High-frequency superficial skin surface conductance in dorsal skin of wildtype and Pig-a null mice at 6 8 h after birth (mean SE, four mice per group); po0.01.

3 466 HARA-CHIKUMA ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY Figure 2 Defect of SC lipid lamellar and numerous abnormal lamellar bodies in SG in Pig-a null mice. Lipid lamellar membrane structures examined by electron microscopy with RuO 4 postfixation (A, Pig-a null mice; B, wild-type mice). Scale bars ¼ 0.1 mm. d ¼ desmosome. (C) Stratum granulosum (SG) in Pig-a null examined by electron microscopy with OsO 4 postfixation. Arrowheads point to abnormal LB in which lamellar tructures were not clear; (inset) normal lamellar body of wild-type littermate. Scale bars ¼ 0.5 mm, inset ¼ 0.1 mm. breakdown products could also contribute to the functional abnormalities in these animals. Western blot analysis revealed that the levels of profilaggrin, filaggrin dimer (Fig 3A), involucrin (Fig 3B) and loricrin (Fig 3C) were not altered in the Pig-a null versus normal control mice. The Pig-a null epidermis, however, exhibited a virtual absence of filaggrin monomer (Fig 3A). To determine whether the lack of filaggrin monomer in the Pig-a null mice reflects defective processing of profilaggrin to filaggrin or increased degradation of filaggrin to free amino acids, we next analyzed levels of free amino acids in the SC. Total free amino acid content was significantly reduced in the Pig-a null mice ( vs nmol per nmol, po0.001). The distribution of amino acids was also altered in Pig-a null mouse SC in comparison to normal control mice (data not shown). Specially, the levels of arginine, glycine, histidine, and serine, which are the major components of filaggrin, were selectively decreased in Pig-a null mouse SC (data not shown). Moreover, a high correlation was found between the amino acid composition of wild type SC and that of murine filaggrin (Genebank accession #A31488) (r ¼ 0.99, po0.01) (Fig 4A), while the same correlation between Pig-a null mice and murine filaggrin were low (r ¼ 0.56) (Fig 4B). In contrast, the composition of defective amino acids (i.e., the amino acid content of wild-type minus Pig-a null for each) correlated highly with the amino acid composition of normal murine filaggrin (r ¼ 0.92, po0.01) (Fig 4C). Together, these results suggest that the lack of filaggrin monomer reflects a failure to generate downstream products of filaggrin, rather than increased rates of profilaggrin degradation. Finally, we investigated whether the decrease in filaggrin monomer reflects defective processing (dephosphorylation) of profilaggrin to filaggrin by the responsible phosphatase. Specifically, we compared to the activity and protein levels of the phosphatase, PP2A, which is known to dephosphorylate profilaggrin (Kam et al, 1993). PP2A activity was estimated as okadaic acid-sensitive PP activity, which has been shown repeatedly to correspond closely with PP2A activity (e.g., Kam et al, 1997). As shown in Fig 5A, PP2A activity was significantly lower in Pig-a null than that in wild-type epidermis, while conversely levels of PP2Ac, a catalytic subunit of PP2A (Kam et al, 1997), were not different in Pig-a null vs. wild-type mice (Fig 5B). Together, these results show that the PP2A-mediated processing of profilaggrin to filaggrin monomer is impaired in Pig-a null mouse epidermis. Discussion Figure 3 Defect of filaggrin monomer in Pig-a null mice. Each lane contains an equal amount (10 mg) of epidermal extracts from Pig-a null ( / ) and wild-type ( þ / þ ) littermate. Filaggrin (A), involucrin (B) and loricrin (C) proteins were detected by Western immunoblotting as described under Methods. Presented is a representative immunoblot of four separate experiments. (A). Arrows indicate filaggrin monomer and dimer. (B) Arrow: involucrin. (C) Arrow: loricrin. We previously demonstrated that epidermal-specific elimination of GPI anchor and GPI-anchored proteins in Pig-a null mice, results in early demise and an abnormal skin phenotype with shiny wrinkled skin and hyperkeratosis (Tarutani et al, 1997). We investigated further here the functional, structural and biochemical basis for these findings. Pig-a null mice displayed abnormalities in two key epidermal functions: first, a severely compromised epidermal permeability barrier that is likely to account for the premature death of these mice. Extracellular lipid-enriched lamellar membranes in the SC interspaces are responsible for epidermal permeability barrier function. These lipids and their precursors are packaged in epidermal LB in the SG, and then secreted into extracellular domains during the transition from the SG to the SC (Elias and Menon, 1991). We show here that the barrier abnormality, in turn, could be explained by profound abnormalities in the LB secretory system, which include defects in organelle assembly and post-secretory processing of secreted LB contents. Our studies did not, however, address which specific protein(s) of organellegenesis, secretion and/or lipid processing is/are

4 123 : 3 SEPTEMBER 2004 REQUIREMENT OF GPI ANCHOR FOR EPIDERMIS 467 Figure 4 Correlation of SC free amino acid (FAA) composition. The defective free amino acid contents were obtained subtracting free amino acid content of Pig-a null from that of wild-type and then converting to amino acid composition (Pig-a null defected FFA: %). The composition in mfg (%) is the amino acid profile of murine filaggrin (Genebank accession #A31488). (A) Wild-type versus mfg; (B) Pig-a null versus mfg ; (C) Pig-a null defected FFA versus mfg. Figure 5 Reduced PP2A activity in Pig-a null mice. (A) Protein phosphatase (PP) activity in epidermis of Pig-a null and wild-type mice. PP activity was calculated from the rate of dephosphorylation of a radioactive substrate as c.p.m. per minute permicrogram protein in the presence and absence of OA (1 mm), a potent inhibitor of PP2A but not PP1 (mean SE, five mice per group); po0.05. po0.01. (B) PP2Ac, catalytic subunit of PP2A, proteins were detected by western immunoblotting as described under Materials and Methods. Equal protein (10 mg) were loaded in a lane. A representative immunoblot from three separate experiments is performed. defective. Nevertheless, our results strongly suggest that a primary abnormality of the LB secretory system produces the barrier abnormality that is responsible for the premature death of Pig-a null mice. We also identified an abnormality in SC hydration that could also contribute to the disease phenotype. The hydration abnormality is, in part, due to the failure to generate filaggrin monomer, which further results in the downstream depletion of humectant amino acids and their deiminated products (Rawlings et al, 1994) in Pig-a null SC. We therefore sought additional or alternative pathological mechanisms to explain the altered hydration status in Pig-a epidermis. Profilaggrin, a high molecular weight, oligomeric protein (4400 kda), is synthesized in the SG cell layer of the epidermis, where it localizes with in F-type keratohyalin granules (Presland et al, 2000). During terminal differentiation, profilaggrin is processed to filaggrin dimer, which is followed by further dephosphorylation of filaggrin by PP2A, along with site-specific proteolysis yielding the filaggrin monomer (Presland et al, 1997). The observed decrease in PP2A activity correlates with, and likely explains, the lack of filaggrin monomer in the Pig-a null mouse epidermis. Yet, despite decreased enzyme activity, PP2A protein levels remained normal in Pig-a null mouse epidermis. Our results suggest further that GPI anchor and GPIanchored proteins could be linked to the human disorder of cornification, HI, which bears clinical, structural, and biochemical resemblance to the skin phenotype of pig-a null mice (Table I). HI is a severe, autosomal recessive skin disorder (Dale et al, 1990; Stewart et al, 2001), whose genetic defect still remains unidentified. The most distinctive metabolic features of human HI are a paucity of LB and defective filaggrin, both of which are dominant features of the Piga null mouse. The defective processing of profilaggrin to filaggrin monomer has been demonstrated in HI types I and II (Dale et al, 1990), which also was evident in Pig-a null epidermis. Moreover, as in the Pig-a null mouse, decreased PP2A activity, but not PP2Ac protein levels, are found in HI type II keratinocytes (Kam et al, 1997). An HI-like mouse has also been identified as a spontaneous mutation at Jackson Laboratory (Sundberg et al, 1997). Although the prominent hyperkeratosis in these mice more clearly resembles human HI, recent studies have revealed the abnormality as due to mutations in cystatin M/E, a serine protease inhibitor (Zeeuwen et al, 2002). Moreover, the LB secretory system is normal in these mice; and furthermore, recent studies exclude cystatin M/E mutations in HI patients (Zeeuwen et al, 2003). Not only do these HI mice display normal membrane and internal contents of LB, 2 there is no evidence of aberrant filaggrin processing. 3 On the other hand, the Pig-a null mouse reveals close 2 Elias PM, Crumrine D, Sunberg JA: Unpublished observations. 3 Hara M, Endo Y, Inoue S: Unpublished results.

5 468 HARA-CHIKUMA ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY Table I. Comparison of symptoms between Pig-a null mice and human HI Symptom Wild-type Pig-a null HI a Morphology Hyperkeratosis þ þ Lamellar body Normal Abnormal Abnormal SC lamellar Present Absent Absent Filaggrin ProFG! FG ProFG! FG ProFG only ProFG only or Absent b PP2A activity Normal Decrease Decrease c PP2Ac protein þ þ þ c a Harlequin ichthyosis (HI) is subdivided into three subtype; type I, II and III (Dale and Kam, 1993). b Only profilaggrin (ProFil) was expressed in type I and II HI. Both ProFil and filaggrin (Fil) are absent in type III HI case (Dale and Kam, 1993). c The decrease in PP2A activity has been reported in type II HI case (Kam et al, 1997). Not reported in type I and III HI. ProFG, profilaggrin; FG, filaggrin. physiological, structural, and biochemical similarities to HI. Interestingly, Pig-a is an X-linked gene, while HI is an autosomal recessive disease. Therefore it is not likely that Pig-a mutation alone accounts for HI. Our results, however, suggest that another gene(s) involved in GPI-anchor biosynthesis, or a GPI-anchored protein, might be defective in HI. For example, GPI anchor or GPI-anchored protein(s) might be critical for cornification, such that alterations would affect both LB biosynthesis/secretion and filaggrin processing, as seen in both Pig-a-deficient mice and HI patients (type I and II). These results suggest that the Pig-a null mouse might be an appropriate model for future studies on the etiology and pathogenesis of HI. In summary, our results indicate that GPI anchor and GPI-anchored proteins appear to be important both for the development of a normal permeability barrier, and for the processing of profilaggrin to filaggrin monomers. Finally, our results show that defects of the GPI anchor and GPI-anchored proteins in the epidermis result in a HI-like disease with hyperkeratosis, abnormal both lamellar body secretory system and filaggrin processing, providing the first evidence that such defects could be etiological/pathogenic features of the human disorder. Materials and Methods Mice The generation of epidermal-specific GPI-anchor deficient mice through Pig-a gene disruption has been described previously (Tarutani et al, 1997). Newborn Pig-a null ( / ) mice were compared to wild-type ( þ / þ ) littermates. Genotypes were confirmed by PCR and Southern blot analysis (Tarutani et al, 1997). All animal experiments were approved by the Animal research Committee of Kanebo Basic Research laboratory in accordance with the National Research Council Guide (National Research Council, 1996). TEWL and skin conductance TEWL was measured gravimetrically from samples taken 6 8 h after birth, as described previously (Nolte et al, 1993). Briefly, skin explants (1.8 cm 2 ) were placed dermis-side down onto parafilm squares, then the lateral edges were sealed with petrolatum in order to prevent water loss from areas other than the epidermal surface. The skin samples were then weighed hourly at ambient temperature and humidity (25 21C, 50% 55%) over a period of 6 h. SC hydration (water content) was measured by high-frequency surface electrical conductance using a Skicon-200 (IBS, Hamamatsu, Japan). Electron microscopy Skin samples obtained 36 h after birth were minced to o0.5 mm 3, fixed in modified Karnovsky s fixative overnight, and post-fixed in either 1.5% aqueous OsO 4 containing 1.5% potassium ferrocyanide, or in 0.2% RuO 4 in 0.1 M sodium cacodylate buffer, as previously described (Holleran et al, 1997). Ultrathin sections were examined by an electron microscope (Zeiss 10A, Carl Zeiss, Thornwood, New York) operated at 60 kv, with or without further contrasting with lead citrate. Immunoblotting Immunoblotting was performed as described previously (Sakai et al, 2003) using polyclonal antibodies to mouse filaggrin (AF 111, BabCo, Berkeley, California), involucrin (BabCo) and loricrin (AF 62, BabCo), and monoclonal antibodies to keratin 6 (LHK6B, NeoMarkers, Fremont, California) and PP2Ac (Upstate Biotechnology, Lake Placid, New York). Assay of protein phosphatase activities Epidermal sheets were separated from the dermis by incubation in PBS solution containing 10 mm EDTA at 371C for 60 min, and homogenized as previously described (Kam et al, 1997). Protein phosphatase (PP) assay was performed as previously described (Kam et al, 1997). Phosphorylated glycogen phophorylase (Phos A) was prepared from purified glycogen phosphorylase (Phos B, Sigma, St Louis, Missouri) and [ 32 P]-g-ATP (Amersham, Piscataway, New Jersey, 110 GBq per mmol) by incubation with phosphorylase kinase and used as the substrate for the PP kinetic assay. PP activity was calculated as radioactivity (c.p.m.) from the dephosphorylation of the substrate per minute per microgram of protein. Okadaic acid (OA), a potent PP inhibitor of PP2A rather than PP1, was added to the assay mixture, and then the reaction was started by adding substrate. Protein was determined using the Bio-Rad DC protein assay kit (Bio-Rad Laboratories, Hercules, California). Free amino acid content SC samples were removed from the skin by tape stripping and incubated in 10 mm HCl to extract the water-soluble fractions. Amino acids were analyzed using a Hitachi Amino Acid Analyzer L-8800 as described previously (Sakai et al, 2003). The total amino acid content derived from the SC proteins was assayed as described previously (Sakai et al, 2000). The free amino acid content per total amino acid content from the SC proteins (nmol/nmol) was calculated as described previously (Sakai et al, 2000). Statistics Statistical significance was assessed using SAS analysis system (SAS Institute Japan, Tokyo, Japan). The authors thank Debbie Crumrine for expert technical assistance. DOI: /j X x Manuscript received November 8, 2003; revised March 16, 2004; accepted for publication March 31, 2004 Address correspondence to: Shintaro Inoue, Basic Research Laboratory, Kanebo Ltd, 3-28, 5-chome Kotobuki-cho, Odawara-shi, Kanagawa-ken , Japan. inoshin@oda.cos.kanebo.co.jp References Dale BA, Holbrook KA, Fleckman P, Kimball JR, Brumbaugh S, Sybert VP: Heterogeneity in harlequin ichthyosis, an inborn error of epidermal

6 123 : 3 SEPTEMBER 2004 REQUIREMENT OF GPI ANCHOR FOR EPIDERMIS 469 keratinization: Variable morphology and structural protein expression and a defect in lamellar granules. J Invest Dermatol 94:6 18, 1990 Dale BA, Kam E: Harlequin ichthyosis. Variability in expression and hypothesis for disease mechanism. Arch Dermatol 129: , 1993 Elias PM, Cullander C, Mauro T, Rassner U, Komuves L, Brown BE, Menon GK: The secretory granular cell: The outermost granular cell as a specialized secretory cell. J Investig Dermatol Symp Proc 3:87 100, 1998 Elias PM, Menon GK: Structural and lipid biochemical correlates of the epidermal permeability barrier. Adv Lipid Res 24:1 26, 1991 Gao XH, Kondoh G, Tarutani M, et al: Rapid compensation for glycosylphosphatidylinositol anchor deficient keratinocytes after birth: Visualization of glycosylphosphatidylinositol-anchored proteins in situ. J Invest Dermatol 118: , 2002 Holleran WM, Uchida Y, Halkier-Sorensen L, Haratake A, Hara M, Epstein JH, Elias PM: Structural and biochemical basis for the UVB-induced alterations in epidermal barrier function. Photodermatol Photoimmunol Photomed 13: , 1997 Horikoshi T, Igarashi S, Uchiwa H, Brysk H, Brysk MM: Role of endogenous cathepshin D-like and chymotrypsin-like proteolysis in human epidermal desquamation. Br J Dermatol 141: , 1999 Ishida-Yamamoto A, Tanaka H, Nakane H, Takahashi H, Hashimoto Y, Iizuka H: Programmed cell death in normal epidermis and loricrin keratoderma. Multiple functions of profilaggrin in keratinization. J Investig Dermatol Symp Proc 4: , 1999 Kam E, Nirunsuksiri W, Hager B, Fleckman P, Dale BA: Protein phosphatase activity in human keratinocytes cultured from normal epidermis and epidermis from patients with harlequin ichthyosis. Br J Dermatol 137: , 1997 Kam E, Resing KA, Lim SK, Dale BA: Identification of rat epidermal profilaggrin phosphatase as a member of the protein phosphatase 2A family. J Cell Sci 106: , 1993 Kinoshita T, Inoue N, Takeda J: Defective glycosyl phosphatidylinositol anchor synthesis and paroxysmal nocturnal hemoglobinuria. Adv Immunol 60:57 103, 1995 Kinoshita T, Ohishi K, Takeda J: GPI-anchor synthesis in mammalian cells: Genes, their products, and a deficiency. J Biochem (Tokyo) 122: , 1997 National Research Council: National Research Council (NRC) guide. Washington, DC: National Academy Press, 1996 Nolte CJ, Oleson MA, Bilbo PR, Parenteau NL: Development of a stratum corneum and barrier function in an organotypic skin culture. Arch Dermatol Res 285: , 1993 Pearton DJ, Nirunsuksiri W, Rehemtulla A, Lewis SP, Presland RB, Dale BA: Proprotein convertase expression and localization in epidermis: Evidence for multiple roles and substrates. Exp Dermatol 10: , 2001 Presland RB, Boggess D, Lewis SP, Hull C, Fleckman P, Sundberg JP: Loss of normal profilaggrin and filaggrin in flaky tail (ft/ft) mice: An animal model for the filaggrin-deficient skin disease ichthyosis vulgaris. J Invest Dermatol 115: , 2000 Presland RB, Kimball JR, Kautsky MB, Lewis SP, Lo CY, Dale BA: Evidence for specific proteolytic cleavage of the N-terminal domain of human profilaggrin during epidermal differentiation. J Invest Dermatol 108: , 1997 Rawlings AV, Scott IR, Harding CR, Bowser PA: Stratum corneum moisturization at the molecular level. J Invest Dermatol 103: , 1994 Sakai S, Endo Y, Ozawa N, et al: Characteristics of the epidermis and stratum corneum of hairless mice with experimentally induced diabetes mellitus. J Invest Dermatol 120:79 85, 2003 Sakai S, Sasai S, Endo Y, Matue K, Tagami H, Inoue S: Characterization of the physical properties of the stratum corneum by a new tactile sensor. Skin Res Technol 6: , 2000 Steven AC, Steinert PM: Protein composition of cornified cell envelopes of epidermal keratinocytes. J Cell Sci 107: , 1994 Stewart H, Smith PT, Gaunt L, et al: De novo deletion of chromosome 18q in a baby with harlequin ichthyosis. Am J Med Genet 102: , 2001 Sundberg JP, Boggess D, Hogan ME, et al: Harlequin ichthyosis (ichq): A juvenile lethal mouse mutation with ichthyosiform dermatitis. Am J Pathol 151: , 1997 Tarutani M, Itami S, Okabe M, et al: Tissue-specific knockout of the mouse Pig-a gene reveals important roles for GPI-anchored proteins in skin development. Proc Natl Acad Sci USA 94: , 1997 Watanabe R, Inoue N, Westfall B, Taron CH, Orlean P, Takeda J, Kinoshita T: The first step of glycosylphosphatidylinositol biosynthesis is mediated by a complex of PIG-A, PIG-H, PIG-C and GPI1. EMBO J 17: , 1998 Zeeuwen PL, Dale BA, de Jongh GJ, et al: The human cystatin M/E gene (CST6): Exclusion candidate gene for harlequin ichthyosis. J Invest Dermatol 121:65 68, 2003 Zeeuwen PL, van Vlijmen-Willems IM, Hendriks W, Merkx GF, Schalkwijk J: A null mutation in the cystatin M/E gene of ichq mice causes juvenile lethality and defects in epidermal cornification. Hum Mol Genet 11: , 2002