E XP ER I ME NTAL C E LL RE S E ARCH

] (]]]]) ]]]–]]]

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/yexcr

Research Article

Isolation of a hemidesmosome-rich fraction from a human squamous cell carcinoma cell line Yoshiaki Hirakoa,n, Yuki Yonemotoa, Tomoe Yamauchia, Yuji Nishizawab, Yoshiyuki Kawamotob, Katsushi Owaribea a

Division of Biological Science, Graduate School of Science, Nagoya University, Furo-cho, Chikusa-ku, Nagoya 464-8602, Japan Department of Biomedical Sciences, Chubu University, 1200 Matsumoto-cho, Kasugai 487-8501, Japan

b

article information

abstract

Article Chronology:

Hemidesmosomes are cell-to-matrix adhesion complexes anchoring keratinocytes to basement

Received 20 December 2013

membranes. For the first time, we present a method to prepare a fraction from human cultured

Received in revised form

cells that are highly enriched in hemidesmosomal proteins. Using DJM-1 cells derived from

23 March 2014

human squamous cell carcinoma, accumulation of hemidesmosomes was observed when these

Accepted 1 April 2014

cells were cultured for more than 10 days in a commercial serum-free medium without supplemental calcium. Electron microscopy demonstrated that numerous electron-dense adhe-

Keywords: Hemidesmosome Basement membrane Keratinocyte Laminin Autoimmune disease

sion structures were present along the basal cell membranes of DJM-1 cells cultured under the aforementioned conditions. After removing cellular materials using an ammonia solution, hemidesmosomal proteins and deposited extracellular matrix were collected and separated by electrophoresis. There were eight major polypeptides, which were determined to be plectin, BP230, BP180, integrin α6 and β4 subunits, and laminin-332 by immunoblotting and mass spectrometry. Therefore, we designated this preparation as a hemidesmosome-rich fraction. This fraction contained laminin-332 exclusively in its unprocessed form, which may account for the promotion of laminin deposition, and minimal amounts of Lutheran blood group protein, a nonhemidesmosomal transmembrane protein. This hemidesmosome-rich fraction would be useful not only for biological research on hemidesmosomes but also for developing a serum test for patients with blistering skin diseases. & 2014 Elsevier Inc. All rights reserved.

Introduction Hemidesmosomes (HDs) are cell–matrix adhesion complexes that anchor intermediate filaments at their cytoplasmic plaques and

bind to the epidermal basement membrane via their transmembrane receptors [1]. The absence of hemidesmosomal constituents in patients affected with several different types of inherited skin blistering diseases have confirmed the significance of the HD as a

Abbreviations: B-CAM, basal cell adhesion molecule; BMP-1, bone morphogenetic protein 1; BP, bullous pemphigoid; CBB, Coomassie Brilliant Blue; EMEM, Eagle's minimum essential medium; HD, hemidesmosome; KGM, keratinocyte growth medium; Lu, Lutheran blood group protein; MAb, monoclonal antibody; mTLD, mammalian Tolloid; PAb, polyclonal antibody n

Corresponding author. Fax: þ81 52 789 4818. E-mail address: [email protected] (Y. Hirako).

http://dx.doi.org/10.1016/j.yexcr.2014.04.002 0014-4827/& 2014 Elsevier Inc. All rights reserved.

Please cite this article as: Y. Hirako, et al., Isolation of a hemidesmosome-rich fraction from a human squamous cell carcinoma cell line, Exp Cell Res (2014), http://dx.doi.org/10.1016/j.yexcr.2014.04.002

2

E XP E RI ME N TAL CE L L R ES E ARC H

dermal-epidermal adhesion apparatus [2]. HD proteins are also known as autoantigens in human autoimmune skin blistering diseases [3]. We previously developed a method to isolate an “HD fraction” from bovine corneas [4]. Analyses of the isolated bovine HD fraction revealed five major HD constituents that were subsequently identified as plectin, bullous pemphigoid antigen 230 (BP230)/BPAG1, integrin α6 and β4 subunits, and bullous pemphigoid antigen 180 (BP180)/type XVII collagen. In addition to the five major HD proteins, this HD fraction contained several minor polypeptides. We attempted to generate specific antibodies against these polypeptides to determine if they were novel HD components; however, these attempts failed. Thus, we considered analyzing this HD fraction by mass spectrometry (MS). However, because of the lack of a peptide database for bovine proteins, such an analysis would have been problematic. Moreover, the recent epidemic of bovine spongiform encephalopathy in Japan would have made it nearly impossible for us to use bovine corneas. To overcome these difficulties, we planned to develop a novel method to isolate an HD-rich fraction from human cultured cells. This fraction then could be analyzed using MS with the advantage of a well-developed human peptide database and easier accessibility to the material. In addition, an HD-rich fraction of human origin would be a much more suitable substrate compared with the bovine HD fraction for detecting antigens targeted by patients' autoantibodies. It is known that a matrix containing laminin-332 can be isolated by treating cultured keratinocytes with an ammonia solution [5,6]. We previously found that trace amounts of integrin β4 and BP180 that were isolated together with laminin-332 from DJM-1 cells, a cell line derived from human squamous cell carcinoma [7,8]. This finding prompted us to establish a culture condition that would promote the formation of HDs in DJM-1 cells. Because HDs are very tough structures that bind tightly to the basement membrane including laminin-332 [4], we anticipated that mature HDs formed in cultured cells would be strong enough to resist treatment with an ammonia solution, so that they could be isolated together with laminin-332. In this study, we report the following results: the formation/ accumulation of HDs was promoted in DJM-1 cells that were cultured long-term in keratinocyte growth medium (KGM) without supplemental calcium; a fraction enriched with the five major HD proteins and laminin-332 was successfully prepared from DJM-1 cells; DJM-1 cells deposited laminin-332 exclusively in its unprocessed form under the culture condition. The biological mechanisms underlying the formation/accumulation of HDs in the cultured cells and the possible clinical applications of this isolated HD-rich fraction are also discussed.

Materials and methods Antibodies Mouse MAb-279 against BP230, MAb-1A3 against the intercellular domain of integrin β4 subunit, MAb-1A8c against the intercellular domain of BP180, and MAb-BMM62 against the ectodomain of BP180 were produced as described previously [4]. Mouse MAbC34 was raised against the COOH-terminal fragment (Ile1188Pro1497) of human BP180. Mouse MAb-PC742 was raised against

] (]]]]) ]]]–]]]

the COOH-terminal fragment (Asp2930-Pro3153) of human plectin. Mouse MAbs specific to laminin α3 (BM515) and γ2 (YN557) chains were prepared by immunizing mice with a basement membrane fraction of bovine cornea [9]. Mouse MAbs against Lu/B-CAM (BRIC221), laminin γ1 chain (sc-13144) and laminin β3 chain (sc-81809) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit PAb to Lu/B-CAM is a kind gift from Dr. Kikkawa at Tokyo Pharmaceutical University [10]. Rat MAb to integrin α6 (GoH3) is a kind gift from Dr. Sonnenberg at the Netherlands Cancer Institute [11]. Goat PAb to human BMP-1 was raised against the NH3-terminal portion (Ala121-Gln730) that is common to both BMP-1 and mTLD (R&D systems, Minneapolis, MN) [12]. Mouse MAb against laminin β1 chain (MAB1921) was purchased from Chemicon International (Temecula, CA). Secondary antibodies labeled by Alexa Fluor 488 or Alexa Fluor 594 were purchased from Molecular Probes (Eugene, OR).

Cultured cells DJM-1 cells, a human skin squamous carcinoma cell line, were kindly provided by Dr. Kitajima at Gifu University [13]. Normally, DJM-1 cells are cultured in the Eagle's minimum essential medium (EMEM) from Nissui pharmaceutical (Tokyo, Japan) supplemented with 400 ng/ml hydrocortisone (Sigma-Aldrich, St. Louis, MO), 20 ng/ml epidermal growth factor (Sigma-Aldrich), 2 mM glutamine, and 10% FCS. Normal human epidermal keratinocytes (NHK) from neonatal foreskin were purchased from Cell Applications (San Diego, CA). NHK were grown in EpiLife medium containing 60 μM calcium and supplemented with HKGS (Life Technologies, Carlsbad, CA). A431 cells (a human epidermoid carcinoma cell line) and BxPC-3 cells (a human pancreatic adenocarcinoma cell line) were obtained from Japanese Collection of Research Bioresources Cell Bank (Osaka, Japan). A431 cells were grown in Dulbecco's modified Eagle medium with high glucose containing 10% FCS, and Bx-PC3 cells were grown in RPMI1640 medium with 10% FCS.

Isolation of the HD-rich fraction from DJM-1 cell DJM-1 cells were grown to 80%–100% confluency under the normal culture condition as described above. Subsequently, the culture medium was completely removed and replaced with KGM (Cat. No. CC-3112; Lonza, Walkersville, MD) without supplemental calcium (KGM minus Ca). KGM consisted of keratinocyte basal medium (Cat. No. CC-3104; Lonza), 0.4% bovine pituitary extract, 5 μg/ml bovine insulin, 0.5 μg/ml hydrocortisone, 0.1 ng/ml human epidermal growth factor, 50 μg/ml gentamicin and 50 ng/ml amphotericin-B. Cells cultured for 10–14 days with KGM minus Ca (the medium was renewed every 3 or 4 days) were washed thoroughly with PBS and then treated with 20 mM ammonia hydroxide for 5 min. The dish was shaken to detach disrupted cells. Subsequently, the solution was discarded along with the floating materials. After this treatment, matrices deposited on the dish were washed three times with PBS and then treated with 0.1% Triton X-100 in PBS for 5 min to solubilize and remove membranous debris. The buffer was discarded, and the deposited material was then washed three times with PBS. When using DJM-1 cells from a 14-day culture, dead cells sometimes remain on the bottom of the dish even after ammonia solution treatment. In this case, these cells could be removed by gently

Please cite this article as: Y. Hirako, et al., Isolation of a hemidesmosome-rich fraction from a human squamous cell carcinoma cell line, Exp Cell Res (2014), http://dx.doi.org/10.1016/j.yexcr.2014.04.002

E X PE R IM EN TA L C ELL R E S EA RC H

scraping the bottom surface of the dish with a cell scraper in PBS. Matrices and attached HD proteins were finally solubilized in the 5  SDS sample buffer (200 mM Tris–HCl, pH 6.8, 10% SDS, 0.05% bromophenol blue, 50% glycerol, 100 mM dithiothreitol). Since HDs are highly insoluble structures, the bottom surface of the dish should be scraped several times using a cell scraper with the SDS sample buffer. We designated this sample solution as the HD-rich fraction. We typically used four confluent 10-cm dishes of DJM-1 cells to prepare 400–600 μl of the HD-rich fraction. A 20–40-μl aliquot of the HD-rich fraction was sufficient to isolate clear

KGM–Ca

polypeptide bands using SDS-PAGE (7.5% mini-gel) that were stained with Coomassie Brilliant Blue (CBB).

Immunofluorescence microscopy DJM-1 cells grown on glass coverslips were treated with 20 mM ammonia hydroxide to remove cellular materials. After this treatment, matrices deposited on the coverslips were fixed with acetone or methanol at  20 1C for 5 min and then air dried. These coverslips were processed for immunofluorescence staining with

KGM+Ca

3 days

EMEM

3

] (]]]]) ]]]–]]]

Supplemental CaCl2 concentration 0.5 mM 0.1 mM

0 mM

10 days

14 days

7 days

1 mM

KGM–Ca

KGM+Ca

BP230

14 days

EMEM

LM α3

Laminin α3 (MAb-BM515)

Integrin β4

Integrin α6

Laminin γ2

BP180

Phase contrast

10 days

Plectin

0 mM

14 days

Treated by ammonia solution

Supplemental CaCl2 concentration 0.5 mM 0.1 mM

7 days

1 mM

BP230 (MAb-279) Treated by ammonia solution

Fig. 1 – DJM-1 cells form HD-like adhesions under long-term culture conditions. DJM-1 cells were cultured on glass coverslips in Eagle's minimum essential medium (EMEM) with 10% FCS and in keratinocyte growth medium (KGM) either with (KGM plus Ca) or without CaCl2 (KGM minus Ca) supplementation for 3 to 14 days after achieving confluent growth in EMEM. For immunostaining, glass coverslips were treated with an ammonia solution to remove cellular materials. (A) Phase contrast images of DJM-1 cells cultured for either 3 or 14 days are shown. (B) Glass coverslips from 14-day cultures were stained with MAbs against BP230, a cytoplasmic HD plaque protein, and against the α3 chain of laminin. (C) DJM-1 cells cultured in KGM minus Ca for 14 days after reaching confluent growth were stained by antibodies against plectin (MAb-PC742), integrin β4 (MAb-1A3), BP180 (MAb-BMM62), integrin α6 (MAb-GoH3), and laminin γ2 (MAb-YN557). Panels for integrin α6 and laminin γ2 are double immunofluorescence images. After treatment with an ammonia solution, no cells remained on the glass coverslips, similar to the bottom right panel showing a phase contrast image of the sample immuno-stained for integrin α6 and laminin γ2. (D) DJM-1 cells were cultured for 7, 10, or 14 days in KGM supplemented with 0.1, 0.5, or 1 mM CaCl2 or without CaCl2 (0 mM). After the treatment with an ammonia solution, matrices were stained with MAb-279 specific for BP230 and MAb-BM515 specific for the α3 chain of laminin. Exposure time was fixed for the comparison in B and D. Bars¼ 20 μm. Please cite this article as: Y. Hirako, et al., Isolation of a hemidesmosome-rich fraction from a human squamous cell carcinoma cell line, Exp Cell Res (2014), http://dx.doi.org/10.1016/j.yexcr.2014.04.002

4

E XP E RI ME N TAL CE L L R ES E ARC H

] (]]]]) ]]]–]]]

primary antibodies followed by secondary antibodies as described previously [14]. Human skin sections were kindly provided by Dr. Takashi Hashimoto at Kurume University, Japan. The skin samples were obtained from a healthy volunteer following the guidelines of the Medical Ethics Committees of Kurume University School of Medicine. These specimens were observed under an Axiovert 200 microscope (Carl Zeiss, Oberkochen, Germany).

Mass analysis of peptides was performed using a LTQ XL linear ion trap mass spectrometer (Thermo Scientific, San Jose, CA). The data were analyzed with the Xcalibur 2.0.5 and BioworksBrowser 3.3 software (Thermo Scientific), and the proteins were identified against human protein database obtained from NCBI Reference Protein Sequence Database (September 2012) modified with carbamidomethylation on cysteine.

Electron microscopy

Preparation of conditioned media and immunoblotting

DJM-1 cells were cultured on 3.5-cm plastic dishes. After washing with PBS, cells were fixed with glutaraldehyde, post-fixed with osmium tetroxide, dehydrated in a graded ethanol series, and embedded in Epon. Thin sections were stained with 5% uranyl acetate in 50% ethanol followed by 0.4% lead citrate and examined under a JEM 1200EX electron microscope (JEOL, Tokyo, Japan) at 100 kV.

Conditioned media from EMEM cultures were concentrated using ammonium sulfate precipitation as described previously [16]. To concentrate proteins from KGM cultures, conditioned media were concentrated using Amicon Ultra (Merck Millipore, Darmstadt, Germany). Immunoblotting was performed as described previously [14]. To determine the relative intensity of immunoblots, protein bands were visualized using an enhanced chemiluminescent detection kit (GE Healthcare, Little Chalfont, UK) and detected by the Light Capture AE-2150 system (ATTO, Tokyo, Japan). Analysis of the detected bands was conducted using ImageJ 1.46 software (http://rsbweb.nih.gov/ij/).

Protein identification by MS Liquid chromatography–tandem MS analysis was performed as described [15], with slight modifications. Briefly, the HD-rich fraction of about 40 μl was subjected to SDS-PAGE (7.5% minigel). Bands of interest were manually excised from the gel stained with SimplyBlue SafeStain (Invitrogen, Carlsbad, CA). Gel pieces were washed three times with 50% acetonitrile in 25 mM ammonium bicarbonate, and then dehydrated with 100% acetonitrile and dried in a vacuum-evaporator. The gel pieces were digested with 20 μg/ml trypsin (Promega, Madison, WI) solution at 37 1C overnight. The resultant peptide mixtures were extracted twice with 0.1% trifluoroacetic acid/50% acetonitrile. The extracted peptides were concentrated in a centrifugal vacuum-evaporator and subjected to liquid chromatography–tandem MS analysis.

BP230

Long-term culture of DJM-1 cells in KGM without supplemental calcium promotes the deposition of HD proteins on the matrix After reaching confluent growth, DJM-1 cells cultured on glass coverslips were further cultured for 14 days using Eagle's minimum essential medium (EMEM) with 10% FCS to promote the formation/maturation of HDs (Fig. 1A, left panels). After removing

BP180

Integrin β4

Plectin

KGM+FCS

KGM+Ca

KGM–Ca

Laminin α3

Results

Fig. 2 – Immunofluorescence microscopy of DJM-I cells long-term cultured with KGM minus Ca, KGM plus Ca, or KGM plus FCS. DJM-1 cells were cultured in KGM minus Ca, KGM plus Ca, or KGM supplemented with 10% FCS for 14 days after achieving confluent growth in EMEM plus FCS. Cells were fixed and stained by antibodies without pretreatment of ammonium solution. These specimens were observed under an FV1000 confocal laser scanning microscope (Olympus, Tokyo, Japan). Bar ¼20 μm. Please cite this article as: Y. Hirako, et al., Isolation of a hemidesmosome-rich fraction from a human squamous cell carcinoma cell line, Exp Cell Res (2014), http://dx.doi.org/10.1016/j.yexcr.2014.04.002

E X PE R IM EN TA L C ELL R E S EA RC H

cellular materials by treatment with an ammonia solution, proteins remaining on the glass coverslips were examined using immunofluorescence microscopy (Fig. 1B, left panels). A monoclonal antibody (MAb) against BP230, a hemidesmosomal plaque protein, did not produce any detectable signals in the specimen, whereas a MAb against the laminin α3 chain showed weak punctate staining. This finding indicated the absence of HD-like adhesion complexes that were deposited on the laminin-332 matrix under this condition. Subsequently, we used serum-free keratinocyte growth medium (KGM) with or without supplemental CaCl2 for the long-term culture of DJM-1 cells. Cells cultured in KGM without supplemental CaCl2 (KGM minus Ca) had loose intercellular contacts, whereas cells cultured in KGM with supplemental 1 mM CaCl2 (KGM plus Ca) tightly adhered to each other, similar to the cells cultured in EMEM (Fig. 1A). Glass coverslips from the 14-day cultures in KGM minus Ca and KGM plus Ca were treated with an ammonia solution and then stained with the MAbs against BP230 and laminin α3 (Fig. 1B). BP230 and laminin α3 staining showed a clear punctate pattern of HD proteins that were deposited on matrices prepared from DJM-1 cells cultured in KGM minus Ca, while the fluorescence signals of these proteins were much weaker in the cells cultured in KGM plus Ca. In addition, we observed other HD proteins that were deposited on the matrices prepared from the cells cultured in KGM minus Ca (Fig. 1C). To define the culture conditions required for the deposition of HD proteins on these matrices, DJM-1 cells were cultured in KGM with different calcium concentrations (Fig. 1D). The immunofluorescence signals of laminin α3 chain and BP230 decreased as the concentration of supplemental CaCl2 increased from 0 to 0.1, 0.5, and 1 mM, indicating that calcium concentration is one of the key factors involved in the deposition. In addition, we examined the deposition of laminin α3 and BP230 in 7-, 10-, and 14-day cultures of DJM-1 cells. Deposition of laminin α3 chain was most prominent in samples taken from 10- and 14-day cultures with the KGM minus Ca condition. Deposition of BP230 also became evident in samples taken from the 10-day culture. From these observations, we can conclude that culturing DJM-1 cells for more than 10 days in KGM minus Ca was optimal culture condition to enhance the deposition of HD proteins on the laminin-332 matrix. Depositions of HD-related proteins were also examined in DJM-1 cells that were not treated with ammonia solution (Fig. 2). DJM-1 cells cultured in KGM minus Ca showed prominent accumulation of laminin α3 chain, BP230, and BP180, while depositions of these proteins were markedly decreased in the cells cultured in KGM plus Ca. Plectin and integrin β4 subunit showed scattered linear staining pattern in KGM plus Ca condition. The addition of 10% FCS to the KGM minus Ca culture decreased the depositions of hemidesmosomal proteins and laminin-332 nearly to the level observed in EMEM plus FCS condition.

Electron microscopy revealed mature HD structures in DJM-1 cells To examine whether DJM-1 cells formed morphologically mature HD-like adhesions under the culture conditions described above, electron microscopy was used to examine these cells. Low magnification images showed that DJM-1 cells cultured in KGM minus Ca for 14 days did not form tight cell-to-cell contacts (Fig. 3A, arrow). In contrast, the cells cultured in KGM plus Ca

] (]]]]) ]]]–]]]

5

tightly adhered to each other along their lateral cell borders (Fig. 3A, dashed lines). At higher magnification, DJM-1 cells cultured in KGM minus Ca showed numerous electron-dense structures at the basal cell membrane (Fig. 3B, arrow heads). In contrast, cells cultured in EMEM and KGM plus Ca did not have such structures at their basal cell membranes. This finding demonstrated that DJM-1 cells cultured under the KGM minus Ca condition had formed morphologically mature HDs.

Identification of HD proteins and laminin-332 in the HD-rich fraction by immunoblotting and MS To analyze the proteins deposited on the matrix, DJM-1 cells cultured for 11 days in KGM minus Ca were treated with an

KGM-Ca

KGM+Ca

EMEM+10%FCS

KGM-Ca

KGM+Ca

Fig. 3 – Electron microscopy demonstrating that DJM-1 cells cultured in KGM minus Ca have numerous electron-dense HD structures. (A) Low magnification images of DJM-1 cells cultured for 14 days after reaching confluent growth. A large gap (arrow) was found at the border between the two cells cultured in KGM minus Ca. Cells cultured in KGM plus Ca adhered tightly to each other at their cell borders (dashed lines). Bar¼ 2.5 μm. (B) High magnification images showing numerous electron-dense HD-like structures along the basal side of DJM-1 cells cultured in KGM minus Ca (arrowheads). Arrow indicates bundles of keratin filaments. In contrast, these structures were barely detectable in cells cultured in EMEM or in KGM plus Ca. Bar¼ 500 nm.

Please cite this article as: Y. Hirako, et al., Isolation of a hemidesmosome-rich fraction from a human squamous cell carcinoma cell line, Exp Cell Res (2014), http://dx.doi.org/10.1016/j.yexcr.2014.04.002

1, 1 2 3

200

5 116

Matrices KGM -Ca +Ca

EM EM

ch -ri D H

M

] (]]]]) ]]]–]]]

p 2, lect B in 3, P23 in 0 4, teg la rin 5, mi β B nin 4 6, P18 α 3 la 0 7, min la in m γ in 2 in β3

E XP E RI ME N TAL CE L L R ES E ARC H

Matrices KGM -Ca +Ca Blot: anti-

EM EM

6

BP230 200

1

100

41

0

100

47

0

100

21

4 6

116

7

97

8

Integrin β4 BP180

66 97

CBB staining

K

EM EM

(kDa)

Matrices Conditioned media KGM KGM -Ca +Ca -Ca +Ca Blot: anti-

EM EM

66

Conditioned media KGM -Ca +Ca 200

LM α3

CBB staining

116

LM γ2

97

Blot: anti-BMP-1/mTLD LM β3 58 100 38

100 14

31

Fig. 4 – Immunoblotting demonstrating the enrichment of HD proteins and laminin-332 in the fraction isolated from DJM-1 cells. (A) An HD-rich fraction was prepared from DJM-1 cells cultured for 11 days in KGM minus Ca. SDS-PAGE analysis followed by CBB staining showed that this fraction contained eight major polypeptide bands ranging in molecular mass from 500 to 120 kDa (bands 1–8). Square bracket indicates keratins (K). Asterisk indicates the position of the Lu/B-CAM protein. (B) The HD-rich fraction was analyzed using immunoblotting with MAbs against HD components and laminin-332. Bands 1, 2, 3, 4, 5, 6, and 7 were plectin, BP230, integrin β4, unprocessed laminin α3 chain, BP180, unprocessed laminin γ2 chain, and laminin β3 chain, respectively. (C) Matrices prepared from cells cultured in EMEM, KGM minus Ca, and KGM plus Ca were CBB stained or immunoblotted with MAbs against BP230, integrin β4, and BP180. Numbers below blots represent relative band intensities. (D) Matrices (left) and conditioned media (right) prepared from DJM-1 cells cultured in EMEM, KGM minus Ca, and KGM plus Ca were immunoblotted with MAbs against the α3, γ2, and β3 chains of laminin. Arrows and arrowheads indicate the positions of the unprocessed and processed forms of the laminin chains, respectively. Numbers below blots represent relative band intensities. (E) Conditioned media obtained from the cells cultured in KGM minus Ca and KGM plus Ca were immunoblotted with a PAb against BMP-1/mTLD. This antibody detected a 130-kDa band (arrow) in conditioned KGM minus Ca and a 120-kDa band (arrow head) in KGM plus Ca.

ammonia solution, and the resulting matrices were dissolved in the SDS sample buffer. These samples were subjected to SDSPAGE. Coomassie brilliant blue (CBB) staining revealed eight major bands ranging in molecular mass from 500 kDa to 150 kDa (numbered from 1 to 8 from the top of the gel in Fig. 4A). Immunoblotting demonstrated that bands 1, 2, 3, 4, 5, 6, and 7 include plectin, BP230, integrin β4 subunit, unprocessed laminin α3 chain, BP180, unprocessed laminin γ2 chain, and laminin β3 chain, respectively (Fig. 4B). To identify a major polypeptide found in band 8 and to confirm the immunoblot results, bands 1 to 8 were cut out from the gel and analyzed using MS. As shown in Table 1, the integrin α6 subunit was the protein with the highest score for band 8. Regarding bands from 1 to 7, the proteins identified using immunoblotting demonstrated the top scores for each band. These results indicated that HD proteins and laminin-332 had been enriched in this preparation. Thus, we designated this preparation as the HD-rich fraction. Relative quantification of HD proteins in matrices isolated from the cells

cultured in EMEM, KGM minus Ca, and KGM plus Ca confirmed that the HD-rich fraction had the most abundant amounts of hemidesmosomal components (Fig. 4C). To examine if the culture condition that promotes deposition of HD proteins in DJM-1 cells is applicable to other cells, normal human epidermal keratinocytes, A431 human squamous carcinoma cells, and BxPC-3 human pancreatic adenocarcinoma cells [17] were cultured in KGM minus Ca. Among these cells, only A431 cells survived under the KGM minus Ca culture condition. Therefore, normal keratinocytes and BxPC-3 cells were cultured in KGM supplemented with 50 μM CaCl2 instead. After long-term culture (14 days), these cells were treated with ammonium solution to prepare matrix fractions. The proteins contained in these fractions were analyzed by SDS-PAGE and immunoblotting (Supplementary figure 1). A431 cells and normal human keratinocytes deposited larger amounts of laminin-332 than DJM-1 cells. However, among the three types of the cells, only the A431 cells showed substantial depositions of HD proteins.

Please cite this article as: Y. Hirako, et al., Isolation of a hemidesmosome-rich fraction from a human squamous cell carcinoma cell line, Exp Cell Res (2014), http://dx.doi.org/10.1016/j.yexcr.2014.04.002

E X PE R IM EN TA L C ELL R E S EA RC H

7

] (]]]]) ]]]–]]]

Table 1 – Identifying the proteins contained in the HD-rich fraction using liquid chromatography-tandem MS. Banda

No.b

Protein namec

Scored

Peptide hite

Nominal mass

Accession numberf

1

1 2 3

Plectin Perlecan Laminin α5

1156 120 50

132 12 5

518,172 466,305 178,103

1477646 11602963 3043590

2

1 2

BP230 Plectin

848 444

114 50

306,563 518,172

4502443 1477646

3

1 2

Integrin β4 BP230

460 168

57 17

207,485 306,563

6453380 4502443

4

1 2 3

Laminin α3 Plectin Integrin β4

350 286 148

49 30 15

147,912 518,172 207,485

1149515 1477646 6453380

5

1 2 3

BP180 BP230 Laminin α3

194 142 140

32 15 17

150,367 306,563 147,912

1877435 4502443 1149515

6

1 2 3

Laminin γ2 Plectin Laminin α3

352 276 150

54 28 18

130,857 518,172 147,912

452,755 1477646 1149515

7

1 2 3

Laminin β3 Integrin α6 Laminin α3

360 228 118

53 26 12

129,492 118,824 147,912

2429079 33942 1149515

8

1 2 3

Integrin α6 Pyruvate carboxylaseg Laminin γ2

240 170 140

28 17 15

118,824 129,552 130,857

33942 106049292 452755

a

: Band numbers as indicated in Fig. 4A. : Ranked in descending order of scores. c : Proteins identified by liquid chromatography-tandem MS. d : The score is based on the number of peptides and the probability. e : Number of valid peptide matches found for the given protein. f : Accession numbers from NCBI database. g : Pyruvate carboxylase is a mitochondorial protein. b

Effect of calcium concentration on processing, deposition, and secretion of laminin-332 The matrices prepared from DJM-1 cells cultured in KGM minus Ca (the HD-rich fraction) contained mostly unprocessed forms of the α3 (190 kDa) and γ2 (150 kDa) chains of laminin (Fig. 4D, left panels, arrows), whereas the matrices obtained from the cells cultured in EMEM and KGM plus Ca comprised the processed forms of α3 (160 kDa) and γ2 (105 kDa) chains of laminin (arrowheads) in addition to the unprocessed forms of the latter. Previous studies reported that unprocessed laminin-332 was preferentially deposited on the extracellular matrix [18–20]. In agreement with these observations, the relative amounts of the β3 chain in matrices isolated from the cells cultured in EMEM, KGM minus Ca, and KGM plus Ca were 58:100:38. This finding showed that the deposition of laminin-332 was the greatest for the HD-rich fraction from the cells cultured in KGM minus Ca. It is known that in cultured cells, laminin-332 is present in both an insoluble form deposited as a matrix and a soluble form secreted into the medium. To test whether the promoted deposition of laminin-332 resulted in a decrease of its soluble form in the medium, an immunoblot analysis was used to detect the α3, β3, and γ2 chains in conditioned media obtained from cells under three different culture conditions (Fig. 4D, right panels). The relative amounts of the β3 chain in conditioned media obtained

from cells cultured in EMEM, KGM minus Ca, and KGM plus Ca were 100:14:59, which indicated a large decrease of soluble laminin-332 in the KGM minus Ca condition. Metalloproteinases of the bone morphogenetic protein 1/mammalian Tolloid (BMP-1/mTLD) family are believed to be enzymes responsible for the processing of the α3 and γ2 chains of laminin332 [21,22]. Keratinocytes predominantly secrete mTLD, an alternatively spliced variant of the Bmp 1 among BMP-1 isoenzymes [23]. To determine whether the impaired processing of laminin332 in the KGM minus Ca condition was due to the absence or dysfunction of mTLD, immunoblotting was used to detect this processing enzyme in conditioned media obtained from DJM-1 cells cultured in KGM minus Ca and KGM plus Ca. For the KGM minus Ca condition, we detected a 130-kDa polypeptide using a PAb against BMP-1/mTLD that probably corresponded to an inactive precursor of mTLD with an N-terminal prodomain (Fig. 4E, arrow), whereas we detected a 120-kDa polypeptide that was probably an active processed form of mTLD for the KGM plus Ca condition (arrow head) [23].

Detection of Lutheran blood group protein/basal cell adhesion molecule in the HD-rich fraction In our SDS-PAGE analysis of the HD-rich fraction, we noted several minor bands between band 8 and keratins (see Fig. 4A).

Please cite this article as: Y. Hirako, et al., Isolation of a hemidesmosome-rich fraction from a human squamous cell carcinoma cell line, Exp Cell Res (2014), http://dx.doi.org/10.1016/j.yexcr.2014.04.002

8

] (]]]]) ]]]–]]]

E XP E RI ME N TAL CE L L R ES E ARC H

MS analysis of these minor polypeptides indicated the presence of the Lutheran blood group protein/basal cell adhesion molecule (Lu/B-CAM), which is an immunoglobulin superfamily transmembrane receptor and localizing at dermal–epidermal junctions in human skin (Fig. 5A) [24]. Immunoblotting results confirmed that this protein was in the HD-rich fraction with an approximate molecular mass of 100 kDa (Fig. 5B). To determine whether the Lu/B-CAM protein was an HD constituent, its localization on the HD-rich matrix was examined using double immunofluorescence microscopy (Fig. 5, C–H). A mouse MAb specific for the Lu/B-CAM protein showed a punctate staining pattern (Fig. 5C) that resembled HDs that were as visualized using a rat MAb against integrin α6 (Fig. 5D). However, a merged image (Fig. 5E, Lu/B-CAM in red and integrin α6 in green) showed that the Lu/B-CAM protein was adjacent to integrin α6 but did not colocalize with this integrin in most cases. On the other hand, the localization pattern of BP180 (Fig. 5F) nearly completely coincided with that of integrin α6 (Fig. 5G), as shown by the yellow color in the merged image (Fig. 5H, BP180 in red and integrin α6 in green). Lu protein is known as a receptor for laminins containing α5 chain (laminin-511/521). We examined if laminin chains other

than laminin-332 were included in the HD-rich fraction (Fig. 5I). Immunoblot analysis using PAb to laminin-111 and MAbs to laminin β1 and γ1 chains detected two bands at molecular mass of more than 200 kDa. The upper and lower polypeptides probably corresponded to β1 and γ1 chains of laminin, respectively. These results indicated that the HD-rich fraction also contained a nonhemidesmosomal adhesion receptor for laminins as a minor polypeptide localized at the dermal–epidermal junction.

Discussion In the epidermis, HDs are found in basal cells, which are in a less differentiated state compared with other cells in the upper layers. Because it is known that primary cultured keratinocytes remain in an undifferentiated state at low extracellular calcium concentrations (0.03–0.1 mM) [25], we postulated that DJM-1 cells would also remain in a less differentiated state under similar culture conditions, which would promote these cells to form HDs, similar to epidermal basal cells. In the present study, DJM-1 cells did form

Matrices KGM -Ca +Ca

EM

EM

Lu/Human skin 200 116

HD-rich

97 66 (kDa)

Lu

int α6

200

Blot: anti-Lu

Lu/α6

116 97 66

BP180

BP180/α6

-γ 1 LM

LM

int α6

LM

-1

11

CBB

-β 1

44

Blot

Fig. 5 – Inclusion of Lutheran blood group protein/basal cell adhesion molecule (Lu/B-CAM) in the HD-rich fraction. (A) MAb against Lu/B-CAM demonstrated that the Lu/B-CAM protein was localized at the dermal–epidermal junction (arrowheads) in human skin. Bar¼ 50 μm. (B) Rabbit PAb against Lu/B-CAM detected a 100-kDa band in the matrix obtained from cells cultured in KGM minus Ca. (C–H) HD-rich matrices were double immmunostained with a mouse MAb for Lu/B-CAM (C) and a rat MAb for integrin α6 (D) or a mouse MAb for BP180 (F) and a rat MAb for integrin α6 (G). Panel E is a merged image of C (red) and D (green). Panel H is a merged image of F (red) and G (green). Bar¼2 μm. (I) The HD-rich fraction was analyzed by immunoblotting with rabbit PAb against murine Engelbreth-Holm-Swarm-laminin (LM-111) and with mouse MAbs against laminin β1 (LM-β1) and γ1 (LM-γ1) chains. The PAb detected two bands at molecular mass of more than 200 kDa, which probably corresponded to β1 (arrowhead) and γ1 (arrow) chains of laminin, as detected by the MAbs. Please cite this article as: Y. Hirako, et al., Isolation of a hemidesmosome-rich fraction from a human squamous cell carcinoma cell line, Exp Cell Res (2014), http://dx.doi.org/10.1016/j.yexcr.2014.04.002

E X PE R IM EN TA L C ELL R E S EA RC H

mature HDs when cultured in KGM without supplemental calcium for more than 10 days after they had achieved confluent growth. However, “KGM without supplemental calcium” did not imply that this medium was completely calcium-free. Given that the calcium concentration of the supplemented bovine pituitary extract is 1–2 mM, we estimated that the calcium concentration of KGM (2 ml of bovine pituitary extract added to 500 ml of medium) was 4–8 μM. Under this culture condition, A431 cells survived and deposited laminin-332 and HD proteins, while normal human keratinocytes and BxPC-3 cells were not viable. This finding indicates that the culture condition used for DJM-1 cells is not applicable to all type of cells producing HD proteins, but only to some cells such as A431. In addition to the calcium concentration, there might be other factors that affect the deposition of HD proteins. Because DJM-1 cells long-term cultured with KGM plus Ca deposited larger amounts of HD proteins than the cells cultured with EMEM plus FCS (Fig. 4C), some factors included in the KGM may have had a positive effect on HD-protein deposition. Conversely, factors contained in FCS may have had a negative effect on the formation of HDs in the long-term culture of DJM-1 cells. In the KGM plus Ca condition, DJM-1 cells still deposited HD proteins in the absence of apparent electron-dense hemidesmosomal plaque structures. We speculate that hemidesmosomal proteins formed immature adhesion structures under the KGM plus Ca condition. Type II HDs consisting of integrin α6β4 and plectin are examples of hemidesmosomal adhesion structures that do not have electron-dense plaques [26–28]. Mutant mice deficient for BP230 or BP180 also form rudimentary HDs without distinctive plaque structures [29,30]. The reproducible formation/accumulation of HDs in DJM-1 cells cultured under the KGM minus Ca condition indicated that it was a good model for analyzing the biological processes involved in HD assembly in vitro. However, the intensity of band 1 for plectin was obviously increased in the matrix from KGM plus Ca culture (Fig. 4C). This may suggest that addition of CaCl2 at a very low concentration or transient cultivation in the KGM plus Ca could further promote the maturation of HDs in the long-term cultured DJM-1 cells. One aim of our study was to identify novel HD components. Here, the eight major polypeptides in the HD-rich fraction were found to be five known HD proteins and three laminin-332 chains using immunoblotting and MS. Because an HD fraction isolated from bovine corneas also contained these five major HD proteins [4], our results confirmed that the number of higher molecular weight HD components is five. At present, we have not yet identified a novel HD component from our analyses of minor polypeptides contained in the HD-rich fraction. However, Lu/BCAM was a nonhemidesmosomal protein in this fraction. Lu/BCAM is expressed in epithelial tissues, including epidermis, and is a receptor for laminins that contain the α5 chain [31]. In agreement with this, MS analysis detected the laminin α5 chain in band 1 (Table 1), β1 and γ1 chains in band 2, and β2 chain in band 3 of the HD-rich fraction. Because their scores were relatively low (highest scores for the α5, β1, β2, and γ1 chains were 50, 90, 70, and 60, respectively), the HD-rich fraction probably contained a small amount of laminin-511/521 complexes, which may serve as ligands for Lu/B-CAM in DJM-1 cells. One notable feature of this HD-rich fraction was that the α3 and γ2 chains of laminin were present in their unprocessed

] (]]]]) ]]]–]]]

9

forms. Our immunoblot results showed that the mTLD protease secreted by DJM-1 cells was in an inactive precursor form in the KGM minus Ca condition. The NH2-terminal prodomain of BMP1/mTLD is excised by furin-like proprotein convertases to yield the COOH-terminal active protease [12,32]. Furin is calcium dependent and requires 1 mM calcium for complete activity [33]. Therefore, the paucity of calcium in the KGM minus Ca condition might impair the enzymatic activity of furin-like proprotein convertases and resulted in the secretion of premature mTLD that was incompetent for processing the α3 and γ2 chains of laminin-332. It has been shown that laminin-332 molecules containing processing-resistant mutant α3 or γ2 chains were preferentially deposited in the matrix on cultured keratinocytes [18–20]. The deposition process of laminin-332 may involve cell surface sulfated glycosaminoglycans, but not laminin-332 binding integrins [34]. In our experiments, enhanced deposition of unprocessed laminin-332 was observed in the cells cultured in KGM minus Ca. In contrast to this deposition, DJM-1 cells cultured in KGM minus Ca secreted a much less soluble laminin-332 into the medium compared with cells cultured under the other two culture conditions. In addition, the secreted laminin-332 exclusively contained the α3 chain in the processed form. The immunofluorescence studies of DJM-1 cells not treated with ammonia solution showed a decreased staining of laminin α3 chain in the KGM plus Ca culture. The result excluded the possibility that the ammonia treatment removed the fully processed lamnin-332 and consequently, HD proteins were lost from the isolated matrix. Based on these observations, we speculate that the processing of α3 chain may facilitate the transition of laminin-332 from the matrix to the medium. Thus, under the normal calcium concentration, cells cannot maintain processed laminin-332 in the matrix deposited on a plastic dish in the long-term culture, which results in reduced HDs that use this laminin as a major extracellular ligand. In contrast, cells cultured in KGM minus Ca deposited the unprocessed form of laminin-332, which was minimally released from the matrix. The enhanced deposition of laminin might promote the formation/accumulation of HDs. We finally consider the possible applications of this HD-rich fraction for screening of the sera of patients with blistering skin diseases. Because the HD-rich fraction contained all five of the major HD components and three chains of laminin-332 in their full-length forms, the reactivities to these proteins can be tested at one time by immunoblotting. The enrichment of this fraction with HD proteins and laminin-332 would contribute to reducing the background staining, which is often problematic when immunoblotting with human patients' sera. In addition, this fraction probably contained nonhemidesmosomal proteins other than Lu/B-CAM that localize at the dermal–epidermal junction. An interesting example is that the γ1 chain of laminin, which is the autoantigen of anti-p200 pemphigoid [35], was detected in our HD-rich fraction by MS and immunoblot analyses. We anticipate that this fraction would become a powerful tool for identifying hemidesmosomal (and possibly nonhemidesmosomal) antigens targeted by patients' autoantibodies. This HD-rich fraction is easy to prepare and does not need special techniques or instruments, except for DJM-1 cells and the commercially available medium. Using this HD-rich fraction for serological tests in clinical laboratories would facilitate the diagnostic identification of autoantibodies in patients' sera.

Please cite this article as: Y. Hirako, et al., Isolation of a hemidesmosome-rich fraction from a human squamous cell carcinoma cell line, Exp Cell Res (2014), http://dx.doi.org/10.1016/j.yexcr.2014.04.002

10

E XP E RI ME N TAL CE L L R ES E ARC H

Acknowledgments We are grateful to Dr. Yasuo Kitajima (Kizawa memorial hospital, Japan) for sharing DJM-1 cells. We thank Dr. Takashi Hashimoto (Kurume University, Japan) for critical reading of the manuscript. This study was partly supported by the Grant-in-Aid for Scientific Research 20591341 (to Y. H.) from the Ministry of Education, Culture, Sports, Science and Technology of Japan.

Appendix A.

Supporting information

Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.yexcr.2014.04.002.

references [1] S.H. Litjens, J.M. de Pereda, A. Sonnenberg, Current insights into the formation and breakdown of hemidesmosomes, Trends Cell. Biol. 16 (2006) 376–383. [2] D. Sawamura, H. Nakano, Y. Matsuzaki, Overview of epidermolysis bullosa, J. Dermatol. 37 (2010) 214–219. [3] D. Zillikens, Acquired skin disease of hemidesmosomes, J. Dermatol. Sci. 20 (1999) 134–154. [4] K. Owaribe, Y. Nishizawa, W.W. Franke, Isolation and characterization of hemidesmosomes from bovine corneal epithelial cells, Exp. Cell Res. 192 (1991) 622–630. [5] M. Langhofer, S.B. Hopkinson, J.C. Jones, The matrix secreted by 804 G cells contains laminin-related components that participate in hemidesmosome assembly in vitro, J. Cell Sci. 105 (1993) 753– 764. [6] D. Gospodarowicz, Preparation of extracellular matrices produced by cultured bovine corneal endothelial cells and PF-HR-9 endodermal cells: their use in cell culture, in: D.W. Barnes, D.A. Sirbasku, G.H. Sato (Eds.), Methods for Preparation of Media, Supplements, and Substrata for Serum-Free Animal Cell Culture, Alan R. Liss, New York, 1984, pp. 275–293. [7] Y. Hirako, K. Yoshino, D. Zillikens, K. Owaribe, Extracellular cleavage of bullous pemphigoid antigen 180/type XVII collagen and its involvement in hemidesmosomal disassembly, J. Biochem. 133 (2003) 197–206. [8] M. Leverkus, K. Bhol, Y. Hirako, H. Pas, C. Sitaru, G. Baier, E.B. Brö cker, M.F. Jonkman, A.R. Ahmed, D. Zillikens, Cicatricial pemphigoid with circulating autoantibodies to beta4 integrin, bullous pemphigoid 180 and bullous pemphigoid 230, Br. J. Dermatol. 145 (2001) 998–1004. [9] J. Uematsu, Y. Nishizawa, Y. Hirako, K. Kitamura, J. Usukura, T. Miyata, K. Owaribe, Both type-I hemidesmosomes and adherens-type junctions contribute to the cell-substratum adhesion system in myoepithelial cells, Eur. J. Cell Biol. 84 (2005) 407–415. [10] Y. Kikkawa, R. Sudo, J. Kon, T. Mizuguchi, M. Nomizu, K. Hirata, T. Mitaka, Laminin alpha 5 mediates ectopic adhesion of hepatocellular carcinoma through integrins and/or Lutheran/basal cell adhesion molecule, Exp. Cell Res. 314 (2008) 2579–2590. [11] A. Sonnenberg, H. Janssen, F. Hogervorst, J. Calafat, J. Hilgers, A complex of platelet glycoproteins Ic and IIa identified by a rat monoclonal antibody, J. Biol. Chem. 262 (1987) 10376–10383. [12] R. Jasuja, G. Ge, N.G. Voss, J. Lyman-Gingerich, A.M. Branam, F.J. Pelegri, D.S. Greenspan, Bone morphogenetic protein 1 prodomain specifically binds and regulates signaling by bone morphogenetic proteins 2 and 4, J. Biol. Chem. 282 (2007) 9053–9062.

] (]]]]) ]]]–]]]

[13] Y. Kitajima, S. Inoue, S. Nagao, K. Nagata, H. Yaoita, Y. Nozawa, Biphasic effects of 12-O-tetradecanoylphorbol-13-acetate on the cell morphology of low calcium-grown human epidermal carcinoma cells: involvement of translocation and down regulation of protein kinase C, Cancer Res. 48 (1988) 964–970. [14] Y. Hirako, J. Usukura, Y. Nishizawa, K. Owaribe, Demonstration of the molecular shape of BP180, a 180-kDa bullous pemphigoid antigen and its potential for trimer formation, J. Biol. Chem. 271 (1996) 13739–13745. [15] A. Furukawa, Y. Kawamoto, Y. Chiba, S. Takei, S. Hasegawa-Ishii, N. Kawamura, K. Yoshikawa, M. Hosokawa, S. Oikawa, M. Kato, A. Shimada, Proteomic identification of hippocampal proteins vulnerable to oxidative stress in excitotoxin-induced acute neuronal injury, Neurobiol. Dis. 43 (2011) 706–714. [16] Y. Hirako, J. Usukura, J. Uematsu, T. Hashimoto, Y. Kitajima, K. Owaribe, Cleavage of BP180, a 180-kDa bullous pemphigoid antigen, yields a 120-kDa collagenous extracellular polypeptide, J. Biol. Chem. 273 (1998) 9711–9717. [17] T. Tani, A. Lumme, A. Linnala, E. Kivilaakso, T. Kiviluoto, R.E. Burgeson, L. Kangas, I. Leivo, I. Virtanen, Pancreatic carcinomas deposit laminin-5, preferably adhere to laminin-5, and migrate on the newly deposited basement membrane, Am. J. Pathol. 151 (1997) 1289–1302. [18] L. Gagnoux-Palacios, M. Allegra, F. Spirito, O. Pommeret, C. Romero, J.P. Ortonne, G. Meneguzzi, The short arm of the laminin gamma2 chain plays a pivotal role in the incorporation of laminin 5 into the extracellular matrix and in cell adhesion, J. Cell Biol. 153 (2001) 835–850. [19] R.O. Sigle, S.G. Gil, M. Bhattacharya, M.C. Ryan, T.M. Yang, T.A. Brown, A. Boutaud, Y. Miyashita, J. Olerud, W.G. Carter, Globular domains 4/5 of the laminin alpha3 chain mediate deposition of precursor laminin 5, J. Cell Sci. 117 (2004) 4481–4494. [20] D.E. Frank, W.G. Carter, Laminin 5 deposition regulates keratinocyte polarization and persistent migration, J. Cell Sci. 117 (2004) 1351–1363. [21] S. Amano, I.C. Scott, K. Takahara, M. Koch, M.F. Champliaud, D.R. Gerecke, D.R. Keene, D.L. Hudson, T. Nishiyama, S. Lee, D.S. Greenspan, R.E. Burgeson, Bone morphogenetic protein 1 is an extracellular processing enzyme of the laminin 5 gamma 2 chain, J. Biol. Chem. 275 (2000) 22728–22735. [22] Y. Tsubota, H. Mizushima, T. Hirosaki, S. Higashi, H. Yasumitsu, K. Miyazaki, Isolation and activity of proteolytic fragment of laminin-5 alpha3 chain, Biochem. Biophys. Res. Commun. 278 (2000) 614–620. [23] D.P. Veitch, P. Nokelainen, K.A. McGowan, T.T. Nguyen, N.E. Nguyen, R. Stephenson, W.N. Pappano, D.R. Keene, S.M. Spong, D.S. Greenspan, P.R. Findell, M.P. Marinkovich, Mammalian tolloid metalloproteinase, and not matrix metalloprotease 2 or membrane type 1 metalloprotease, processes laminin-5 in keratinocytes and skin, J. Biol. Chem. 278 (2003) 15661–15668. [24] M. Schö n, C.E. Klein, V. Hogenkamp, R. Kaufmann, B.G. Wienrich, M.P. Schon, Basal-cell adhesion molecule (B-CAM) is induced in epithelial skin tumors and inflammatory epidermis, and is expressed at cell–cell and cell–substrate contact sites, J. Invest. Dermatol. 115 (2000) 1047–1053. [25] H. Hennings, D. Michael, C. Cheng, P. Steinert, K. Holbrook, S.H. Yuspa, Calcium regulation of growth and differentiation of mouse epidermal cells in culture, Cell 19 (1980) 245–254. [26] Y. Hieda, Y. Nishizawa, J. Uematsu, K. Owaribe, Identification of a new hemidesmosomal protein, HD1: a major, high molecular mass component of isolated hemidesmosomes, J. Cell. Biol. 116 (1992) 1497–1506. [27] J. Uematsu, Y. Nishizawa, A. Sonnenberg, K. Owaribe, Demonstration of type II hemidesmosomes in a mammary gland epithelial cell line, BMGE-H, J. Biochem. (Tokyo). 115 (1994) 469–476. [28] L. Fontao, J. Stutzmann, P. Gendry, J.F. Launay, Regulation of the type II hemidesmosomal plaque assembly in intestinal epithelial cells, Exp. Cell Res. 250 (1999) 298–312.

Please cite this article as: Y. Hirako, et al., Isolation of a hemidesmosome-rich fraction from a human squamous cell carcinoma cell line, Exp Cell Res (2014), http://dx.doi.org/10.1016/j.yexcr.2014.04.002

E X PE R IM EN TA L C ELL R E S EA RC H

[29] W. Nishie, D. Sawamura, M. Goto, K. Ito, A. Shibaki, J.R. McMillan, K. Sakai, H. Nakamura, E. Olasz, K.B. Yancey, M. Akiyama, H. Shimizu, Humanization of autoantigen, Nat. Med. 13 (2007) 378–383. [30] L. Guo, L. Degenstein, J. Dowling, Q.C. Yu, R. Wollmann, B. Perman, E. Fuchs, Gene targeting of BPAG1: abnormalities in mechanical strength and cell migration in stratified epithelia and neurologic degeneration, Cell 81 (1995) 233–243. [31] C.L. Moulson, C. Li, J.H. Miner, Localization of Lutheran, a novel laminin receptor, in normal, knockout, and transgenic mice suggests an interaction with laminin alpha5 in vivo, Dev. Dyn. 222 (2001) 101–114.

] (]]]]) ]]]–]]]

11

[32] M. Leighton, K.E. Kadler, Paired basic/Furin-like proprotein convertase cleavage of Pro-BMP-1 in the trans-Golgi network, J. Biol. Chem. 278 (2003) 18478–18484. [33] G. Thomas, Furin at the cutting edge: from protein traffic to embryogenesis and disease, Nat. Rev. Mol. Cell Biol. 3 (2002) 753–766. [34] Y. Kariya, H. Sato, N. Katou, Y. Kariya, K. Miyazaki, Polymerized laminin-332 matrix supports rapid and tight adhesion of keratinocytes, suppressing cell migration, PLoS One 7 (2012) e35546. [35] T. Dainichi, S. Kurono, B. Ohyama, N. Ishii, N. Sanzen, M. Hayashi, C. Shimono, Y. Taniguchi, H. Koga, T. Karashima, S. Yasumoto, D. Zillikens, K. Sekiguchi, T. Hashimoto, Anti-laminin gamma-1 pemphigoid, Proc. Natl. Acad. Sci. USA 106 (2009) 2800–2805.

Please cite this article as: Y. Hirako, et al., Isolation of a hemidesmosome-rich fraction from a human squamous cell carcinoma cell line, Exp Cell Res (2014), http://dx.doi.org/10.1016/j.yexcr.2014.04.002