QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

Vol. 18, Issue 9, 3512-3522, September 2007

This Article Abstract Full Text (PDF) Supplemental Material All Versions of this Article: E07-04-0306v1 18/9/3512    most recent Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Wilhelmsen, K. Articles by Sonnenberg, A. Search for Related Content PubMed PubMed Citation Articles by Wilhelmsen, K. Articles by Sonnenberg, A.

Serine Phosphorylation of the Integrin 4 Subunit Is Necessary for Epidermal Growth Factor Receptor–induced Hemidesmosome Disruption

Division of Cell Biology, The Netherlands Cancer Institute, 1066 CX Amsterdam, The Netherlands

Submitted April 5, 2007; Accepted June 27, 2007
Monitoring Editor: Mark Ginsberg

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Hemidesmosomes (HDs) are multiprotein adhesion complexes that promote attachment of epithelial cells to the basement membrane. The binding of 64 to plectin plays a central role in their assembly. We have defined three regions on 4 that together harbor all the serine and threonine phosphorylation sites and show that three serines (S1356, S1360, and S1364), previously implicated in HD regulation, prevent the interaction of 4 with the plectin actin-binding domain when phosphorylated. We have also established that epidermal growth factor receptor activation, which is known to function upstream of HD disassembly, results in the phosphorylation of only one or more of these three residues and the partial disassembly of HDs in keratinocytes. Additionally, we show that S1360 and S1364 of 4 are the only residues phosphorylated by PKC and PKA in cells, respectively. Taken together, our studies indicate that multiple kinases act in concert to breakdown the structural integrity of HDs in keratinocytes, which is primarily achieved through the phosphorylation of S1356, S1360, and S1364 on the 4 subunit.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Hemidesmosomes (HDs) are adhesive superstructures present primarily in the basal cells of epithelial layers. They guarantee stable adhesion of these cells to the underlying basement membrane. HDs are present in complex and (pseudo) stratified epithelia and are composed of at least six proteins: the two subunits of the integrin 64 (Stepp et al., 1990; Jones et al., 1991; Sonnenberg et al., 1991), the bullous pemphigoid antigens 180 (Giudice et al., 1992) and 230 (BP230; Stanley et al., 1981), the cytoskeletal linker protein plectin (Hieda et al., 1992; Gache et al., 1996), and the tetraspanin CD151 (Sterk et al., 2000). The integrin 64 binds to its ligand laminin-332 (laminin-5) in the basement membrane, whereas the two members of the plakin family, plectin and BP230, link the cell surface receptors 64 and BP180 to the intermediate filament cytoskeleton. This ensures resistance of the adhesive contacts of keratinocytes to mechanical separation.

Although HDs present in skin keratinocytes, bladder epithelial cells, and myoepithelial cells contain all six hemidesmosomal proteins (type I), those present in intestinal epithelial cells contain only the integrin 64 and plectin (type II; Uematsu et al., 1994; Orian-Rousseau et al., 1996). The binding of plectin to 64 is essential for the integrity of the adhesive complex in both HD types, as indicated by patients with mutations in the 4 integrin subunit that prevent this interaction (reviewed in Litjens et al., 2006). Furthermore, it has previously been reported that the recruitment of plectin by 64 is one of the first steps in type I HD assembly (Koster et al., 2004). The actin-binding domain (ABD) of plectin binds to the first pair of fibronectin type III (FnIII) domains and a small region of the connecting segment (CS) that separates the first pair of FnIII domains from the second pair on 4 (Geerts et al., 1999). This interaction is strengthened by the binding of the plectin plakin domain to the CS and C-tail of 4 (Koster et al., 2004; Rezniczek et al., 1998). Additionally, in vitro analysis clearly indicates that if the interaction between plectin and 64 is disrupted, HD integrity is compromised (Geerts et al., 1999; Koster et al., 2001).

Although HDs are adhesive protein complexes, they nevertheless are dynamic structures because their components are rearranged during cell processes such as migration and division. However, the interaction between plectin and 4 reduces the dynamics when laminin-332 is clustered beneath HDs (Geuijen and Sonnenberg, 2002). Epidermal growth factor (EGF), macrophage-stimulating protein (MSP), and phorbol 12-myristate acetate (PMA) have been implicated in the disassembly of HDs through signaling pathways that result in the phosphorylation of the 4 cytoplasmic domain by protein kinase C (PKC). Activation of PKC downstream of the EGF receptor has been suggested to lead to phosphorylation of serine residues in the 4 CS that results in a translocation of 64 to lamellipodia, where it associates with actin-rich protrusions (Rabinovitz et al., 1999, 2004). Moreover, activation of PKC downstream of the Ron (MSP) receptor also results in a mobilization of 64 to actin-rich protrusions in lamellipodia, where it is reported to associate with the Ron receptor through 14-3-3 proteins (Santoro et al., 2003). However, it is clear that other kinases, or mechanisms, are involved in the disassembly of HDs because EGF, MSP, and PMA did not cause complete disassembly of them as appears in these reports.

Previously, Rabinovitz et al. (2004) showed that the serine residues S1356, S1360, and S1364 in the 4 CS are phosphorylated downstream of the EGF receptor, and they suggest that this is mediated by PKC. Moreover, substitution of these three residues by aspartic acid results in the partial loss of the colocalization of plectin with 64 in transfected COS-7 cells. Therefore, we decided to investigate the role of these serines in the disassembly of HDs under more physiological conditions in keratinocytes. By phosphatase inhibition assays three regions of 4 were identified that were heavily phosphorylated on serine and threonine residues in cells, including the region of the CS containing S1356, S1360, and S1364. We further show that activation of endogenous PKC results in the phosphorylation of only S1360.

We also show through phosphomimicry studies that phosphorylation of two or more of these serines prevents the primary interaction of the plectin ABD with 4. Furthermore, we identified S1364 as an endogenous PKA phosphorylation site in keratinocytes and showed that its activation in conjunction with PKC reduces the ability of the plectin ABD to associate with 4. And finally, we established that activation of the EGF receptor results in the phosphorylation of only these three residues in keratinocytes, which leads to a partial disassembly of HDs. Taken together, our results suggest that in keratinocytes, multiple kinases act in concert to induce the disassembly of HDs through the phosphorylation of S1356, S1360, and S1364.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Cell Lines
PA-JEB (4-null) keratinocytes have been described previously (Schaapveld et al., 1998) and were maintained in keratinocyte serum-free medium (SFM; Invitrogen, Rockville, MD) supplemented with 50 µg/ml bovine pituitary gland extract, 5 ng/ml EGF, 100 U/ml penicillin, and 100 U/ml streptomycin. Stable expression of 4 mutants in cells was performed as described previously (Sterk et al., 2000; Geuijen and Sonnenberg, 2002). COS-7 cells were grown in DMEM (Invitrogen) containing 10% fetal bovine serum, 100 U/ml penicillin, and 100 U/ml streptomycin. They were transiently transfected using the DEAE-dextran method (Seed and Aruffo, 1987).

4 and 4 Phospho-specific Antibodies
4 rabbit polyclonal Abs were raised against the first pair of FNIII repeats and part of the CS (4A residues 1115–1355) fused to an amino-terminal glutathione S-transferase (GST) moiety. GST preparations were done as previously described (Wilhelmsen et al., 2002), and fusion proteins were eluted from glutathione-Sepharose beads using 15 mM reduced glutathione. The GST moiety was removed by Factor VIII protease treatment followed by column chromatography and subsequently, the 4 fragment was injected into rabbits four times.

The phospho-specific 4 rabbit polyclonal antibodies (4 pS-CS) were raised against a synthetic peptide (sequence: CAQSGEDYDSFLMYSDDVLRpSPSGpSQRPpSVSDD) coupled to an activated mcKLH moiety (Pierce, Rockford, IL) and subsequently, the coupled peptides were injected into rabbits four times. The antibodies were first affinity-purified on a column containing SulfaLink beads (Pierce) cross-linked to the phosphorylated peptides, eluted with 1 M glycine-HCl, pH 2.7, and immediately neutralized with an equal volume of 5% NHCO₃. Bovine serum albumin (1%) was then added, and the protein mixture was dialyzed overnight against phosphate-buffered saline (PBS). Finally, the dialyzed antibody sample was run through a column containing SulfaLink beads (Pierce) cross-linked to the equivalent nonphosphorylated peptides and the nonbound fraction collected and stored at 4°C with azide.

cDNA Constructs
For PCR the proofreading Pwo DNA polymerase (Roche Molecular Biochemicals, Indianapolis, IN) was used. All plasmids were verified by sequencing, and protein expression and sizes were confirmed by Western blotting. The full-length 4 cDNA in pUC18 has been described previously (Niessen et al., 1997). Point mutants were generated by the PCR overlap extension method. Subsequently 4 was isolated using the EcoRI restriction site and it was ligated into the retroviral LZRS-IRES-zeo vector as described before (Sterk et al., 2000). BssHII-EcoRV fragments were used for exchange with fragments from IL2R-4 in pCMV (Nievers et al., 1998) to generate IL2R-4 mutants. The plasmid pCMV contains a Cytomegalovirus LTR directing the transcription of cDNA sequences. Full-length 4 and deletion mutants of 4 in pCMV have been described before (Niessen et al., 1997). Also, the plectin-1A ABD in pcDNA3-HA (hemagglutinin) has been previously described (Litjens et al., 2003).

Phosphopeptide Mapping
For in vivo phosphopeptide mapping experiments, PA-JEB or COS-7 cells expressing 4 were phosphate-starved for 45 min at 37°C. Subsequently, 2 mCi of [³²P]orthophosphate was added and incubated for 3 h at 37°C. Calyculin A (100 nM; Cell Signaling, Beverly, MA), 1 µM PMA (Sigma-Aldrich, St. Louis, MO), 25 µM forskolin (FSK; Calbiochem, San Diego, CA), and 0.1 mM 3-isobutyl-1-methylxanthine (IBMX; Calbiochem) or 50 ng/ml EGF (Sigma-Aldrich) were added for 30 min at 37°C. 4 was immunoprecipitated with the mAb 450-11A (PharMingen International, San Diego, CA). For in vitro phosphopeptide mapping experiments, IL2R-4 was immunoprecipitated with 450-11A from lysates of transfected COS-7 cells. Subsequently, the isolated protein was phosphorylated in vitro with 20 U of PKA (Sigma-Aldrich) in the presence of 50 µCi [-³²P]ATP, 40 mM HEPES, pH 7.5, 3 mM MnCl₂, 10 mM MgCl₂, and 1 mM dithiothreitol (DTT) for 2 h at 30°C. The in vivo and in vitro samples were subjected to SDS-PAGE, and the gel was dried. The film was exposed at room temperature for 10 min for in vitro phosphorylations or 1 h for in vivo phosphorylations. Radioactive 4 was isolated from the gel and digested with trypsin. Phosphopeptide mapping was performed as described previously (van der Geer et al., 1993; Wilhelmsen et al., 2002).

Immunofluorescence
PA-JEB/4 keratinocytes were seeded on glass coverslips in keratinocyte SFM. After 24 h the medium was replaced by high-calcium medium (Ham's F12/DMEM, 1:3). After overnight incubation, cells were stimulated with 50 ng/ml EGF (Sigma-Aldrich) for 30 min at 37°C. Cells were fixed and permeabilized as described before (Litjens et al., 2003). Primary antibodies used were the rat mAb GoH3 against 6 (Sonnenberg et al., 1987) and the mouse mAb 121 against plectin/HD1 (Hieda et al., 1992), which was a generous gift from Dr. K. Owaribe (Nagoya University, Nagoya, Japan). Secondary antibodies were fluorescein isothiocyanate (FITC)-conjugated goat anti-rat antibodies, purchased from Rockland (Gilbertsville PA), and Cy5-coupled donkey anti-rat and Texas Red–coupled donkey anti-mouse antibodies, obtained from Jackson ImmunoResearch Laboratories (West Grove, PA). Secondary antibodies were selected for minimal cross-reactivity against the other species. The coverslips were mounted onto glass slides in Mowiol mounting medium (Calbiochem) containing 2.5% DABCO (Sigma-Aldrich). Immunofluorescence images were taken using a Leica confocal laser scanning microscope (Deerfield, IL).

Scatter Plots and Colocalization Images
Colocalization between 6 and plectin images (eight-bit grayscale values) was investigated by a custom-made Visual Basic (v6.0) program (Microsoft Corp.). Pixel intensity for 6 was plotted against the pixel intensity of plectin in a scatter plot. Multiple occurrences of identical coordinates are color coded as indicated in Figure 5. For a pair of colocalized images, the scatter plot shows a distribution along the diagonal with some divergence because of photon noise. Noncolocalized structures appear as off-diagonal clusters. Superimposed on a grayscale overlay image, the bright colocalized pixels, the noncolocalized 6 pixels, and the noncolocalized plectin pixels are shown in color as indicated in Figure 5.

Pulldown and Phospho-specific Antibody Assays
COS-7 cells were transfected with indicated cDNA constructs. When indicated, COS-7 or PA-JEB/4 cells were stimulated with either 1 µM PMA (Sigma-Aldrich), 25 µM FSK (Calbiochem), and 0.1 mM IBMX (Calbiochem) or 50 ng/ml EGF (Sigma-Aldrich) for 30 min (or for the indicated times in the time-course stimulation assays) at 37°C. Cells were lysed in M-PER buffer (Mammalian Protein Extraction Reagent; Pierce), containing 1 mM sodium vanadate and a cocktail of protease inhibitors (Sigma-Aldrich). Immunoprecipitations were performed with mAb HA 12CA5 (Santa Cruz Biotechnology, Santa Cruz, CA) and immunoblottings were done with either pAb anti-IL2R (sc-665; Santa Cruz Biotechnology), pAb anti-HA (sc-805; Santa Cruz Biotechnology), mAb anti-pTyr (4G10), pAb anti-EGF receptor (Ab-17; Neomarkers or 1005; Santa Cruz Biotechnology), mAb anti-EGFR pY845 (Cell Signaling, Beverly, MA), a mixture of mAbs antibodies against PKC, , and (Transduction Laboratories, Lexington, KY) or pAb anti-phospho-PKC (pan; betaII Ser 660; 9371; Cell Signaling) as described previously (Litjens et al., 2003). Gö6983 (Sigma-Aldrich) was used at a concentration of 100 nM for 30 min before PMA addition.

Migration Assays
PA-JEB, PA-JEB/4^WT, PA-JEB/4^{S1356/1360/1364A} (4^3xA), and PAJEB/4^{S1356/1360/1364D} (4^3xD) keratinocytes were grown to confluence in 24-well tissue culture plates in normal keratinocyte growth medium. Cells were then serum-starved overnight in keratinocyte-SFM and treated with 10 µg/ml mitomycin C (Sigma) 2 h before wounding. Monolayers were scratched with a yellow pipette tip and washed twice in keratinocyte-SFM to remove cell debris, before stimulation with EGF (5 ng/ml). Cells were then incubated at 37°C, and scratched areas were photographed during at least 48 h at 10x magnification. Relative migration was calculated from the scratch area at t = 48 over the scratch area at t = 0 and averaged from three independent experiments done in duplicate.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Phosphorylation of Serine/Threonine on 4 Occurs Primarily in the CS
To define the regions of the 4 subunit that are serine- and threonine-phosphorylated, various deletion mutants of the 4 cytoplasmic domain were expressed in COS-7 cells, and subsequently the cells were treated with calyculin A in the presence of [³²P]orthophosphate. Calyculin A inhibits protein phosphatase-1 and -2A (Ishihara et al., 1989), which together account for most of the serine/threonine phosphatase activity in a cell. Their inhibition results in an increase in the serine and threonine phosphorylation of proteins. The 4 deletion mutants were then isolated and subjected to phosphopeptide mapping. The results show that by deleting the C-terminal tail (4¹⁶⁷¹) two peptides are no longer phosphorylated. Further deletion of the cytoplasmic domain until amino acid 1487 (4¹⁴⁸⁷) does not result in the obvious loss of phosphorylation of other peptides. However, when the third FnIII domain and part of the CS were deleted (4¹³⁸²), five peptides were no longer phosphorylated. The three remaining phosphopeptides were no longer present after further deletion of the 4 cytoplasmic domain to residue 1355 (4¹³⁵⁵; Figure 1A). These results indicate that most of the serine and threonine phosphorylations on 4 occur in the region between residues 1487 and 1355, which comprises a large part of the CS, with some also occurring in the C-terminal tail (Figure 1B).

View larger version (31K): [in this window] [in a new window]   Figure 1. Serine/threonine phosphorylation of 4 occurs predominantly in the CS and the C-terminal tail. (A) In vivo phosphopeptide maps of wild type (A) and four sequentially truncated (B–E) 4 subunits derived from COS-7 cells that were treated with calyculin A in the presence of [32P]orthophosphate. (F) A schematic diagram of the results shown in A–E. (B) Schematic drawings of the integrin 64 depicting the regions serine and threonine phosphorylated on the 4 subunit after calyculin A treatment and the interaction sites with plectin reported in the literature. Residue numbers indicate where the 4 deletions were made. Roman numerals I–IV show the positions of the four FnIII domains. The CS is between FnIII II and III, whereas the C-terminal tail is after FnIII IV. CalX, Ca-Na exchanger motif. 4 numbering is based on the human 4A sequence.

Phosphorylation of S1356, S1360, and S1364 Disrupts the Bond between 4 with the Plectin-ABD

Curiously, the primary site on 4 that interacts with the plectin ABD lies outside the regions that are subject to phosphorylation, whereas those that interact with the plakin domain do not (Figure 1B). As we showed previously, the primary interaction between plectin and 4 occurs between the first pair of FnIII domains on 4 and the plectin ABD (Geerts et al., 1999). In fact, this interaction is sufficient for the formation of normal HD structures in cells (Niessen et al., 1997; Nikolopoulos et al., 2005). The plakin domain of plectin in turn associates with the CS and C-terminal tail of 4, but these regions in 4 are not essential for the localization of plectin in HDs (Niessen et al., 1997; Schaapveld et al., 1998). Therefore, we decided to concentrate on the question whether phosphorylation of S1356, S1360, and/or S1364 influences the binding of the ABD to 4. To this end, we performed coimmunoprecipitation experiments in COS-7 cells using an HA-plectin-1A ABD construct in conjunction with IL2R-4 fusion proteins containing all possible combinations of aspartic acid residue substitutions of these three serine residues. The results show that the single point mutants IL2R-4^S1356D and IL2R-4^S1360D did not bind as well to the plectin-1A ABD as either the wild-type IL2R-4 or the IL2R-4^S1364D mutant (Figure 2A), indicating that the phosphorylation of a single residue is not sufficient to prevent the binding of 4 to the ABD. However, when two, or all three, of these residues are replaced by aspartic acid, binding between the two proteins does not occur (Figure 2, A and B). Importantly, binding occurs, and possibly increases, when the three serines are mutated to alanine (Figure 2B), indicating that the loss of binding is not due to mutation of this area of 4 per se. These results indicate that phosphorylation of any two, or all three, of these serine residues on 4 can prevent the interaction with the plectin-1A ABD.

View larger version (29K): [in this window] [in a new window]   Figure 2. Phosphomimick aspartic acid substitutions of S1356, S1360, and S1364 in the 4 CS prevents the association with the Plectin-1A ABD. (A) COS-7 cells were transiently transfected with either IL2R-4WT (lanes 1 and 6), IL2R-4S1356D (lane 2), IL2R-4S1360D (lane 3), IL2R-4S1364D (lane 4), or IL2R-43xD (lane 5) cDNA constructs or a control plasmid (lane 7) and an expression construct for the HA-plectin-1A ABDWT (lanes 1–5 and 7) or a control plasmid (lane 6). The cells were lysed in M-PER buffer and HA-IPs were probed for the presence of IL2R-4 (top) and the HA-tagged ABDs (bottom). Whole cell lysates (WCLs) were probed for the expression level of IL2R-4 proteins (middle). (B) The experiment was performed as in A, except that either IL2R-4WT (lanes 1 and 7), IL2R-4S1356/1360D (lane 2), IL2R-4S1360/1364D (lane3), IL2R-4S1356/1364D (lane 4), IL2R-43xD (lane 5), or IL2R-43xA (lane 6) cDNA constructs, or a control plasmid (lane 8), were instead cotransfected with the HA-tagged plectin-1A ABD cDNA construct (lanes 1–6 and 8) or a control plasmid (lane 7).

View larger version (65K): [in this window] [in a new window]   Figure 3. S1360 and S1364 are in vivo PKC and PKA phosphorylation sites on 4, respectively, that cooperate to reduce the association with the plectin-1A ABD. (A) In vivo phosphopeptide maps of 4 isolated from either PA-JEB/4WT cells that were serum-starved (A), treated with PMA (B) or calyculin A (D), PA-JEB/4S1360A cells that were treated with PMA (C) or PA-JEB/43xA cells that were treated with calyculin A (E) in the presence of [32P]orthophosphate. (F) A schematic diagram of the results shown in A–E. Gray ovals, peptides containing 4 specific phosphorylation sites. (B) PA-JEB/4WT cells were either left serum-starved (lane 1), treated with PMA for 30 min (lane 2), or treated with PMA for 30 min in the presence of the compound Gö6983 (lane 3) before lysis in mPER buffer. Similarly, PA-JEB/43xA (lanes 4 and 5) and PA-JEB/43xD (lanes 6 and 7) cells were either left serum-starved (lanes 4 and 6) or treated with PMA for 30 min (lanes 5 and 7) before lysis. WCLs were probed for the presence of 4 phosphorylated on residues S1356, S1360, and/or S1364 using our rabbit polyclonal phospho-specific antibody 4 pS-CS (top). The expression level of the 4 proteins was detected using a rabbit polyclonal antibody raised against the first pair of FNIII repeats in 4 (middle). To verify the activation of PKC, WCLs were also probed for the presence of phospho-PKC using pan antibodies raised against phosphorylated Thr 660 of PKC II (bottom). (C) Alignment of the human 4A sequence from residue 1352-1368 with the same region in cow, dog, mouse, rat, chicken, and zebrafish. The kinases identified from the phospho-phosphorylation site prediction programs phospho-ELM and GPS are shown below their corresponding residue. The human 4 sequence was used as the input. (D) In vivo phosphopeptide maps of 4 isolated from either serum-starved PA-JEB/4WT cells (A) or PA-JEB/4WT (B), PA-JEB/4S1360/1364A (C) or PA-JEB (an equivalent area of gel were 4 would run was excised; D) cells treated for 30 min with FSK/IBMX in the presence of [32P]orthophosphate. (E) A schematic diagram of the results shown in A–D. Note that the peptides in these maps migrate to a lower part on the thin-layer chromatography plates because the chromatography buffer used was not optimal. Gray ovals, peptides containing 4 specific phosphorylation sites. (E) PA-JEB/4WT cells were either serum-starved (lane 1) or treated with Forskolin/IBMX for the indicated times (in minutes; lanes 2–7) before lysis in mPER buffer. WCLs were probed for the presence of 4 phosphorylated on residues S1356, S1360, and/or S1364 using our rabbit polyclonal phospho-specific antibody 4 pS-CS (top) and the expression level of the 4 proteins with our rabbit polyclonal antibody raised against the first pair of FNIII repeats in 4 (bottom). (F) COS-7 cells were transiently transfected with either IL2R-4WT (lanes 1–4 and 7) or IL2R-4S1360/1364A (lanes 5 and 6) cDNA constructs or a control plasmid (lane 8), and an expression construct for the HA-plectin-1A ABDWT (lanes 1–6 and 8) or a control plasmid (lane 7). The cells were left untreated (lanes 1, 5, 7 and 8) or treated with either FSK/IBMX (lane 2), PMA (lane 3), or all three compounds together (lanes 4 and 6) 30 min before lysis in mPER buffer. HA-IPs were probed for the presence of IL2R-4 (top) and the HA-tagged ABDs (middle). WCLs were probed for the expression level of IL2R-4 proteins (bottom). Quantitation was done in ImageJ and is a ratio of the band intensity shown in the top panel to those in the middle panel for each lane, relative to lane 1.

In search of other kinases that may phosphorylate this region of 4, we again used GPS and phospho-ELM and identified many other kinases that could in principle phosphorylate serines 1356 and 1364, but also 1360 (Figure 3C). We paid particular attention to PKA because this kinase appeared to be the only one capable of phosphorylating S1364 and because there are compounds, FSK and IBMX, that specifically activate this kinase. We first tested the ability of FSK/IBMX to induce phosphorylation of S1364 in cells by using the previously mentioned 4^WT-expressing keratinocytes and keratinocytes expressing a 4^S1360/1364A double mutant in the presence of [³²P]orthophosphate. We used this double mutant to lower the level of background phosphorylation, which would facilitate interpretation of the phosphopeptide maps. Indeed, we see that FSK/IBMX induces phosphorylation of two tryptic peptides, which was abolished by mutation of S1360 and S1364 (Figure 3D). The results were later verified in in vitro phosphopeptide mapping experiments that conclusively show that S1364 is the only PKA phosphorylation site on 4 (Supplementary Figure 1). We also utilized our 4 phospho-specific rabbit polyclonal antibody (4 pS-CS) to independently determine whether PKA can indeed induce phosphorylation of 4. To this end, a FSK/IBMX time course was performed on 4^WT-expressing keratinocytes. The results show that phosphorylation of 4 is maximal at around 30 min after PKA activation, after which it diminishes to background levels by 2 h (Figure 3E). After taking into account the specificity of our new antibody, this data again indicates that the 4 subunit is phosphorylated on S1364 by PKA in keratinocytes.

The results from Figures 2B and 3, A–E, suggest that activation of PKA and PKC should be sufficient to prevent the association of 4 with the plectin-1A ABD. To prove this, we transiently transfected cDNAs expressing the IL2R-4^WT and HA-plectin-1A ABD proteins into COS-7 cells and stimulated the cells with PMA and/or FSK/IBMX. We also utilized an IL2R-4 construct containing S1360A and S1364A point mutations (IL2R-4^S1360/1364A) as a control because this 4 subunit mutant is not phosphorylated by either PKC or PKA in cells. The results show that PMA and FSK/IBMX by themselves do not influence the binding of the plectin-1A ABD to 4, but together they reduce binding, although only to a degree (40% reduction; Figure 3F). Importantly, the reduction in binding was less with the IL2R-4^S1360/1364A mutant, than for the IL2R-4^WT, after PMA and FSK/IBMX treatment (15% reduction; Figure 3F). This indicates that the loss in binding of IL2R-4^WT is in fact mostly due to phosphorylation of S1360 and S1364 by PKC and PKA, respectively.

Effects of EGF Receptor Activation on HD Integrity
The above results suggest the involvement of PKC and PKA in the regulation of the interaction between 4 and plectin in COS-7 cells, but there are other kinases that are predicted to phosphorylate S1356, S1360, and S1364. However, the involvement of these other kinases is difficult to determine because of the lack of activators that specifically activate them, and we therefore decided to focus on the above three serine residues in a more physiological context. Activation of the EGF receptor is known to induce HD disassembly and has been shown to activate some of the kinases capable of phosphorylating S1356, S1360, and/or S1364 (Oda et al., 2005). Therefore, we first tested the ability of EGF to induce phosphorylation of 4 on S1356, S1360, and S1364 in keratinocytes. The 4^WT- or 4^3xA-expressing cells were stimulated with EGF in the presence of [³²P]orthophosphate and phosphopeptide mapping was performed on the isolated 4 subunits. The results indicate that after activation of the EGF receptor, one or more of these residues are phosphorylated in cells and importantly, these serine residues are the only ones phosphorylated (Figure 4A).

View larger version (61K): [in this window] [in a new window]   Figure 4. EGF induces phosphorylation of 4 only on S1356, S1360, and S1364, which reduces the affinity of 4 for the plectin-1A ABD. (A) In vivo phosphopeptide maps of 4 isolated from either serum-starved PA-JEB/4WT cells (A) or PA-JEB/4WT (B), PA-JEB/43xA (C), or PA-JEB (an equivalent area of gel where 4 would run was excised, D) cells treated for 30 min with EGF in the presence of [32P]orthophosphate. (E) A schematic diagram of the results shown in A–D. Gray ovals, peptides containing 4 specific phosphorylation sites. (B) PA-JEB/4WT cells were either serum-starved (lane 1) or treated with 50 ng/ml EGF for the indicated times (in minutes; lanes 2–6) before lysis in mPER buffer. WCLs were probed for the presence of 4 phosphorylated on residues S1356, S1360, and/or S1364 using our rabbit polyclonal phospho-specific antibody 4 pS-CS (top), and the expression level of the 4 proteins was detected using a rabbit polyclonal antibody raised against the first pair of FNIII repeats in 4 (second from top panel). To verify the activation of the EGF receptor and PKC isoforms, WCLs were probed using an antibody that recognizes phospho-tyrosine 845 on the activated EGF receptor (third from top panel) and antibodies raised against phosphorylated Thr 660 of PKC II (pan-pPKC; second from bottom panel). The expression levels of the EGF receptor and PKC isoforms were detected using an antibody against the EGF receptor (1005; third from bottom panel) and a mixture of antibodies against PKC , , and (bottom). Interestingly, the phospho-tyrosine 845 EGFR antibody recognizes a cleaved protein fragment of the full-length EGF receptor that contains the phosphorylated cytoplasmic domain. (C) COS-7 cells were transiently transfected with either IL2R-4WT (lanes 1, 2, and 5) or IL2R-43xA (lanes 3, 4, and 6) cDNA constructs or a control plasmid (lane 7) or an expression construct for the HA-plectin-1A ABDWT (lanes 1–4 and 7) or a control plasmid (lanes 5 and 6). The cells were left untreated (lanes 1, 3, and 5–7) or treated with EGF (lanes 2 and 4) 30 min before lysis in mPER buffer. HA-IPs were probed for the presence of IL2R-4 (top) and the HA-tagged ABDs (second from top panel). WCLs were probed for the expression level of IL2R-4 proteins (third from top panel) and the EGF receptor (second from bottom panel). The activation of the EGF receptor was visualized using phosphotyrosine antibodies (third from bottom panel; the location of the EGFR is denoted by an asterisk) and the classical PKCs using pan antibodies raised against phosphorylated Thr 660 of PKC II (bottom panel; the location of the phospho-PKC is denoted by two asterisks). Quantitation was done in ImageJ and is a ratio of the band intensity shown in the top panel to those in the second to top panel for each lane, relative to lane 1.

Our next aim was to determine whether EGF-induced phosphorylation of serines 1356, 1360, and/or 1364 prevents the association of 4 with the plectin-1A ABD. Again, transient transfection experiments in COS-7 cells were used. IL2R-4^WT or IL2R-4^3xA mutant chimeras were coexpressed with the HA-plectin-1A ABD and were left untreated or were treated with EGF for 30 min. The activation of the endogenous EGF receptor was confirmed by Western blotting with phosphotyrosine antibodies. The results show that after EGF treatment, there is approximately a 55% reduction of the binding of the IL2R-4^WT with the HA-plectin-1A ABD, whereas there is no loss of binding when S1356, S1360, and S1364 are replaced by alanine (Figure 4C). In addition, we used antibodies that recognize the phosphorylated, and thus activated, forms of the classical PKCs to show that at least one isoform is indeed activated after EGF treatment (Figure 4C). These results confirm that EGF-induced phosphorylation of these residues reduces the affinity of the plectin-1A ABD for 4.

To see whether EGF can induce the disassembly of HDs in keratinocytes through the phosphorylation of S1356, S1360, and S1364, we used scatter plot analysis to quantify the results of immunofluorescence assays. Briefly, if the colocalization of two proteins is perfect, all pixels in the scatter plot will align along the diagonal and the further the pixels deviate from this line, the weaker the colocalization of the two proteins. Steady-state 4^WT-expressing PA-JEB cells show typical hemidesmosomal patterning with strong colocalization of plectin and 6 (blue pixels; Figure 5, A–E). However, after EGF treatment, HDs are still present, but significant proportions of plectin (yellow pixels) and 6 (red pixels) are no longer colocalized, which is highlighted in the scatter plot (Figure 5, F–J). Analysis of their subcellular distribution revealed that the noncolocalized 6 occurred in clusters, whereas the noncolocalized plectin (yellow) was diffusely spread over the cytoplasm (Figure 5J). These results indicate that activation of the EGF receptor causes partial disruption of HDs.

View larger version (82K): [in this window] [in a new window]   Figure 5. The effect of S1356/S1360/S1364 phosphorylation on HD formation and plectin binding. PA-JEB/4WT (A–J), PA-JEB/43xA (K–T), or PA-JEB/43xD (U–D') cells were either left untreated (A–E, K–O, and U–Y) or treated for 30 min with EGF (F–J, P–T, or Z–D'), and immunofluorescence studies were performed to locate endogenous plectin using the mAb 121 (red) and 6 using GoH3 (green). Colocalization appears as yellow. Scatter plots are produced as described in Materials and Methods and show the amount of colocalization between 6 and plectin. In the right panels, high intensity colocalizing pixels are shown in blue, noncolocalized high intensity 6 pixels are shown in red, whereas noncolocalized high-intensity plectin pixels are shown in yellow.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
In this study we have assessed the importance of serine phosphorylation of the integrin 4 subunit for the integrity of HDs. We have identified three regions of 4 that are phosphorylated in cells after treatment with calyculin A and show that activation of PKC and PKA in cells induces phosphorylation of S1360 and S1364 on 4, respectively. We also show that activation of the EGF receptor results in the phosphorylation of only S1356, S1360, and/or S1364, and this causes a partial disassembly of HDs in keratinocytes by disrupting the association between 4 with plectin. The results of the in vitro assays indicate that this is due to the inability of the plectin ABD to bind to the first pair of FnIII domains.

Independently from Rabinovitz et al. (2004), we identified S1360 as an important PKC phosphorylation site in 4. In contrast to their interpretation, however, we believe that this is the only PKC phosphorylation site on 4. Our phosphopeptide maps clearly demonstrate that replacement of S1360 by alanine prevents increased phosphorylation of the tryptic peptides derived from this region of 4 in cells upon PMA addition. We believe that background phosphorylation and a different stoichiometry of phosphorylation through the basal activity of other kinases are the cause of the phosphorylation of two tryptic phosphopeptides that contain S1360. This was supported in our calyculin A–derived maps that showed that mutation of S1356, S1360, and S1364 completely prevents phosphorylation of these two tryptic phosphopeptides, consistent with previous results after EGF treatment (Rabinovitz et al., 2004). However, it cannot be ruled out that the in vitro PKC phosphorylation studies performed by Rabinovitz and coworkers resulted in the phosphorylation of residues, other than S1360, that are not normally phosphorylated on 4 in keratinocytes in cells or that there were nonradiolabeled phosphorylated residues in this region of 4 before isolation. Additionally, the presence of two phosphotryptic peptides indicates that phosphorylation of S1360 possibly augments the phosphorylation of either S1356 or S1364 by another kinase that is activated downstream of PMA or that the above result is due to incomplete digestion of a tryptic peptide containing only phosphoserine 1360.

We also demonstrated that S1364 is the sole PKA phosphorylation site on 4 in keratinocytes. Our studies using phosphomimic aspartic acid residues indicate that at least two of the serines at positions 1356, 1360, and 1364 on 4 must be phosphorylated to prevent it from binding to the plectin-1A ABD, which is consistent with the results obtained previously that suggest that more than one phosphorylation event in this region of 4 is necessary to cause disassembly of HDs (Rabinovitz et al., 2004). Therefore, we reasoned that activation of PKA and PKC might be sufficient to prevent the interaction. However, in our COS-7 studies we only saw a limited decrease of the binding between 4 and the ABD. This may be explained in several ways: First, the phosphorylation of the S1360 and S1364 sites is weak in our experiments and, therefore, binding is not prevented to the same extent as in the phosphomimicry assays. Second, activation of PKA and/or PKC may not be sufficient to prevent this association in cells and that still another kinase is necessary for the phosphorylation of S1356. And lastly, in COS-7 cells activation of PKA and PKC at the 4 cytoplasmic domain is not spatially coordinated (e.g., a scaffolding protein is absent). The role of PKC in 4 phosphorylation and HD disassembly has been well reported, but the possibility that PKA is not involved cannot be ruled out. Interestingly, if PKA would not be involved, another kinase(s) must be and it (they) most likely phosphorylate(s) S1356 because no other kinases are known that can phosphorylate S1364.

To simplify our studies, we decided to use EGF receptor activation to definitively determine the importance of these three phosphoserine residues in HD disassembly. We conclusively show that in keratinocytes activation of the EGF receptor induces the phosphorylation of only the peptide containing S1356, S1360, and S1364 in cells, and these events can disrupt the association of plectin with 64, leading to the disassembly of type II HDs as has been previously documented in COS-7 cells, and likely also of type I HDs because their formation is critically dependent on the association of 4 with plectin (Geerts et al., 1999; Koster et al., 2001; Rabinovitz et al., 2004). Because EGF is not the only factor necessary to induce cell migration, other growth factors or cytokines might activate downstream kinases that result in a more robust phosphorylation of 4 at S1356, S1360, and/or S1364 or activate kinases necessary to phosphorylate 4 in other regions to induce HD disassembly. Although our calyculin A phosphopeptide maps clearly show that other serine or threonine residues can be phosphorylated on 4, these studies are complicated by the fact that there are 36 of these residues in the region between amino acids 1382–1487 in 4, and therefore we initially decided to focus on S1356, S1360, and S1364 because they had been previously implicated in HD disassembly. Furthermore, it cannot be ruled out that in addition to the phosphorylation of the above serines, phosphorylation of other components of HDs or other mechanisms altogether might lead to their disassembly. In fact, plectin is phosphorylated at its C-terminus by both p34cdc2 and PKA (Malecz et al., 1996), whereas PKC phosphorylates BP180 and thus dissociates it from HDs (Kitajima et al., 1999).

Interestingly, activation of the EGF receptor has been shown to induce the phosphorylation of 4 on multiple residues, including tyrosine, in carcinoma cells (Mainiero et al., 1996; Mariotti et al., 2001). Because 4 is often not localized at the basal side of carcinoma cells, as it is in keratinocytes, but is instead distributed over the entire plasma membrane, 4 has access to kinases that it does normally not have access to at the basal side of keratinocytes. The resulting aberrant phosphorylation of the 4 subunit most likely explains its assumed function as an adapter protein for tyrosine kinases and their associated signaling proteins in cancer cells (reviewed in Wilhelmsen et al., 2006). In any case, these studies highlight the necessary tight control and spatiotemporal regulation of the phosphorylation of 4 in basal keratinocytes.

We found it interesting that the regions of 4 that are phosphorylated after calyculin A, PMA, FSK/IBMX, or EGF treatment are not directly involved in binding to the plectin ABD, although they regulate this binding. In fact, it has been previously shown that the regions that are phosphorylated can be completely deleted, while the ability of 4 to bind plectin and to form type II HDs is retained (Niessen et al., 1997). There are several possible explanations for this regulation of binding: First, electrostatic repulsion weakens the affinity of 4 for plectin. Second, upon phosphorylation, the 4 cytoplasmic domain may adopt a conformation that is suboptimal for plectin binding. In fact, we have shown previously that the conformation of the 4 cytoplasmic domain may be important for plectin recruitment (Koster et al., 2004). Thirdly, and the most intriguing explanation to us, the phosphorylated CS may compete with plectin for binding to the first pair of FnIII domains. The second FnIII domain contains two arginine residues (R1225 and R1281) that are essential for plectin binding (Koster et al., 2001). In fact, mutation of these residues in human patients causes a nonlethal form of junctional epidermolysis bullosa (Pulkkinen et al., 1998; Nakano et al., 2001). It is spatially possible that two of the three phosphorylated residues in the CS bind to these arginine residues, thereby sterically interfering with plectin binding, and thus regulating the interaction between plectin and 4.

	ACKNOWLEDGMENTS

 
We thank Dr. K. Owaribe for the mAb 121 against HD1/plectin. This work was supported by a grant from the Dutch Cancer Society.

	Footnotes

 
This was published online ahead of print in MBC in Press (http://www.molbiolcell.org/cgi/doi/10.1091/mbc.E07-04-0306) on July 5, 2007.

The online version of this article contains supplemental material at MBC Online (http://www.molbiolcell.org).

Address correspondence to: Arnoud Sonnenberg (a.sonnenberg{at}nki.nl).

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
Bertotti, A., Comoglio, P. M., and Trusolino, L. (2006). 4 integrin activates a Shp2-Src signaling pathway that sustains HGF-induced anchorage-independent growth. J. Cell Biol 175, 993–1003.[Abstract/Free Full Text]

Boyle, W. J., van der Geer, P., and Hunter, T. (1991). Phosphopeptide mapping and phosphoamino acid analysis by two-dimensional separation on thin-layer cellulose plates. Methods Enzymol 201, 110–148.[Medline]

Gache, Y., Chavanas, S., Lacour, J. P., Wiche, G., Owaribe, K., Meneguzzi, G., and Ortonne, J. P. (1996). Defective expression of plectin/HD1 in epidermolysis bullosa simplex with muscular dystrophy. J. Clin. Invest 97, 2289–2298.[Medline]

Geerts, D., Fontao, L., Nievers, M. G., Schaapveld, R. Q., Purkis, P. E., Wheeler, G. N., Lane, E. B., Leigh, I. M., and Sonnenberg, A. (1999). Binding of integrin 64 to plectin prevents plectin association with F-actin but does not interfere with intermediate filament binding. J. Cell Biol 147, 417–434.[Abstract/Free Full Text]

Geuijen, C. A., and Sonnenberg, A. (2002). Dynamics of the 64 integrin in keratinocytes. Mol. Biol. Cell 13, 3845–3858.[Abstract/Free Full Text]

Giudice, G. J., Emery, D. J., and Diaz, L. A. (1992). Cloning and primary structural analysis of the bullous pemphigoid autoantigen BP180. J. Invest. Dermatol 99, 243–250.[CrossRef][Medline]

Hieda, Y., Nishizawa, Y., Uematsu, J., and Owaribe, K. (1992). Identification of a new hemidesmosomal protein, HD1, a major, high molecular mass component of isolated hemidesmosomes. J. Cell Biol 116, 1497–1506.[Abstract/Free Full Text]

Ishihara, H. et al. (1989). Calyculin A and okadaic acid: inhibitors of protein phosphatase activity. Biochem. Biophys. Res. Commun 159, 871–877.[CrossRef][Medline]

Jones, J. C., Kurpakus, M. A., Cooper, H. M., and Quaranta, V. (1991). A function for the integrin 64 in the hemidesmosome. Cell Regul 2, 427–438.[Medline]

Kitajima, Y., Aoyama, Y., and Seishima, M. (1999). Transmembrane signaling for adhesive regulation of desmosomes and hemidesmosomes, and for cell-cell detachment induced by pemphigus IgG in cultured keratinocytes: involvement of protein kinase C. J. Investig. Dermatol. Symp. Proc 4, 137–144.[Medline]

Koster, J., Kuikman, I., Kreft, M., and Sonnenberg, A. (2001). Two different mutations in the cytoplasmic domain of the integrin 4 subunit in nonlethal forms of epidermolysis bullosa prevent interaction of 4 with plectin. J. Invest. Dermatol 117, 1405–1411.[CrossRef][Medline]

Koster, J., van Wilpe, S., Kuikman, I., Litjens, S. H., and Sonnenberg, A. (2004). Role of binding of plectin to the integrin 4 subunit in the assembly of hemidesmosomes. Mol. Biol. Cell 15, 1211–1223.[Abstract/Free Full Text]

Litjens, S. H., de Pereda, J. M., and Sonnenberg, A. (2006). Current insights into the formation and breakdown of hemidesmosomes. Trends Cell Biol 16, 376–383.[CrossRef][Medline]

Litjens, S. H., Koster, J., Kuikman, I., van Wilpe, S., de Pereda, J. M., and Sonnenberg, A. (2003). Specificity of binding of the plectin actin-binding domain to 4 integrin. Mol. Biol. Cell 14, 4039–4050.[Abstract/Free Full Text]

Mainiero, F., Pepe, A., Yeon, M., Ren, Y., and Giancotti, F. G. (1996). The intracellular functions of 64 integrin are regulated by EGF. J. Cell Biol 134, 241–253.[Abstract/Free Full Text]

Malecz, N., Foisner, R., Stadler, C., and Wiche, G. (1996). Identification of plectin as a substrate of p34cdc2 kinase and mapping of a single phosphorylation site. J. Biol. Chem 271, 8203–8208.[Abstract/Free Full Text]

Mariotti, A., Kedeshian, P.A., Dans, M., Curatola, A. M., Gagnoux-Palacios, L., and Giancotti, F. G. (2001). EGF-R signaling through Fyn kinase disrupts the function of integrin 64 at hemidesmosomes: role in epithelial cell migration and carcinoma. invasion. J. Cell Biol 155, 447–458.

Nakano, A., Pulkkinen, L., Murrell, D., Rico, J., Lucky, A. W., Garzon, M., Stevens, C. A., Robertson, S., Pfendner, E., and Uitto, J. (2001). Epidermolysis bullosa with congenital pyloric atresia: novel mutations in the 4 integrin gene (ITGB4) and genotype/phenotype correlations. Pediatr. Res 49, 618–626.[Medline]

Nikolopoulos, S. N., Blaikie, P., Yoshioka, T., Guo, W., Puri, C., Tacchetti, C., and Giancotti, F. G. (2005). Targeted deletion of the integrin 4 signaling domain suppresses laminin-5-dependent nuclear entry of mitogen-activated protein kinases and NF-B, causing defects in epidermal growth and migration. Mol. Cell. Biol 25, 6090–6102.[Abstract/Free Full Text]

Niessen, C. M., Hulsman, E. H., Oomen, L. C., Kuikman, I., and Sonnenberg, A. (1997). A minimal region on the integrin 4 subunit that is critical to its localization in hemidesmosomes regulates the distribution of HD1/plectin in COS-7 cells. J. Cell Sci 110, 1705–1716.[Abstract]

Niessen, C. M., Hulsman, E. H., Rots, E. S., Sanchez-Aparicio, P., and Sonnenberg, A. (1997). Integrin 64 forms a complex with the cytoskeletal protein HD1 and induces its redistribution in transfected COS-7 cells. Mol. Biol. Cell 8, 555–566.[Abstract]

Nievers, M. G., Kuikman, I., Geerts, D., Leigh, I. M., and Sonnenberg, A. (2000). Formation of hemidesmosome-like structures in the absence of ligand binding by the 64 integrin requires binding of HD1/plectin to the cytoplasmic domain of the 4 integrin subunit. J. Cell Sci 113, 963–973.[Abstract]

Nievers, M. G., Schaapveld, R. Q., Oomen, L. C., Fontao, L., Geerts, D., and Sonnenberg, A. (1998). Ligand-independent role of the 4 integrin subunit in the formation of hemidesmosomes. J. Cell Sci 111, 1659–1672.[Abstract]

Oda, K., Matsuoka, Y., Funahashi, A., and Kitano, H. (2005). A comprehensive pathway map of epidermal growth factor receptor signaling. Mol. Syst. Biol 1, 2005–0010.

Orian-Rousseau, V., Aberdam, D., Fontao, L., Chevalier, L., Meneguzzi, G., Kedinger, M., and Simon-Assmann, P. (1996). Developmental expression of laminin-5 and HD1 in the intestine: epithelial to mesenchymal shift for the laminin 2 chain subunit deposition. Dev. Dyn 206, 12–23.[CrossRef][Medline]

Pulkkinen, L., Rouan, F., Bruckner-Tuderman, L., Wallerstein, R., Garzon, M., Brown, T., Smith, L., Carter, W., and Uitto, J. (1998). Novel ITGB4 mutations in lethal and nonlethal variants of epidermolysis bullosa with pyloric atresia: missense versus nonsense. Am. J. Hum. Genet 63, 1376–1387.[CrossRef][Medline]

Rabinovitz, I., Toker, A., and Mercurio, A. M. (1999). Protein kinase C-dependent mobilization of the 64 integrin from hemidesmosomes and its association with actin-rich cell protrusions drive the chemotactic migration of carcinoma cells. J. Cell Biol 146, 1147–1160.[Abstract/Free Full Text]

Rabinovitz, I., Tsomo, L., and Mercurio, A. M. (2004). Protein kinase C-alpha phosphorylation of specific serines in the connecting segment of the 4 integrin regulates the dynamics of type II hemidesmosomes. Mol. Cell. Biol 24, 4351–4360.[Abstract/Free Full Text]

Rezniczek, G. A., de Pereda, J. M., Reipert, S., and Wiche, G. (1998). Linking integrin 64-based cell adhesion to the intermediate filament cytoskeleton: direct interaction between the 4 subunit and plectin at multiple molecular sites. J. Cell Biol 141, 209–225.[Abstract/Free Full Text]

Santoro, M. M., Gaudino, G., and Marchisio, P. C. (2003). The MSP receptor regulates 64 and 31 integrins via 14-3-3 proteins in keratinocyte migration. Dev. Cell 5, 257–271.[CrossRef][Medline]

Schaapveld, R. Q., Borradori, L., Geerts, D., van Leusden, M. R., Kuikman, I., Nievers, M. G., Niessen, C. M., Steenbergen, R. D., Snijders, P. J., and Sonnenberg, A. (1998). Hemidesmosome formation is initiated by the 4 integrin subunit, requires complex formation of 4 and HD1/plectin, and involves a direct interaction between 4 and the bullous pemphigoid antigen 180. J. Cell Biol 142, 271–284.[Abstract/Free Full Text]

Seed, B., and Aruffo, A. (1987). Molecular cloning of the CD2 antigen, the T-cell erythrocyte receptor, by a rapid immunoselection procedure. Proc. Natl. Acad. Sci. USA 84, 3365–3369.[Abstract/Free Full Text]

Sonnenberg, A. et al. (1991). Integrin 64 complex is located in hemidesmosomes, suggesting a major role in epidermal cell-basement membrane adhesion. J. Cell Biol 113, 907–917.[Abstract/Free Full Text]

Sonnenberg, A., Janssen, H., Hogervorst, F., Calafat, J., and Hilgers, J. (1987). A complex of platelet glycoproteins Ic and IIa identified by a rat monoclonal antibody. J. Biol. Chem 262, 10376–11083.[Abstract/Free Full Text]

Stanley, J. R., Hawley-Nelson, P., Yuspa, S. H., Shevach, E. M., and Katz, S, I. (1981). Characterization of bullous pemphi