Biochemical characterization and tissue distribution of the A and B variants of the integrin alpha 6 subunit.

Two cytoplasmic variants of the alpha 6 integrin, alpha 6A and alpha 6B, have been identified previously (Hogervorst, F., I. Kuikman, A. G. van Kessel, and A. Sonnenberg. 1991. Eur. J. Biochem. 199:425-433; Cooper, H. M., R. N. Tamura, and V. Quaranta. 1991. J. Cell Biol. 115:843-850). Using synthetic peptides, containing sequences of their cytoplasmic domains, we have produced mAbs specific for either of the variants. These antibodies reacted with a variety of different epithelial tissues. In some tissues (e.g., salivary gland) both variants could be detected while in others only one of the variants was found (e.g., alpha 6A in epidermis and alpha 6B in kidney). Among nonepithelial cells and tissues, perineural fibroblasts and Schwann cells in peripheral nerves and platelets reacted with anti-alpha 6A, while microvascular endothelia reacted with both anti-alpha 6A and anti-alpha 6B. From our immunohistochemical results there is not evidence that combination with beta 1 or beta 4 is restricted to one of the two variants of alpha 6. This was confirmed by immunoprecipitation studies which showed that both beta 1 and beta 4 were coprecipitated by both anti-alpha 6A or anti-alpha 6B antibodies from cells. Also, the distribution of alpha 6A and alpha 6B subunits associated with beta 1 on cells attached to laminin was similar: both were found in focal contacts colocalizing with vinculin. In contrast, the alpha 6A subunit, associated with beta 4 in cultures of a squamous cell carcinoma cell line, was found to codistribute with bullous pemphigoid antigen 230 in hemidesmosomal-like structures. The alpha 6A and alpha 6B variants, immunoprecipitated from various cell lines, exhibited slightly different electrophoretic mobilities. Analysis of the antigens under reducing conditions showed that the mobility of the light chains, but not of the heavy chains, is different. In addition, in some cells the light chains of alpha 6A and alpha 6B, each are of two different sizes. Treatment with N-glycanase showed that these two light chain variants of alpha 6A and alpha 6B are not due to differences in N-linked glycosylation, and may therefore represent alternative proteolytic products of the alpha 6 precursor. We further demonstrate that alpha 6A, but not alpha 6B, is a major target for PMA-induced phosphorylation. Phosphorylated alpha 6A contained phosphoserine and a small amount of phosphotyrosine. There are also two variants of the integrin alpha 3 subunit with different cytoplasmic domains, but in the cell lines examined only alpha 3A could be demonstrated by RT-PCR.(ABSTRACT TRUNCATED AT 400 WORDS)

Induction of cell scattering by expression of beta1 integrins in beta1-deficient epithelial cells requires activation of members of the rho family of GTPases and downregulation of cadherin and catenin function.

Adhesion receptors, which connect cells to each other and to the surrounding extracellular matrix (ECM), play a crucial role in the control of tissue structure and of morphogenesis. In this work, we have studied how intercellular adhesion molecules and beta1 integrins influence each other using two different beta1-null cell lines, epithelial GE11 and fibroblast-like GD25 cells. Expression of beta1A or the cytoplasmic splice variant beta1D, induced the disruption of intercellular adherens junctions and cell scattering in both GE11 and GD25 cells. In GE11 cells, the morphological change correlated with the redistribution of zonula occluden (ZO)-1 from tight junctions to adherens junctions at high cell confluency. In addition, the expression of beta1 integrins caused a dramatic reorganization of the actin cytoskeleton and of focal contacts. Interaction of beta1 integrins with their respective ligands was required for a complete morphological transition towards the spindle-shaped fibroblast-like phenotype. The expression of an interleukin-2 receptor (IL2R)-beta1A chimera and its incorporation into focal adhesions also induced the disruption of cadherin-based adhesions and the reorganization of ECM-cell contacts, but failed to promote cell migration on fibronectin, in contrast to full-length beta1A. This indicates that the disruption of cell-cell adhesion is not simply the consequence of the stimulated cell migration. Expression of beta1 integrins in GE11 cells resulted in a decrease in cadherin and alpha-catenin protein levels accompanied by their redistribution from the cytoskeleton-associated fraction to the detergent-soluble fraction. Regulation of alpha-catenin protein levels by beta1 integrins is likely to play a role in the morphological transition, since overexpression of alpha-catenin in GE11 cells before beta1 prevented the disruption of intercellular adhesions and cell scattering. In addition, using biochemical activity assays for Rho-like GTPases, we show that the expression of beta1A, beta1D, or IL2R-beta1A in GE11 or GD25 cells triggers activation of both RhoA and Rac1, but not of Cdc42. Moreover, dominant negative Rac1 (N17Rac1) inhibited the disruption of cell-cell adhesions when expressed before beta1. However, all three GTPases might be involved in the morphological transition, since expression of either N19RhoA, N17Rac1, or N17Cdc42 reversed cell scattering and partially restored cadherin-based adhesions in GE11-beta1A cells. Our results indicate that beta1 integrins regulate the polarity and motility of epithelial cells by the induction of intracellular molecular events involving a downregulation of alpha-catenin function and the activation of the Rho-like G proteins Rac1 and RhoA.

Hemidesmosome formation is initiated by the beta4 integrin subunit, requires complex formation of beta4 and HD1/plectin, and involves a direct interaction between beta4 and the bullous pemphigoid antigen 180.

Hemidesmosomes (HDs) are stable anchoring structures that mediate the link between the intermediate filament cytoskeleton and the cell substratum. We investigated the contribution of various segments of the beta4 integrin cytoplasmic domain in the formation of HDs in transient transfection studies using immortalized keratinocytes derived from an epidermolysis bullosa patient deficient in beta4 expression. We found that the expression of wild-type beta4 restored the ability of the beta4-deficient cells to form HDs and that distinct domains in the NH2- and COOH-terminal regions of the beta4 cytoplasmic domain are required for the localization of HD1/plectin and the bullous pemphigoid antigens 180 (BP180) and 230 (BP230) in these HDs. The tyrosine activation motif located in the connecting segment (CS) of the beta4 cytoplasmic domain was dispensable for HD formation, although it may be involved in the efficient localization of BP180. Using the yeast two-hybrid system, we could demonstrate a direct interaction between beta4 and BP180 which involves sequences within the COOH-terminal part of the CS and the third fibronectin type III (FNIII) repeat. Immunoprecipitation studies using COS-7 cells transfected with cDNAs for alpha6 and beta4 and a mutant BP180 which lacks the collagenous extracellular domain confirmed the interaction of beta4 with BP180. Nevertheless, beta4 mutants which contained the BP180-binding region, but lacked sequences required for the localization of HD1/plectin, failed to localize BP180 in HDs. Additional yeast two- hybrid assays indicated that the 85 COOH-terminal residues of beta4 can interact with the first NH2-terminal pair of FNIII repeats and the CS, suggesting that the cytoplasmic domain of beta4 is folded back upon itself. Unfolding of the cytoplasmic domain may be part of a mechanism by which the interaction of beta4 with other hemidesmosomal components, e.g., BP180, is regulated.

Role of binding of plectin to the integrin beta4 subunit in the assembly of hemidesmosomes.

We have previously shown that plectin is recruited into hemidesmosomes through association of its actin-binding domain (ABD) with the first pair of fibronectin type III (FNIII) repeats and a small part of the connecting segment (residues 1328-1355) of the integrin beta4 subunit. Here, we show that two proline residues (P1330 and P1333) in this region of the connecting segment are critical for supporting beta4-mediated recruitment of plectin. Additional binding sites for the plakin domain of plectin on beta4 were identified in biochemical and yeast two-hybrid assays. These sites are located at the end of the connecting segment (residues 1383-1436) and in the region containing the fourth FNIII repeat and the C-tail (residues 1570-1752). However, in cells, these additional binding sites cannot induce the assembly of hemidesmosomes without the interaction of the plectin-ABD with beta4. Because the additional plectin binding sites overlap with sequences that mediate an intramolecular association of the beta4 cytoplasmic domain, we propose that they are not accessible for binding and need to become exposed as the result of the binding of the plectin-ABD to beta4. Furthermore, these additional binding sites might be necessary to position the beta4 cytoplasmic domain for an optimal interaction with other hemidesmosomal components, thereby increasing the efficiency of hemidesmosome assembly.

Distinct and overlapping ligand specificities of the alpha 3A beta 1 and alpha 6A beta 1 integrins: recognition of laminin isoforms.

The ligand specificity of the alpha 3A beta 1 integrin was analyzed using K562 cells transfected with full-length alpha 3A cDNA and was compared with that of alpha 6A beta 1 in similarly transfected K562 cells. Clones were obtained that showed comparable surface expression of either alpha 3A beta 1 or alpha 6A beta 1 integrins. Those expressing alpha 3A beta 1 attached to and spread on immunopurified human kalinin and cellular matrices containing human kalinin, which is a particular isoform of laminin. In addition, alpha 3A transfectants adhered to bovine kidney laminins possessing a novel A chain variant. Binding to kalinin was blocked by a monoclonal antibody against the A chain constituent of kalinin and adhesion to both kalinin and kidney laminins by anti-alpha 3 and beta 1 monoclonal antibodies. The alpha 3A transfected cells bound more strongly to kalinin and bovine kidney laminins after treatment with the beta 1 stimulatory antibody TS2/16. A distinctly weaker and activation-dependent adhesion of alpha 3A transfectants was observed on human placental laminins possessing the Am chain variant (merosin), and no adhesion occurred on bovine heart laminins and murine EHS tumor laminin. Further inactive substrates were fibronectin, nidogen, and collagen types IV and VI, indicating that the alpha 3A beta 1 integrin is a much less promiscuous receptor than thought before. By contrast, alpha 6A transfected cells adhered to all laminin isoforms when stimulated with TS2/16. Adhesion also occurred only on bovine kidney laminins in the absence of TS2/16. These results demonstrate that both alpha 3A beta 1 and alpha 6A beta 1 integrins are typical laminin receptors but that their affinity and activation dependence for binding to various laminin isoforms differ considerably.

Two different mutations in the cytoplasmic domain of the integrin beta 4 subunit in nonlethal forms of epidermolysis bullosa prevent interaction of beta 4 with plectin.

The integrin alpha 6 beta 4 plays a crucial role in the assembly and maintenance of hemidesmosomes. Previous work has shown that the recruitment of plectin into hemidesmosomes is dependent on beta 4 and involves a region of the beta 4 cytoplasmic domain, which contains the first two fibronectin (FNIII) repeats and a short region of the connecting segment. Two missense mutations (R1225H and R1281W) in beta 4 that are responsible for nonlethal forms of epidermolysis bullosa are located in the second FNIII repeat. One of them is confined to a loop region that connects two beta strands (EC') whereas the other is located at the N-terminal end of the second FNIII repeat. We here report that these mutations render beta 4 unable to interact with plectin and prevent the localization of plectin in hemidesmosomes. Substitution of a lysine residue (K1279W) that forms part of the same loop as R1281 had no effect on the ability of beta 4 to recruit plectin. Furthermore, we show that an extended loop structure in beta 4, composed of the amino acids DDN (1262--1264), which resembles the RGD integrin-binding loop in fibronectin, is not involved in the binding to plectin. These results further demonstrate that binding of beta 4 to plectin is essential for the proper formation and function of hemidesmosomes and that loss of the interaction between beta 4 and plectin is associated with a mild form of epidermolysis bullosa.

A novel beta 1 integrin isoform produced by alternative splicing: unique expression in cardiac and skeletal muscle.

The mRNA's of several integrin subunits are alternatively spliced in the region encoding cytoplasmic domains, that may potentially provide alternative integrin-cytoskeleton interactions and transmembrane signaling pathways. We identified a novel cytoplasmic tail variant of the human beta 1 subunit by reverse transcriptase polymerase chain reaction. This fourth beta 1 variant, named beta 1D, is specific for skeletal and cardiac muscle. The determined genomic organization of the 3'-region of the human beta 1 gene reveals that beta 1D is produced by alternative splicing of mRNA. In addition, we show that the expression of beta 1D is developmentally regulated during murine myoblast differentiation, suggesting a role for beta 1D in myogenesis.

The role of phosphorylation in activation of the alpha 6A beta 1 laminin receptor.

The phorbol ester phorbol 12-myristate 13-acetate (PMA) induces phosphorylation of serine residues in the cytoplasmic domain of the alpha 6A integrin subunit, as well as activation of the alpha 6A beta 1 laminin receptor. We examined whether phosphorylation correlates with the induction of high affinity binding of laminin by the alpha 6A beta 1 receptor. Two potential phosphorylation sites for protein kinase C, serine 1041 and serine 1048, are present in the cytoplasmic domain of the alpha 6A subunit. We introduced point mutations into the alpha 6A cDNA, replacing either one or both of the serine residues with alanine. Wild-type and mutant alpha 6A cDNAs were transfected into K562 cells. All alpha 6A subunit mutants were expressed at levels similar to those of wild-type alpha 6A and formed heterodimers with endogenous beta 1. Analysis of the phosphorylation state of wild-type and mutant alpha 6A subunits in resting K562 cells and after treatment with PMA showed that serine 1041, but not serine 1048, is the target residue of PMA-induced phosphorylation. Cells expressing alpha 6A mutant subunits or wild-type alpha 6A transfectants all bound laminin in the presence, but not in the absence of PMA; however, the extent of binding differed. Cells transfected with alpha 6A containing the serine to alanine mutation showed a 2-3-fold higher binding to laminin than cells transfected with alpha 6A containing serine 1041. The results indicate that phosphorylation of the alpha 6A cytoplasmic domain is not required for the induction of high affinity of the alpha 6A beta 1 receptor by PMA, and suggest that, in contrast, it may reduce the affinity of this integrin for ligand.

The unique cytoplasmic domain of the human integrin variant beta4E is produced by partial retention of intronic sequences.

A novel cytoplasmic splice variant of the human beta4 integrin subunit has been identified by reverse transcription polymerase chain reaction using mRNA from cultured keratinocytes as the template. This fifth beta4 variant, called beta4E, is expressed in a wide variety of tissues including the epidermis, lung, duodenum, heart, spleen and stomach and in several human epithelial cell lines. The beta4E cDNA contains an insert of 37 base pairs which produces a frame shift in the sequence encoding the beta4 cytoplasmic domain and generates a new stop codon after a stretch of cDNA encoding a unique 114-amino acid peptide. Analysis of the genomic organization at the site of this insertion in the human beta4 gene reveals that beta4E is produced by partial retention of an intron in the final transcript.

An alternatively spliced exon in the extracellular domain of the human alpha 6 integrin subunit--functional analysis of the alpha 6 integrin variants.

Variants in the extracellular domain of the integrin alpha 7 subunit which arise as a consequence of alternative splicing of mRNA have recently been reported. Two alternative exons, X1 and X2, have been identified in the alpha 7 gene, and homologous exons were found for alpha 6 (Ziober et al., 1993). In this study, we have isolated the region of the alpha 6 gene containing exons X1 and X2 that are, like those of alpha 7, located between stretches of DNA that encode the homologous repeat domains III and IV, proximal to the three divalent cation binding sites of the alpha 6 subunit. We demonstrated by reverse transcriptase polymerase chain reactions and confirmed by sequencing that alpha 6X1 and alpha 6X1X2 mRNAs are generated by alternative splicing of exon X2. The alpha 6X1X2 mRNA is expressed in a limited number of tissues and cell lines and it is always co-expressed with the ubiquitous alpha 6X1 mRNA. Stable transfection of K562 cells with full length cDNAs for the alpha 6AX1X2 and beta 4 subunits resulted in cell populations that expressed the alpha 6AX1X2 variant, in association with either beta 1 or beta 4, on their surface. In addition, a population of cells was isolated that expressed the alpha 6AX1X2 variant at low levels and almost exclusively in association with beta 1. Comparison of the alpha 6AX1X2 integrins with alpha 6AX1 using similarly transfected cells showed no obvious differences between the alternative extracellular alpha 6A isoforms with respect to ligand specificity and activation-dependency of ligand binding. After treatment with the anti-beta 1 stimulatory antibody TS2/16, both the alpha 6AX1 beta 1 and alpha 6AX1X2 beta 1 integrin variants mediated cell adhesion to EHS tumor laminin (laminin-1), kalinin (laminin-5), human placental (laminin-2 and -4) and bovine kidney laminins. In contrast, the alpha 6AX1 beta 4 and alpha 6AX1X2 beta 4 integrins also mediated cell adhesion to laminin and kalinin without stimulation. Furthermore, the different transfectants did not differ in their ability to spread on kalinin. The presented data indicate that the X2 region in alpha 6 is not involved in defining ligand specificity or affinity.

Formation of hemidesmosome-like structures in the absence of ligand binding by the (alpha)6(beta)4 integrin requires binding of HD1/plectin to the cytoplasmic domain of the (beta)4 integrin subunit.

Hemidesmosomes are adhesion structures that mediate anchorage of epithelial cells to the underlying basement membrane. We have previously shown that the (alpha)6(beta)4 integrin can induce the assembly of these multi-protein structures independent of binding to its ligand laminin-5 (ligand-independent formation of hemidesmosomes). Our results suggested a role for HD1/plectin, which binds to the cytoplasmic domain of the (beta)4 integrin subunit, in controlling the clustering of hemidesmosomal components at the basal side of the cell. Using keratinocytes derived from patients lacking HD1/plectin, we now show that ligand-independent formation of hemidesmosomal clusters indeed requires HD1/plectin, in contrast to the ligand-dependent assembly of hemidesmosomes. No clustering of the (alpha)6(beta)4 integrin, or of the bullous pemphigoid antigens BP180 and BP230, was seen when HD1/plectin-deficient keratinocytes were plated on fibronectin or type IV collagen. In (&bgr;)4-deficient keratinocytes, expression of an interleukin 2 receptor (IL2R) transmembrane chimera containing the (beta)4 cytoplasmic tail with the mutation R1281W, which abrogates HD1/plectin binding, resulted in a diffuse distribution of the chimeric receptor. In contrast, a (beta)4(R1281W) mutant that can associate with (alpha)6 and bind ligand, was found to be directed to the basal surface of the cells, at sites where laminin-5 was deposited. In addition, this mutant induced clustering of BP180 and BP230 at these sites. Together, these results show that the formation of hemidesmosomes requires binding of either ligand or HD1/plectin to the (beta)4 integrin subunit. Intriguingly, we found that IL2R/(beta)4 chimeras become localized in pre-existing hemidesmosomes of HD1/plectin-deficient keratinocytes, and that this localization requires a domain in the (beta)4 cytoplasmic tail that is also required for HD1/plectin binding (residues 1115-1356). Because this part of (beta)4 lacks the BP180 binding site, and since we show in this study that it is unable to interact with the same part on another (beta)4 molecule, we suggest that the chimera becomes incorporated into hemidesmosomes of HD1/plectin-deficient keratinocytes by interacting with an as yet unidentified hemidesmosomal component.

Expression and function of the cytoplasmic variants of the integrin alpha 6 subunit in transfected K562 cells. Activation-dependent adhesion and interaction with isoforms of laminin.

Two variants of the cytoplasmic domain of the integrin alpha 6 subunit have been identified (alpha 6A and alpha 6B). To determine the role of each variant in mediating cell adhesion to laminin, we have independently expressed the alpha 6A and alpha 6B subunits in K562 cells. Both variants associated with endogenous beta 1 and were present at comparable levels on the surface of transfected K562 cells. After activation with phorbol ester (phorbol 12-myristate 13-acetate; PMA) or the stimulatory anti-beta 1 antibody TS2/16, alpha 6A beta 1 as well as alpha 6B beta 1 mediated cell adhesion to laminin and more specifically to its fragment E8. Furthermore, both integrin variants interacted with the laminin isoforms kalinin and merosin. Cell adhesion to laminin isoforms was inhibited by the alpha 6-specific monoclonal antibody GoH3. PMA was less efficient in stimulating adhesion than TS2/16 and stimulated adhesion of alpha 6B transfectants better than of alpha 6A transfectants. In contrast, TS2/16 stimulated the adhesion of the alpha 6A and alpha 6B transfectants to laminin to a similar extent. These findings indicate that the cells may regulate the activation of the two alpha 6 variants independently. Activation by PMA was associated with the phosphorylation of both alpha 6A and alpha 6B subunits, but there was no relationship between the degree of phosphorylation and the ability of the transfectants to adhere to laminin since alpha 6A became phosphorylated much more strongly by PMA than alpha 6B. Thus, both alpha 6A beta 1 and alpha 6B beta 1 on K562 cells are activation-dependent receptors for different isoforms of laminin.

The interaction of plectin with actin: evidence for cross-linking of actin filaments by dimerization of the actin-binding domain of plectin.

Plectin is a major component of the cytoskeleton and is expressed in a wide variety of cell types. It plays an important role in the integrity of the cytoskeleton by cross-linking the three filamentous networks and stabilizing cell-matrix and cell-cell contacts. Sequence analysis showed that plectin contains a highly conserved actin-binding domain, consisting of a pair of calponin-like subdomains. Using yeast two-hybrid assays in combination with in vitro binding experiments, we demonstrate that the actin-binding domain of plectin is fully functional and preferentially binds to polymeric actin. The sequences required for actin binding were identified at the C-terminal end of the first calponin homology domain within the actin-binding domain of plectin. We found that the actin-binding domain of plectin is able to bundle actin filaments and we present evidence that this is mediated by the dimerization of this domain. In addition we also show that plectin and another member of the plakin family, dystonin, can heterodimerize by their actin-binding domains. We propose a new mechanism by which plectin and possibly also other actin-binding proteins can regulate the organization of the F-actin network in the cell.

A minimal region on the integrin beta4 subunit that is critical to its localization in hemidesmosomes regulates the distribution of HD1/plectin in COS-7 cells.

The integrin alpha6 beta4 is a major component of hemidesmosomes, in which it mediates firm adhesion to laminin 5. Previous studies have shown that the incorporation of alpha6 beta4 into hemidesmosomes requires a 303 amino acid stretch of the cytoplasmic domain of beta4, comprising part of the first fibronectin type III (FNIII) repeat, the second FNIII repeat and the segment that connects the second to the third FNIII repeat (connecting segment). Now, we have further defined sequences within beta4 that are critical for its localization in hemidesmosomes and we demonstrate that these sequences also induce the redistribution of HD1/plectin into junctional complexes containing the integrin alpha6 beta4 in COS-7 cells, transfected with cDNAs encoding alpha6A and beta4. Truncation of the cytoplasmic domain of beta4 after amino acids 1,382 or 1,355 in the connecting segment, by which a potential tyrosine activation motif (TAM) is removed, does not prevent the localization of alpha6 beta4 in hemidesmosomes in the rat bladder carcinoma cell line 804G and neither did it eliminate the ability of alpha6 beta4 to change the subcellular distribution of HD1/plectin in COS-7 cells. In contrast, beta4 subunits in which the entire connecting segment had been deleted or which were truncated after amino acid 1,328, which removes almost the complete segment, had lost both of these functions. Furthermore, when beta4 subunits with either a deletion of the second FNIII repeat or a small deletion in this repeat were co-expressed with alpha6, the integrins were not localized in hemidesmosomes and did not induce the redistribution of HD1/plectin in COS-7 cells. Finally, the fourth FNIII repeat of beta4 could not replace the second in either of these activities. These findings establish that a region in beta4, which encompasses the second FNIII repeat and a stretch of 27 amino acids (1,329-1,355) of the connecting segment, is critical for the localization of alpha6beta4 in hemidesmosomes and that it regulates the distribution of HD1/plectin.

Molecular cloning of the human alpha 6 integrin subunit. Alternative splicing of alpha 6 mRNA and chromosomal localization of the alpha 6 and beta 4 genes.

We have isolated cDNAs encoding the alpha 6 subunit from a lambda gt11 expression library from human keratinocytes by combined screening with a rabbit polyclonal anti-alpha 6 antibody and the polymerase chain reaction. The alpha 6 subunit encoded by this cDNA consists of 1050 amino acids with a 991-amino-acid extracellular, a 23-amino-acid transmembrane and a 36-amino-acid cytoplasmic domain. The extracellular domain contains three putative divalent cation-binding sites and nine potential N-linked glycosylation sites. From a cDNA library from normal human mammary gland cells two different cDNAs for alpha 6 were isolated, one of which is identical to the above cDNA. The two alpha 6 subunits, called alpha 6A and alpha 6B, encoded by the two cDNAs each have a unique cytoplasmic domain, that of alpha 6B being 18 amino acids longer than that of alpha 6A. Different carcinoma cell lines contain transcripts for both alpha 6 subunits. K562 leukemic cells have little alpha 6A or alpha 6B mRNAs. The overall level of expression varies in the carcinoma cell lines, but reflects alpha 6 cell surface expression. In A375 melanoma cells, however, cell surface expression of alpha 6 was low in spite of a high level of mRNA. This suggest that other mechanisms may be involved in regulating the expression of alpha 6 on the surface of these cells. The mRNA for both alpha 6 subunits is around 6 kb. The alpha 6 subunits are similar to other alpha subunits (26-31% identity with cleaved alpha subunits) of the integrin family but they are more similar to the alpha 3 subunit (40% identity). This high degree of similarity may be the basis for their functional resemblance since both alpha 3 and alpha 6 subunits, when associated with beta 1, function as laminin receptors and bind to the long arm of laminin. The genes for alpha 6 and beta 4, the alternative beta subunit with which alpha 6 combines on certain epithelial cells, were mapped to chromosome 2 and 17q11-qter, respectively.

Cloning and sequence analysis of beta-4 cDNA: an integrin subunit that contains a unique 118 kd cytoplasmic domain.

The alpha 6 beta 4 complex is a member of the integrin superfamily of adhesion receptors. A human keratinocyte lambda gt11 cDNA library was screened using a monoclonal antibody directed against the beta 4 subunit. Two cDNAs were selected and subsequently used to isolate a complete set of overlapping cDNA clones. The beta 4 subunit consists of 1778 amino acids with a 683 amino acid extracellular domain, a 23 amino acid transmembrane domain and an exceptionally long cytoplasmic domain of 1072 residues. The deduced amino-terminal sequence is in good agreement with the published amino-terminal sequence of purified beta 4. The extracellular domain contains five potential N-linked glycosylation sites and four cysteine-rich homologous repeat sequences. The extracellular part of the beta 4 subunit sequence shows 35% identify with other integrin beta subunits, but is the most different among this class of molecules. The transmembrane region is poorly conserved, whereas the cytoplasmic domain shows no substantial identity in any region to the cytoplasmic tails of the known beta sequences or to other protein sequences. The exceptionally long cytoplasmic domain suggests distinct interactions of the beta 4 subunit with cytoplasmic proteins.

The two phenylalanines in the GFFKR motif of the integrin alpha6A subunit are essential for heterodimerization.

The membrane-proximal domain of the integrin alpha subunit contains a conserved motif of five amino acid residues, GFFKR. We deleted this motif from the human alpha6A subunit and found that in COS-7 cells this mutant cannot associate with the beta1 subunit and is retained in the endoplasmic reticulum. Point mutations in the GFFKR motif of the glycine residue or the two highly charged amino acids, or deletion of the lysine and arginine residues, had no effect on the ability of alpha6 to interact with beta1 and to be expressed at the cell surface. In contrast, by replacing either of the two phenylalanines with alanine, or by deletion of both of these residues, alpha6 was incapable of associating with beta1. The alpha6 point mutants that associated with beta1 were expressed in K562 cells and their responsiveness to integrin-activating factors was determined. None of these transfectants bound spontaneously to laminin-1, but binding could be induced by either PMA or the stimulating anti-beta1 antibody TS2/16 to the same extent as that of the wild-type transfectant. The ability of these mutants to initiate focal-contact formation in CHO cells plated on laminin-1 substrates also appeared to be unaltered. Thus the behaviour of alpha6 mutants involving the glycine, lysine or arginine residues was indistinguishable from that of wild-type alpha6 both in inside-out and outside-in signalling. In contrast, deletion of the cytoplasmic domain of alpha6 C-terminal of the GFFKR motif resulted in a loss of responsiveness of alpha6beta1 to PMA stimulation and formation of focal contacts on laminin-1. However, this mutant was targeted to focal contacts formed by other integrins, even when they had not bound ligand. Together, these results suggest that the two phenylalanine residues of the GFFKR motif provide a site for interaction of the alpha6A subunit with beta1, whereas the cytoplasmic domain C-terminal of this motif is involved in the regulation of bidirectional signalling via alpha6Abeta1.

Identification of the cleavage sites in the alpha6A integrin subunit: structural requirements for cleavage and functional analysis of the uncleaved alpha6Abeta1 integrin.

The alpha6A and alpha6B integrin subunits are proteolytically cleaved during biosynthesis into a heavy chain (120 kDa) that is disulphide-linked to one of two light chains (31 or 30 kDa). Analysis of the structure of the alpha6A subunit on the carcinoma cell line T24 and human platelets demonstrated that the two light chains of alpha6 are not differentially glycosylated products of one polypeptide. Rather they possess different polypeptide backbones, which presumably result from proteolytic cleavage at distinct sites in the alpha6 precursor. Mutations were introduced in the codons for the R876KKR879, E883K884, R890K891 and R898K899 sequences, the potential proteolytic cleavage sites, and wild-type and mutant alpha6A cDNAs were transfected into K562 cells. The mutant alpha6A integrin subunits were expressed in association with endogenous beta1 at levels comparable to that of wild-type alpha6Abeta1. A single alpha6 polypeptide chain (150 kDa) was precipitated from transfectants expressing alpha6A with mutations or deletions in the RKKR sequence. Mutations in the EK sequence yielded alpha6A subunits that were cleaved once into a heavy and a light chain, whereas alpha6A subunits with mutations in one of the two RK sequences were, like wild-type alpha6A, cleaved into one heavy and two light chains. Thus a change in the RKKR sequence prevents the cleavage of alpha6. The EK site is the secondary cleavage site, which is used only when the primary site (RKKR) is intact. Microsequencing of the N-termini of the two alpha6A light chains from platelets demonstrated that cleavage occurs after Arg879 and Lys884. Because alpha6(RKKG), alpha6(GKKR) and alpha6(RGGR) subunits were not cleaved it seems that both the arginine residues and the lysine residues are essential for cleavage of RKKR. alpha6A mutants with the RKKR sequence shifted to the EK site, in such a way that the position of the arginine residue after which cleavage occurs corresponds exactly to Lys884, were partly cleaved, whereas alpha6A mutants with the RKKR sequence shifted to other positions in the alpha6A subunit, including one in which it was shifted two residues farther than the EK cleavage site, were not cleaved. In addition, alpha6A mutants with an alpha5-like cleavage site, i.e. arginine, lysine and histidine residues at positions -1, -2 and -6, were not cleaved. Thus both an intact RKKR sequence and its proper position are essential. After activation by the anti-beta1 stimulatory monoclonal antibody TS2/16, both cleaved and uncleaved alpha6Abeta1 integrins bound to laminin-1. The phorbol ester PMA, which activates cleaved wild-type and mutant alpha6Abeta1, did not activate uncleaved alpha6Abeta1. Thus uncleaved alpha6Abeta1 is capable of ligand binding, but not of inside-out signalling. Our results suggest that cleavage of alpha6 is required to generate a proper conformation that enables the affinity modulation of the alpha6Abeta1 receptor by PMA.

The subcellular distribution of the high molecular mass protein, HD1, is determined by the cytoplasmic domain of the integrin beta 4 subunit.

The high molecular mass protein, HD1, is a structural protein present in hemidesmosomes as well as in distinct adhesion structures termed type II hemidesmosomes. We have studied the distribution and expression of HD1 in the GD25 cells, derived from murine embryonal stem cells deficient for the beta 1 integrin subunit. We report here that these cells possess HD1 but not BP230 or BP180; two other hemidesmosomal constituents, and express only traces of the alpha 6 beta 4 integrin. By immunofluorescence and interference reflection microscopy HD1 was found together with vinculin at the end of actin filaments in focal contacts. In OVCAR-4 cells, derived from a human ovarian carcinoma which, like GD25 cells, only weakly express alpha 6 beta 4, HD1 was also localized in focal contacts. Upon transfection of both GD25 and OVCAR-4 cells with cDNA for the human beta 4 subunit the subcellular distribution of HD1 changed significantly. HD1 is then no longer present in focal contacts but in other structures at cell-substrate contacts, colocalized with alpha 6 beta 4. These junctional complexes are probably the equivalent of the type II hemidesmosomes. Transfection of GD25 cells with beta 1 cDNA did not affect the distribution of HD1, which indicates that the localization of HD1 in focal contacts was not due to the absence of beta 1. Moreover, in GD25 cells transfected with cDNA encoding a beta 4/beta 1 chimera, in which the cytoplasmic domain of beta 4 was replaced by that of beta 1, the distribution of HD1 was unaffected. Our findings indicate that the cytoplasmic domain of beta 4 determines the subcellular distribution of HD1 and emphasize the important role of alpha 6 beta 4 in the assembly of hemidesmosomes and other junctional adhesive complexes containing HD1.

The alpha 6 beta 4 integrin is a receptor for both laminin and kalinin.

Previously, we have establish K562 transfectants that express either alpha 6A beta 1 or alpha 6B beta 1 (K alpha 6A or K alpha 6B) on their surface. Both cell lines bind to laminin and kalinin after treatment with the beta 1-stimulatory antibody TS2/16. Here we introduce the full-length beta 4 cDNA into the alpha 6A- and alpha 6B-expressing K562 cells and selected stably transfected cells. The beta 4 subunit was expressed on the surface of both transfectants and it formed dimers with the alpha 6A or alpha 6B subunits. Immunoprecipitation and preclearing analyses revealed that both transfectants expressed alpha 6 beta 1, in addition to alpha 6 beta 4. While K alpha 6A and K alpha 6B cells required TS2/16 stimulation for binding to laminin or kalinin, adhesion of the unstimulated beta 4-transfected K alpha 6A and K alpha 6B cells to these matrix components was already substantial. This adhesion was mediated by both alpha 6 beta 1 and alpha 6 beta 4 since it was completely blocked by an alpha 6-specific antibody or by a combination of anti-beta 1 and anti-beta 4 antibodies, but only partially by either of these latter two antibodies alone. Adhesion to laminin was completely blocked by an antiserum to laminin fragment E8 as was the adhesion to kalinin by an antibody to kalinin, demonstrating the specificity of adhesion. Both transfectants always adhered more strongly to kalinin than to laminin. Furthermore, binding to kalinin was less well blocked by antibodies to beta 4 than binding to laminin, indicating that the affinity of alpha 6 beta 4 for kalinin is higher than that for laminin. The fact that alpha 6 beta 1 mediated adhesion without TS2/16 stimulation on the beta 4-transfected K alpha 6A and K alpha 6B cells suggests that some activation of alpha 6 beta 1 had occurred in these cells, even though binding was increased when they were actively stimulated by the antibody TS2/16. Finally, we show that Mn2+ induced binding of solubilized alpha 6 beta 4 to matrix containing kalinin, deposited by the murine cell line RAC-11P/SD. This binding was inhibited by the anti-alpha 6 mAb GoH3. Together, these results indicate that both alpha 6 beta 1 and alpha 6 beta 4 are receptors for laminin and kalinin and that there are no differences in ligand specificity between the A and B variants of the alpha 6 subunit when associated with either beta 1 or beta 4.