Identification of specific human epithelial cell integrin receptors as VLA proteins

Defective beta1-integrins expression in arsenical keratosis and arsenic-treated cultured human keratinocytes

BACKGROUND

beta1-integrins, which localize to the basolateral surface of basal keratinocytes, are important in the differentiation control and proliferation of the epidermis. Many cutaneous diseases with perturbed differentiation, including arsenical keratosis, show altered patterns of integrin distribution and expression. Arsenic may induce arsenical keratosis through the differentiation and apoptosis aberration by integrins. The purpose of this study is to investigate the role of integrin and arsenic in the pathogenesis of arsenical keratosis.

METHODS

Twenty-five specimens obtained from 25 patients with arsenical keratosis disease were studied. Immunohistochemistry staining to beta1, alpha2beta1, or alpha3beta1 integrins was performed in arsenical keratosis and clinically normal perilesional skin. Western blotting was used to assess the expression of integrin beta1 and focal adhesion kinase (FAK) in arsenic-treated cultured keratinocytes.

RESULTS

A decreased expression of beta1, alpha2beta1, or alpha3beta1 integrins was demonstrated in arsenical keratosis and clinical normal perilesional skin in a large proportion of arsenical keratosis cases studied. The expressions of integrin beta1 and FAK were both decreased in arsenic-treated keratinocytes.

CONCLUSIONS

Our results suggest that arsenic induces abnormal differentiation in arsenical keratosis via the effects of integrin expression in keratinocytes.

Conjugation of extracellular matrix proteins to basal lamina analogs enhances keratinocyte attachment

The dermal-epidermal junction of skin contains extracellular matrix proteins that are involved in initiating and controlling keratinocyte signaling events such as attachment, proliferation, and terminal differentiation. To characterize the relationship between extracellular matrix proteins and keratinocyte attachment, a biomimetic design approach was used to precisely tailor the surface of basal lamina analogs with biochemistries that emulate the native biochemical composition found at the dermal-epidermal junction. A high-throughput screening device was developed by our laboratory that allows for the simultaneous investigation of the conjugation of individual extracellular matrix proteins (e.g. collagen type I, collagen type IV, laminin, or fibronectin) as well as their effect on keratinocyte attachment, on the surface of an implantable collagen membrane. Fluorescence microscopy coupled with quantitative digital image analyses indicated that the extracellular matrix proteins adsorbed to the collagen-GAG membranes in a dose-dependent manner. To determine the relationship between extracellular matrix protein signaling cues and keratinocyte attachment, cells were seeded on protein-conjugated collagen-GAG membranes and a tetrazolium-based colorimetric assay was used to quantify viable keratinocyte attachment. Our results indicate that keratinocyte attachment was significantly enhanced on the surfaces of collagen membranes that were conjugated with fibronectin and type IV collagen. These findings define a set of design parameters that will enhance keratinocyte binding efficiency on the surface of collagen membranes and ultimately improve the rate of epithelialization for dermal equivalents.

Expression of alpha(v)beta6 integrin in oral leukoplakia

The distribution of alpha(v)beta6 integrin was examined in oral leukoplakia, lichen planus and squamous cell carcinomas using immunohistochemistry. Controls included oral mucosal wounds, chronically inflamed and normal oral mucosa. Integrins beta1, beta3, beta4, beta5, fibronectin and tenascin were also studied. The integrin alpha(v)beta6 was highly expressed throughout the whole lesion of 90% of the squamous cell carcinomas but was not present in any of the normal specimens. alpha(v)beta6 integrin was also expressed in 41% of the leukoplakia specimens, and 85% of the lichen planus samples, but in none of the tissues with inflammatory hyperplasia or chronic inflammation. The expression of beta1 integrins was localized in the basal layer, and that of the beta4 at the cell surface facing the basement membrane of all specimens. The integrins beta3 and beta5 were absent from all normal and leukoplakia specimens. Fibronectin and tenascin were present in the connective tissue underneath the epithelium of all the sections, and their expression was similar in both alpha(v)beta6-positive and alpha(v)beta6-negative tissues. A group of 28 leukoplakia patients were followed 1-4 years after first diagnosis. In this group, initially alpha(v)beta6 integrin-positive leukoplakia specimens had high tendency for disease progression while alpha(v)beta6-negative specimens did not progress. These results suggest that the expression of alpha(v)beta6 integrin could be associated in the malignant transformation of oral leukoplakias.

Cross-species interaction of porcine and human integrins with their respective ligands: implications for xenogeneic tolerance induction

BACKGROUND

Organ transplantation is limited by the number of available donors. One possible solution would be the use of pigs as organ donors. However, current immunosuppressive protocols cannot prevent rejection of these organs. If donor-specific tolerance toward porcine antigens could be induced in recipients, subsequent implantation of porcine organs would be possible without further immunosuppression. Induction of tolerance can be achieved with a bone marrow transplant if donor antigen-presenting cells successfully differentiate in the recipient thymus to induce deletion of donor-reactive host cells. Migration of porcine progenitor cells to the host marrow and thymus and differentiation into tolerance-inducing antigen-presenting cells is likely to require successful interaction of porcine adhesion molecules with human ligands. In this study, we investigated whether very late antigen (VLA)4 and VLA-6 integrins, which play important roles in homing and differentiation of hematopoietic progenitor cells, function across the pig-to-human species barrier.

METHODS

Static cell-to-cell and cell-to-extracellular matrix protein adhesion assays were used to examine the cross-species interaction of porcine adhesion molecules with human ligands.

RESULTS

Our studies show that porcine cells adhere to various human endothelial cell monolayers and extracellular matrix proteins and demonstrate that porcine VLA-4 and VLA-6 appear to be fully cross-reactive to the human ligands vascular cell adhesion molecule-1 and laminin, respectively.

CONCLUSIONS

It is likely that porcine hematopoietic progenitor cells will be able to successfully employ pVLA-4- and pVLA-6-human ligand interactions in a pig-to-human bone marrow transplantation model in order to induce donor-specific tolerance.

Topography and biological role of integrins in human skin

We report the topography of integrins in the human epidermis and in cultured human keratinocytes. Both in situ and in vitro beta 1 integrins are exposed at the cell-cell adhesion interface while beta 4 is located on the basal membrane in contact with the basal lamina. Such defined sorting identifies discrete cell membrane domains that may be involved in defining, building up, and maintaining epithelial cell polarity. The distribution of integrins is deeply altered in hyperproliferative states like those occurring in several experimental conditions and in epidermal diseases.

Expression of the antigen recognized by mAb GB36 in normal skin and in skin tumours

GB36, a mouse monoclonal antibody (mAb) raised against an epithelial antigen of the human trophoblast, reacts with the epithelial basement membrane of chorionic villi; it does not react with the invasive extravillous cytotrophoblast. Expression and characterization of the antigen of GB36 (designated GBA36) were investigated in normal keratinocytes by immunoprecipitation, immunofluorescence and immunoelectron microscopic studies. Immunoprecipitation experiments demonstrated that the proteins identified on keratinocytes by mAb GB36 and a rat mAb anti-integrin alpha 6 (GoH3) were the same. Using immunofluorescence and immunoelectron microscopic methods, GBA36 was localized on the cell membrane facing the epithelial basal lamina of basal keratinocytes. GBA36 distribution in benign and malignant skin tumours was evaluated by immunostaining methods (immunofluorescence and immunoperoxidase). Analysis of tumours revealed that whereas benign epithelial tumours and intradermal naevi displayed high levels of GBA36, the expression of this antigen decreased progressively in spinocellular and basal cell carcinomas, and in cutaneous melanomas in relation to invasiveness. During cell transformation, GBA36 undergoes quantitative alterations, and expression is down-regulated. Although the functional relevance of these changes remains unknown, the correlation of decreased GBA36 expression with tumour progression may indicate a role for altered integrin expression in tissue invasion by human skin carcinoma and melanoma.

alpha 3A beta 1 integrin localizes to focal contacts in response to diverse extracellular matrix proteins

In vitro binding assays and inhibition of cell adhesion with monoclonal antibodies have implicated the integrin alpha 3 beta 1 as a receptor for a variety of extracellular ligands. However, reports of alpha 3 beta 1-ligand interactions are inconsistent, and transfection studies have suggested that alpha 3 beta 1 is not sufficient for cell attachment to ligands other than kalinin/laminin 5. We used immunofluorescence to study subcellular localization of the alpha 3A cytoplasmic domain variant in different cultured cell types. Using standard fixation and permeabilization methods, antibodies specific for alpha 3A stained most cell types in a diffuse pattern, consistent with previous reports. Surprisingly, however, chemical cross-linking of integrins to the extracellular matrix and extraction of the cytoskeleton prior to immunofluorescence revealed alpha 3A in focal contacts of most cells tested, suggesting that the cytoplasmic domain was concealed in intact focal contacts by cytoskeletal or other cytoplasmic proteins. The alpha 3A subunit localized to focal contacts in several cell types cultured on fibronectin, kalinin/laminin 5, EHS-laminin/laminin 1, type IV collagen, or vitronectin. In contrast, alpha 5 and alpha V integrins were detected in focal contacts only in cells grown on their known ligands (fibronectin, and fibronectin or vitronectin, respectively). Therefore, our results show that alpha 3A beta 1 responds to a broad spectrum of extracellular ligands. Time course comparisons of the recruitment of alpha subunits from different fibronectin receptors suggested that localization of alpha 3A beta 1 to fibronectin-induced focal contacts was independent of the recruitment of alpha 5 and alpha 4 integrins. However, other studies have shown that alpha 3A beta 1 does not mediate initial cell adhesion to many of the ligands that induced its focal contact localization, including fibronectin. Therefore, we suggest that alpha 3A beta 1 may be a secondary receptor with post-cell-adhesion functions for a broad spectrum of extracellular matrices.

Expression of beta 1 integrin receptors in transformed mouse epidermal keratinocytes: upregulation of alpha 5 beta 1 in spindle carcinoma cells

The adhesive properties and the expression of extracellular matrix receptors of the beta 1-integrin subfamily were analyzed in transformed epidermal keratinocyte cell lines of different stages of mouse skin carcinogenesis. One- and two-dimensional analyses of the immunoprecipitates obtained with anti-beta 1- and specific anti-alpha-integrin subunits showed qualitative and quantitative changes in the expression of beta 1 integrins by the different cell lines. The polyvalent alpha 3 beta 1 integrin was expressed by all analyzed cell lines, although the levels detected in undifferentiated spindle CarC cells were lower than those present in the rest of keratinocyte cell lines. In contrast, spindle cells expressed high levels of the specific fibronectin receptor alpha 5 beta 1, whereas this integrin was absent or expressed at very reduced levels in the other epithelial cell lines. Expression of alpha 5 beta 1 integrin in spindle cells appeared organized in cell-substratum contact areas on spread cells. In addition, high and homogenous expression of alpha 5 beta 1 was detected in fully undifferentiated tumors induced in nude mice by three independent spindle cell lines. These results suggest that the expression of alpha 5 beta 1 integrin is upregulated during the development of spindle cell carcinomas that occur in the last stages of mouse skin carcinogenesis and can be associated with the acquisition of the fibroblastoid phenotype of spindle cells. On the other hand, expression of the collagen receptor alpha 2 beta 1 was demonstrated in a transformed cell line (PDV), and it was apparently also expressed in two other malignant keratinocyte cell lines (PDVC57 and HaCa4). The expression of alpha 2 beta 1 was correlated with the increased adhesion to collagen type I and collagen type IV exhibited by the tumorigenic cell lines.

Integrin expression by human epidermal keratinocytes can be modulated by interferon-gamma, transforming growth factor-beta, tumor necrosis factor-alpha, and culture on a dermal equivalent

Receptors of the integrin family are largely confined to the basal layer of keratinocytes, both in human epidermis and in stratified cultures of human keratinocytes. However, suprabasal integrin expression is observed during epidermal wound healing and in psoriatic lesions. We have investigated potential stimuli of suprabasal expression. Addition of transforming growth factor-beta (TGF-beta), interferon-gamma (IFN-gamma), or tumor necrosis factor-alpha (TNF-alpha) to keratinocytes cultured with a 3T3 feeder layer did not induce suprabasal expression. The cytokines caused small changes in the levels of alpha 2 beta 1 or alpha 3 beta 1 on the surface of basal keratinocytes but had no significant effect on the proportion of cells adhering to fibronectin, type IV collagen, and laminin, and did not cause changes in the mobility of integrin subunits on polyacrylamide gels. Injection of TNF-alpha or IFN-gamma intradermally into healthy human volunteers induced an inflammatory response but did not induce suprabasal integrin expression. However, we did observe transient suprabasal integrin expression when keratinocytes were grown on a dermal equivalent consisting of fibroblasts in a collagen gel. One week after raising the cultures to the air-liquid interface, beta 1 integrins were found in all the viable cell layers, with suprabasal cells co-expressing integrins and involucrin; 1 week later integrins were confined to the basal layer. Addition of TGF-beta, IFN-gamma, or TNF-alpha to the dermal equivalents neither induced nor inhibited suprabasal integrin expression. We conclude that suprabasal integrin expression is not induced by the inflammatory cytokines tested, and instead may reflect the proliferation/differentiation status of the epidermis.

Expression of ICAM-3 on human epidermal dendritic cells

Three counter-receptors for LFA-1 of the immunoglobulin family have been discovered: ICAM-1, ICAM-2, and ICAM-3. Despite their homologies, their patterns of expression suggest specialized roles. The finding that ICAM-3 is much better expressed than other LFA-1 ligands on monocytes and resting T cells, and that this discovery may be important in the initiation of immune responses prompted us to search for the expression of ICAM-3 by human epidermal Langerhans cells (LC). Six out of eight different ICAM-3 monoclonal antibodies were found to be reactive with epidermal LC. Immunoelectron-microscopy staining revealed that 100% of freshly-isolated, typical Birbeck granules containing LC expressed ICAM-3. After one day and three days of culture, 100% of LC still expressed ICAM-3, but the staining intensity was decreased by 58% and 76% respectively. Immunoprecipitation of 125I surface-labeled LC with anti-ICAM-3 antibodies revealed a polypeptide with apparent M(r) of 122,000-125,000. To determine whether ICAM-3 was involved in LC function, mixed epidermal cell-lymphocyte reactions were performed with freshly isolated LC in the presence of various concentrations of different anti-ICAM-3 antibodies. Among the different antibodies tested, HP2/19 and CBR-IC3/1 were found to partially block the reaction in a dose-dependent manner, suggesting that ICAM-3 represents a new molecule involved in the initiation of the immune response driven by epidermal LC.

Characterization of the antigens recognized by two monoclonal antibodies reactive with basal-layer keratinocytes of human epidermis

Two monoclonal antibodies, GR3 and GR4, reactive with the basal-layer keratinocytes of human epidermis, were derived by immunization of Balb/c mice with glycoproteins isolated from cultured keratinocytes by lectin-affinity chromatography. Immunoprecipitation of Triton X-100 extracts from human keratinocytes metabolically labelled with D-[1-14C]glucosamine revealed that GR3 recognized a major glycoprotein with migration properties identical with those of a glycoprotein (reduced form M(r) 126,000) which was previously shown to be implicated in intercellular adhesion [Roberts and Brunt (1985) Biochem J. 232, 67-70]. In their unreduced forms the antigens recognized by GR3 and GR4 both migrated as two bands with M(r) values of 118,000 and 147,000. Comparison of 125I-labelled glycoproteins immunoprecipitated by GR3, GR4 and integrin antibodies revealed that, under reducing conditions, the major band immunoprecipitated by both GR3 and GR4 co-migrated with the alpha 3 and beta 1 integrin chains. In addition the immunoprecipitate obtained with GR4 contained an additional band co-migrating with the alpha 2 integrin chain. Sequential immunoprecipitation studies with GR3, GR4 and integrin antibodies confirmed that GR3 is directed against the alpha 3 integrin chain, whereas GR4 is directed against the beta 1 chain. These studies also indicate that some of the alpha 2 integrin chains on keratinocytes may be associated with a beta-chain not recognized by the antisera against the beta 1 integrin chain used in this study.

Integrin expression in middle ear cholesteatoma

Cholesteatoma is lined by a squamous keratinizing epithelium exhibiting most of the features of normal epidermis. In this study, we investigated by immunohistochemistry the expression of integrin adhesion molecules in primary acquired and recurrent cholesteatomas, and compared it with common epidermal cysts and normal human skin. The results showed that cholesteatoma epithelium exhibited a markedly augmented expression of alpha v integrin subunit and a corresponding increased deposition of vitronectin (alpha v ligand) in the surrounding stroma as compared to epidermal cyst and normal human skin. In contrast, the expression pattern of alpha 2 beta 1, alpha 3 beta 1, and alpha 6 beta 4 integrins as well as the distribution of laminin, collagen IV and fibronectin were similar in cholesteatomas, epidermal cysts and normal human skin. Similar staining pattern was observed in primary acquired and recurrent cholesteatoma.

Adhesive properties and integrin expression profiles of two colonic cancer populations differing by their spreading on laminin

The mostly undifferentiated parental HT29 (HT29p) human colonic adenocarcinoma cell line and a differentiated subpopulation selected by the anti-cancer drug 5-fluorouracil (HT29-Fu) (Lesuffleur et al. (1991) Int. J. Cancer 49, 721–730) display strikingly different behavior when grown on laminin coatings: the former grows as aggregates while the latter grows as monolayers. In an attempt to explain this difference, we performed a comparative study of cell adhesion properties and of expression, involvement and localization of the alpha 6, beta 1 and beta 4 subunits constituting the integrin family among the two cell populations. HT29p and HT29-Fu cells exhibited a similar adhesion pattern to laminin and laminin fragments E8 and P1. In both cell lines, cell adhesion could be blocked at about 90% with anti-alpha 6 subunit antibodies and around 30–50% with anti-beta 1 antibodies; no inhibition of the cell adhesion was obvious when using anti-beta 4 antibodies. Immunoprecipitations of iodinated membrane-solubilized proteins and immunoblotting experiments showed that all alpha 6 chains expressed in both HT29p and HT29-Fu cell populations exist as alpha 6 beta 4 integrins; beta 1 subunits are associated with alpha 2 and alpha 3 chains. When HT29p or HT29-Fu cells were injected subcutaneously in nude mice, a similar expression pattern of alpha 6, beta 4 and beta 1 integrin subunits was noticeable in the resulting tumors: alpha 6 and beta 4 subunits were localized at the basal surface of the tumor cells facing the stromal elements, and to a lesser extent at the cell-cell contacts within the tumor-cell clumps; beta 1 subunits were mainly found within the cytoplasm of the tumor cells. Despite these overall similarities among the two cell lines, the following changes could account for their different behavior on laminin: less proteolytic processing of the beta 4 integrin subunit occurred in HT29-Fu cells yielding peptidic fragments of 175 kDa, which are absent from the parental cells; the immunostaining pattern of the various subunits demonstrated a segregation of alpha 6, beta 4 and beta 1 integrin subunits on the basal side of the HT29-Fu cells when cultured on laminin to the detriment of their lateral location, a phenomenon that was not obvious in the parental cells. Altogether, these results suggest that the distinct behavior of the undifferentiated versus differentiated HT29 cell populations on laminin is not related to altered adhesion properties of the cells but rather to a deficient stabilization of the adhesion leading to cell spreading.(ABSTRACT TRUNCATED AT 400 WORDS)

Functional down-regulation of alpha 5 beta 1 integrin in keratinocytes is reversible but commitment to terminal differentiation is not

Extracellular matrix receptors of the integrin family have a dual role in the epidermis, regulating both adhesion and differentiation. Loss of contact with the extracellular matrix causes keratinocytes to become committed to terminal differentiation, and results in a decrease in the ability of the alpha 5 beta 1 integrin to bind fibronectin. We have investigated whether the decrease in ligand-binding ability is reversible and, if so, whether commitment to terminal differentiation can also be reversed. Keratinocytes that had been placed in suspension for 5 hours to induce commitment were compared with the starting population (0 hour cells) in the presence or absence of 8A2, an activating anti-beta 1 antibody. 8A2 IgG or FAb fragments increased the amount of alpha 5 beta 1 in cell extracts that bound to fibronectin-Sepharose and in the presence of 8A2 the amount of bound alpha 5 beta 1 in 0 hour and 5 hour extracts was equal. 8A2 also restored alpha 5 beta 1 function in adhesion assays of intact 5 hour cells. Ca2+, Mg2+ and Mn2+ alone, at concentrations of up to 1 mM, did not increase the adhesiveness of 5 hour cells relative to 0 hour cells; however, the effect of 8A2 on keratinocytes was dependent on Ca2+. Although 8A2 restored alpha 5 beta 1 ligand-binding ability it did not prevent committed cells from withdrawing from the cell cycle and expressing involucrin, a differentiation marker.(ABSTRACT TRUNCATED AT 250 WORDS)

Distinctive integrin expression in the newly forming epidermis during wound healing in humans

The integrin receptor family plays a fundamental role in mediating cell attachment to a variety of extracellular matrix molecules. In normal human epidermis, the alpha 2 beta 1, alpha 3 beta 1, alpha 6 beta 4, and alpha v beta 5 integrin heterodimers are expressed and appear largely confined to the basal cell layer. In the present study, beta 1, beta 4, and alpha v integrin expression in the epidermis during wound healing in humans was examined. Punch biopsies were performed on healthy volunteers. At daily intervals up to day 8, and at days 11, 14, 21, and 28, the wound site was surgically removed. Using immunofluorescence microscopy, several modifications of the integrin expression pattern were observed on migrating keratinocytes during the re-epithelialization phase of the wound-healing process: i) alpha v expression was strongly enhanced and polarized at the basal pole of basal keratinocytes; ii) among the beta 1 integrins, alpha 3 beta 1 was overexpressed and distributed over the entire basal keratinocyte membrane and a weak alpha 5 beta 1 reactivity became evident; and iii) alpha 6 beta 4 was detected as a linear staining along the newly forming dermal-epidermal junction. Moreover, both during the re-epithelialization phase and during the first 2 weeks after wound closure, alpha 3, alpha 6, alpha v, beta 1, and beta 4 were no longer confined to the basal layer, as in normal epidermis, but were also found on several suprabasal cell layers. These results suggest that alpha v beta 5, alpha 3 beta 1, and alpha 5 beta 1 may be the main integrin receptors mediating keratinocyte spreading and migration over the provisional matrix of the wound bed.

Regulation of keratinocyte terminal differentiation by integrin-extracellular matrix interactions

Suspension-induced terminal differentiation of human epidermal keratinocytes can be inhibited by fibronectin through binding to the alpha 5 beta 1 integrin. We have investigated the effect of fibronectin on expression of integrins and proteins of the actin cytoskeleton and have explored the nature of the differentiation stimulus by testing different combinations of anti-integrin monoclonal antibodies or extracellular matrix proteins in the suspension assay. Fibronectin prolonged cell surface expression of beta 1 integrins but did not overcome the inhibition of intracellular transport of integrins that occurs when keratinocytes are placed in suspension. Fibronectin did not prevent the suspension-induced decline in the level of mRNAs encoding the beta 1 integrin subunit, actin, filamin and alpha-actinin; furthermore, the inhibition of terminal differentiation did not depend on the state of assembly of microfilaments or microtubules. Terminal differentiation could be partially inhibited by an adhesion-blocking monoclonal antibody to the beta 1 integrin subunit or by a combination of adhesion blocking antibodies recognising the alpha subunits that associate with beta 1 (alpha 2, alpha 3 and alpha 5). Although laminin and type IV collagen do not inhibit terminal differentiation individually, they were inhibitory when added to cells in combination with a low concentration of fibronectin. We conclude that the proportion of keratinocyte beta 1 integrins occupied by ligand can regulate the initiation of terminal differentiation independently of the state of assembly of the actin cytoskeleton.

Expression and topography of integrins and basement membrane proteins in epidermal carcinomas: basal but not squamous cell carcinomas display loss of alpha 6 beta 4 and BM-600/nicein

The expression and topography of some integrins and basement membrane proteins in cutaneous basal cell carcinomas (BCCs) and squamous cell carcinomas (SCCs) have been studied by immunohistochemistry and Western blotting. It has been shown that the typical cell-to-cell distribution of alpha 2 beta 1 and alpha 3 beta 1 found in normal epidermis is replaced by pericellular distribution in both BCC and SCC cells. BCC and SCC also showed different patterns of expression of alpha 6 beta 4, an integrin heterodimer normally lining the basal surface of basal epidermal keratinocytes: whereas SCC showed high expression and pericellular distribution of alpha 6 beta 4, BCC cells did not express this integrin at all. The absence of alpha 6 and beta 4 subunits from BCC extracts was confirmed by Western blotting. The molecular composition of the basement membrane was markedly different in the two types of epidermal tumors. Whereas laminin and collagen type IV were conserved in the basement membrane zone of both tumors, the molecular complex BM-600/nicein, which is recognized by the monoclonal antibody GB3 and is possibly identical to the previously described basement membrane glycoproteins kalinin and epiligrin, was absent from BCC cells. Then, the simultaneous loss of expression of alpha 6 beta 4 and BM-600/nicein in BCC cells but not in SCC cells indicates that alpha 6 beta 4 integrin and one of its potential ligands may be co-regulated in both BCC and SCC, thus suggesting a role for this phenomenon in the pathogenesis and clinical behavior of these epidermal tumors.

Evidence against a major role for integrins in calcium-dependent intercellular adhesion of epidermal keratinocytes

It is well established that integrins mediate keratinocyte adhesion to extracellular matrix proteins, but, in addition, there is some evidence that they mediate intercellular adhesion. We have investigated the role of integrins in keratinocyte-keratinocyte adhesion by adding anti-integrin antibodies to cells in three assays that differ according to the calcium ion concentration of the medium, the presence or absence of an adhesive substrate (glass or tissue culture plastic) and the timing of antibody addition. As previously reported by Larjava et al., (J. Cell Biol. 110:803-815), a monoclonal antibody to the beta 1 subunit perturbed cell-cell adhesion when added to adherent monolayers in low calcium medium (0.1 mM calcium ions), but did not prevent cell-cell adhesion or stratification induced by raising the level of calcium ions to 1.8mM (the concentration in standard medium). Monoclonal antibodies to both the alpha 3 and beta 1 subunits inhibited the attachment, spreading and motility of keratinocytes in low or standard calcium medium when added at the time of plating; however, they had only a modest effect on the accumulation of cells in adherent clusters. Aggregation of keratinocytes in suspension required a calcium ion concentration of greater than 0.1mM and was not inhibited by any of a large panel of anti-integrin antibodies, including three new antibodies that recognise alpha 2 beta 1. We conclude that any inhibitory effects of individual anti-integrin antibodies on cell-cell adhesion are abrogated by a calcium ion concentration above 0.1mM and that in low calcium medium at least some of the inhibition of cell-cell adhesion is a consequence of the inhibition of cell-substrate adhesion and motility.

Integrin expression in normal, hyperplastic, dysplastic, and malignant oral epithelium

We have examined the distribution of a range of integrin subunits in normal and lesional oral mucosa. The alpha 2, alpha 3, alpha 6, beta 1, and beta 4 subunits were highly expressed in normal epithelium, and there was weaker, more variable expression of alpha 5 and alpha v. Expression of all subunits was highest in the basal layer of normal epithelium, but extensive staining above the basal layer was also observed, particularly in the floor of the mouth and the lateral margin of the tongue. In dysplastic lesions and hyperplastic epithelium adjacent to ulcers, suprabasal staining was even more pronounced. Staining patterns in squamous cell carcinomas showed considerable variation, both within and between individual tumours: in some areas there was staining reminiscent of normal epithelium, but uniform staining throughout tumour islands, and patchy and variable cytoplasmic and pericellular staining were also seen. Thirteen out of 17 carcinomas showed some loss of integrin expression: six out of ten moderately well differentiated tumours and all the poorly differentiated tumours. Focal loss of alpha 6 and beta 4 was most commonly observed, but loss of alpha 2 and alpha 3 also occurred. Since integrins regulate not only keratinocyte adhesion, but also the initiation of terminal differentiation, the changes in integrin expression that we have observed may have significance for the behaviour of individual tumours.

Characterization of the internalization of bacillus Calmette-Guerin by human bladder tumor cells

Adjuvant intravesical Mycobacterium bovis BCG is the treatment of choice for recurrent superficial bladder cancer. Fibronectin (FN) was previously demonstrated to be necessary for the retention of BCG within the bladder and for the expression of antitumor activity. Recent studies have demonstrated that BCG attach and are ingested by bladder epithelial cells, suggesting the existence of a second bacterial attachment mechanism. We report the characterization of the molecules involved in BCG attachment and internalization by the human bladder transitional cell carcinoma cell line T-24. Pretreatment of T-24 cells with monoclonal antibodies to either alpha 5 or beta 1 integrin subunits significantly inhibited both BCG attachment and ingestion. Exogenous FN was observed to enhance both attachment and ingestion of BCG, and anti-FN was observed to inhibit both phenomena. Latex beads precoated with either FN or laminin (LN) but not BSA were ingested by T-24 cells, but only FN-coated beads inhibited BCG attachment and ingestion. Pretreatment of BCG with FN augmented both attachment and ingestion. The role of bacterial FN binding proteins was evaluated. A monoclonal antibody to a 55-kD FN-binding protein was observed to abrogate attachment and ingestion. These results demonstrate that attachment and ingestion of BCG are mediated in part by the alpha 5 beta 1 integrin receptor and are dependent on FN. These studies demonstrate a mechanism of entrance of mycobacteria into epithelial cells and suggest a second role for FN in the adjuvant antitumor effect of BCG.

Altered glycosylation and cell surface expression of beta 1 integrin receptors during keratinocyte activation

We studied the mechanism by which cell adhesiveness becomes activated when keratinocytes are removed from skin and placed into cell culture. Our results suggest that activation involves altered beta 1 integrin subunit glycosylation accompanied by an increase in cell surface beta 1 integrin receptors. Activated keratinocytes contained two forms of the beta 1 integrin subunit, approximately 93 kDa and approximately 113 kDa. As shown by pulse-chase experiments, the smaller represented the cytoplasmic precursor of the larger, and only the 113 kDa mature form was detected in integrin receptors expressed at the cell surface. Pre-activated keratinocytes contained beta 1 integrin subunits ranging from approximately 97 to 110 kDa. These beta 1 subunits had been processed through the Golgi, based on resistance to endoglycosidase-H treatment, and were not converted to 113 kDa subunits during subsequent cell culture. Experiments with endoglycosidase-F showed that differences in the apparent sizes of beta 1 integrin subunits observed in pre-activated and activated keratinocytes could be attributed to differences in subunit glycosylation. Smaller beta 1 subunits found in pre-activated keratinocytes, like the precursor beta 1 subunits of activated cells, appeared to be less efficient in reaching the cell surface. Overall, a approximately 10-fold increase in the level of cell surface integrin receptors occurred concomitant with the increased proportion of 113 kDa beta 1 subunits found in activated cells. Endoglycosidase-F experiments also indicated that there were changes in keratinocyte alpha subunits associated with beta 1. In related experiments, keratinocytes cultured in low Ca2+, serum-free MCDB medium for 4 days proliferated but their adhesiveness did not become activated. Therefore, keratinocyte proliferation and activation of adhesion are regulated separately. Finally, substantial activation of keratinocytes was observed when serum was added to cells cultured in MCDB with serum, indicating a role for serum factors in the activation process.

Human epidermal Langerhans cells express beta 1 integrins that mediate their adhesion to laminin and fibronectin

Members of the beta 1 or very late antigen (VLA) integrin family represent the predominant class of integrin extracellular matrix receptors. Adhesion assays were developed for the identification of the beta 1 integrins involved in the adhesive interactions between Langerhans cells (which mainly express alpha 4 beta 1, alpha 5 beta 1, and alpha 6 beta 1) and extracellular matrix proteins. For this purpose, binding assays were performed on fibronectin-, laminin-, collagen type IV-, and collagen type I-coated plates. 59% +/- 21% of Langerhans cells (LC) specifically attached to fibronectin. Using as inhibitory probes monoclonal antibodies against the beta 1, alpha 5, and alpha 3 chains and the synthetic peptide GRGDSP resulted in a decrease of 43%, 41%, 15%, and 42% respectively of LC binding to fibronectin. 76% +/- 20% of LC specifically adhered to laminin. Anti-alpha 6 monoclonal antibody potently inhibited this adhesion, which dropped to 36%, whereas the synthetic peptide GRGDSP was ineffective. A low number of LC adhered to type I and type IV collagen (13-15%). These results indicate that alpha 5 beta 1 and alpha 6 beta 1 were the major beta 1 integrins involved in LC adhesion to fibronectin and laminin. Ultrastructural cell morphology of adherent cells was examined and showed that LC were largely spread on laminin and became tightly bound to the substrate on a large portion of membrane. On fibronectin surface, the contact between LC and substrate was smaller, thus cells could conserve their general round aspect. Moreover, LC binding to fibronectin and laminin induced a significative decrease of the Birbeck granule number. The finding that LC attach to LM and FN in vitro suggests they exist similarly in vivo. By mediating a passage through basement membrane and migration throughout the fibronectin network of the dermis, alpha 5 beta 1 and alpha 6 beta 1 could contribute to the ability of LC to migrate into and out of the epidermis.

Identification and characterization of a novel high-molecular-weight form of the integrin alpha 3 subunit

The integrin alpha 3 beta 1 is a multiligand extracellular matrix receptor found on many cell types. Immunoprecipitations of 125I-surface-labeled prostate carcinoma cell lines, DU145 and PC-3, with the anti-alpha 3 integrin monoclonal antibodies J143 or PIB5, resulted in the coimmunoprecipitation, along with the expected alpha 3 beta 1 heterodimer, of a polypeptide with a molecular mass of 225 kDa. This protein could also be copurified with the 155-kDa alpha 3 and 115-kDa beta 1 subunits upon affinity chromatography of 125I-surface-labeled cell extracts on anti-alpha 3 antibody-Sepharose columns. Upon reduction, this 225-kDa protein generated 130- and 95-kDa polypeptides, while the 155-kDa alpha 3 subunit generated 130- and 25-kDa polypeptides. The 225-kDa protein did not generate a 25-kDa polypeptide. Deglycosylation and reduction of the 225-kDa protein resulted in the generation of 110- and 95-kDa polypeptides, while deglycosylation and reduction of the 155-kDa alpha 3 resulted in a 110-kDa polypeptide identical in size to the 110-kDa polypeptide generated from the 225-kDa protein. Peptide maps generated from the 110-kDa components of the 225-kDa polypeptide and the 155-kDa alpha 3 integrin subunit were identical, as were their N-terminal amino acid sequences. An antibody directed against the cytoplasmic domain of the alpha 3 subunit immunoprecipitated the 225-kDa polypeptide in addition to the 155-kDa alpha 3 subunit. Furthermore, Northern blot analysis of RNA from DU145 and PC-3 cells with a human alpha 3 cDNA probe identified an mRNA species of 6.2 kb in addition to a major mRNA species of 4.3 kb. The larger mRNA species, which is of an appropriate size for encoding a polypeptide of approximately 220-kDa, was not detectable in cells which did not express the 225-kDa protein. These data demonstrate that the 225-kDa polypeptide represents a novel integrin alpha 3 subunit consisting of the alpha 3 integrin heavy chain disulfide-bonded to a 95-kDa polypeptide which may represent an alternative "light" chain to the 25-kDa light chain of the alpha 3 subunit.

Aberrant integrin expression during epidermal wound healing and in psoriatic epidermis

We have examined integrin expression during the remodeling of the epidermis that takes place during wound healing, using a suction blister model in which the epidermis is detached from the dermis, leaving the basement membrane intact. By immunofluorescence microscopy, we found that the same integrin subunits were expressed during wound healing as in normal epidermis with very little change in the relative intensity or distribution of staining at the leading edge of the migrating epidermis. However, at the time of wound closure, when the epidermis is still hyperproliferative, alpha 2, alpha 3, alpha 6, and beta 1 were no longer confined to the basal layer, as in normal epidermis, but were also found in all the living suprabasal cell layers, coexpressed with the terminal differentiation markers involucrin, keratin 10, and keratin 16. Strong suprabasal staining for alpha v was also found in one specimen. beta 4, which normally forms a heterodimer with alpha 6, and alpha 5 remained predominantly basal. Three of the integrin ligands, fibronectin, type IV collagen, and laminin, remained largely confined to the basement membrane zone and dermis. By 14 d after wounding, the integrins were once more restricted to the basal layer. Suprabasal integrin expression was also observed in involved psoriatic lesions. Thus, in two situations in which the epidermis is hyperproliferative, there is a failure to downregulate integrin expression on initiation of terminal differentiation. The functional consequences of this aberrant integrin expression remain to be explored.

Expression of beta 1, beta 3, beta 4, and beta 5 integrins by human epidermal keratinocytes and non-differentiating keratinocytes

We have compared the adhesive properties and integrin expression profiles of cultured human epidermal keratinocytes and a strain of nondifferentiating keratinocytes (ndk). Both cell types adhered to fibronectin, laminin, and collagen types I and IV, but ndk adhered more rapidly and at lower coating concentrations of the proteins. Antibody blocking experiments showed that adhesion of both cell types to fibronectin was mediated by the alpha 5 beta 1 integrin and to laminin by alpha 3 beta 1 in synergy with alpha 2 beta 1. Keratinocytes adhered to collagen with alpha 2 beta 1, but an antibody to alpha 2 did not inhibit adhesion of ndk to collagen. Both cell types adhered to vitronectin by alpha v-containing integrins. Immunoprecipitation of surface-iodinated and metabolically labeled cells showed that in addition to alpha 2 beta 1, alpha 3 beta 1, and alpha 5 beta 1, both keratinocytes and ndk expressed alpha 6 beta 4 and alpha v beta 5. ndk expressed all these integrins at higher levels than normal keratinocytes. ndk, but not normal keratinocytes, expressed alpha v beta 1 and alpha v beta 3; they also expressed alpha 1 beta 1, an integrin that was not consistently detected on normal keratinocytes. Immunofluorescence experiments showed that in stratified cultures of normal keratinocytes integrin expression was confined to cells in the basal layer; terminally differentiating cells were unstained. In contrast, all cells in the ndk population were integrin positive. Our observations showed that the adhesive properties of ndk differ from normal keratinocytes and reflect differences in the type of integrins expressed, the level of expression and the distribution of integrins on the cell surface. ndk thus have a number of characteristics that distinguish them from normal basal keratinocytes.

HIV-1 infection of human T lymphocytes results in enhanced alpha 5 beta 1 integrin expression

Altered T cell adherence after human immunodeficiency virus 1 (HIV-1) infection may contribute to viral pathogenesis in the acquired immune deficiency syndrome. To address this hypothesis, we assessed mechanisms of T cell adherence to extracellular matrix proteins in vitro. We found that after HIV-1 infection, both chronically infected H9 CD4+ T cells and acutely infected primary peripheral blood lymphocytes acquired the ability to adhere to the extracellular matrix glycoprotein fibronectin, to a lesser extent to type IV collagen and laminin, but not to type I collagen. H9 cells chronically infected with two of the three HIV-1 strains studied showed approximately a sevenfold increase in attachment to fibronectin, while the same cells infected with the human retrovirus HIV-2 did not. Adhesion was accompanied by changes in morphology, including marked spreading and increased filopodia. These alterations were not blocked by the protein kinase C inhibitor H-7, which did inhibit TPA-induced T cell attachment to fibronectin. Monoclonal antibodies against both the alpha 5 and the beta 1 subunits of the classical fibronectin receptor as well as an Arg-Gly-Asp (RGD) peptide inhibited attachment, whereas anti-alpha 4 monoclonal antibodies and the CS1 peptide did not. Binding to collagen IV was also inhibited by the anti-beta 1 monoclonal antibody, but not the other antibodies. Cells metabolically labeled with [35S]methionine and analyzed by immunoprecipitation with polyclonal anti-beta 1 integrin antibody showed a 2.5-fold increase in integrin synthesis in infected cells compared to uninfected controls. This increase in synthesis was associated with an increase in cell surface expression of both alpha 5 and beta 1 integrins by FACS (registered trademark of Becton Dickinson for a fluorescence-activated cell sorter) analysis. Enhanced expression of integrins such as alpha 5 beta 1 may cause T cell adherence to a variety of tissues, where released viral gene products may induce some of the tissue-specific manifestations of HIV-1 infection.

Expression and endocytosis of integrin VLA receptors for collagen, fibronectin and laminin by normal human keratinocytes

The Very Late Activation (VLA) antigen family is involved in cell-extracellular matrix interactions and consists of six heterodimeric cell surface receptors with a common beta 1 and a variable alpha subunit. Using a panel of specific antibodies, we showed that human epidermal basal cells expressed VLA-2, VLA-3 and VLA-6 but failed to express VLA-4. Their functional roles were investigated and VLA-2 appeared as a specific receptor for type IV collagen and also as a laminin receptor. VLA-3 appeared as a receptor for fibronectin and laminin and to a lesser extent as a type I collagen receptor. VLA-6 appeared as a specific receptor for laminin. It also appeared that the VLA-alpha subunit specifically mediates the recognition of ligand but the beta 1 subunit is also involved in adhesion and that both subunits have a synergistic influence. Immunoprecipitation analyses confirmed that VLA-2, VLA-3 and VLA-6 were expressed by basal keratinocytes. Endocytosis of VLA-2 and VLA-3 was observed involving coated vesicles and endosomes that are structures characteristic of a receptor-mediated pathway. These findings provide first evidence that normal human basal keratinocytes are able of endocytosis mediated by receptors. Taken together, these results indicate that multiple VLA receptors function in combination to mediate epidermal basal cell adhesion to extracellular matrix.

Human epidermal Langerhans cells express integrins of the beta 1 subfamily

The expression of VLA integrins on human epidermal Langerhans cells (LC) was investigated by indirect immunogold labeling on transmission electron microscopy, followed by a quantitative analysis. Labelings on suspensions enriched in freshly isolated LC were carried out with an antibody recognizing the beta 1 subunit, common to all members of the VLA family, and with antibodies specific for the six different alpha subunits, alpha 1 to alpha 6. Normal human epidermal LC were all beta 1 positive, of which 60% were highly positive. By labelings with different VLA-alpha-chain MoAb, there appeared two subpopulations of LC: one positive and one negative. On average 40% of LC bore small amounts of VLA-1 and VLA-3, 53% and 77% of LC expressed moderate amounts of VLA-2 and VLA-5, respectively, and VLA-4 and VLA-6 were expressed by 67% and 90% of LC, respectively. VLA proteins are mainly extracellular matrix protein receptors. VLA-6 (laminin receptor) and VLA-5 (fibronectin receptor) are expressed mainly by LC, and in this way could subserve LC to leave the epidermal compartment through the basement membrane, to migrate throughout the fibronectin network of the dermis before migrating via the afferent lymphatics to the regional lymph nodes, where they present antigen to T cells. VLA proteins such as VLA-4, VLA-3, or VLA-2, which are found involved in cell-cell contacts, could contribute to the promotion of T-cell activation by facilitating adherence between LC and T cells.

Polarized expression of integrin receptors (alpha 6 beta 4, alpha 2 beta 1, alpha 3 beta 1, and alpha v beta 5) and their relationship with the cytoskeleton and basement membrane matrix in cultured human keratinocytes

In human keratinocytes cultured in conditions which allow differentiation and stratification and are suitable to reconstitute a fully functional epidermis, alpha 6 beta 4 and two members of the beta 1 integrin family (alpha 2 beta 1 and alpha 3 beta 1) were respectively polarized to the basal and lateral domains of the plasmamembrane both in growing colonies and in the reconstituted epidermis. Conversely, the alpha v integrin subunit, presumably in association with beta 5, was expressed at the basal surface in growing and migrating but not in stationary keratinocytes. The integrin alpha 6 beta 4: (a) was organized in typical patches which often showed a "leopard skin" pattern where spots corresponded to microfilament-free areas; (b) was not associated with focal contacts containing vinculin and talin but rather corresponded to relatively removed contact areas of the basal membrane as shown by interference reflection microscopy; and (c) was coherent to patches of laminin secreted and deposited underneath the ventral membrane of individual cells. The two beta 1 integrins (alpha 2 beta 1 and alpha 3 beta 1), both endowed with laminin receptor properties, were not associated with focal adhesions under experimental conditions allowing full epidermal maturation but matched the lateral position of vinculin (but not talin), cingulin, and desmoplakin, all makers of intercellular junctions. Often thin strips of laminin were observed in between the lateral aspects of individual basal keratinocytes. The integrin complex alpha v beta 5 had a topography similar to that of talin- and vinculin-containing focal adhesions mostly in the peripheral cells of expanding keratinocyte colonies and in coincidence with fibronectin strands. The discrete topography of beta 1 and beta 4 integrins has a functional role in the maintenance of the state of aggregation of cultured keratinocytes since lateral aggregation was impaired by antibodies to beta 1 whereas antibodies to beta 4 prevented cell-matrix adhesion (De Luca, M., R. N. Tamura, S. Kajiji, S. Bondanza, P. Rossino, R. Cancedda, P. C. Marchisio, and V. Quaranta. Proc. Natl. Acad. Sci. USA. 87:6888-6892). Moreover, the surface polarization of integrins followed attachment and depended both on the presence of Ca2+ in the medium and on the integrity of the cytoskeleton. We conclude that our in vitro functional tests and structural data suggest a correlation between the pattern of integrin expression on defined plasmamembrane domains and the mechanism of epidermal assembly.

Structural and functional studies of integrin receptors in cultured human keratinocytes

In human keratinocytes cultured in conditions which allow differentiation and stratification and which are suitable to reconstitute a fully functional epidermis, the integrin alpha 6 beta 4 and two members of the beta 1 integrin family (alpha 2 beta 1 and alpha 3 beta 1) were polarized to the basal and lateral domains of the plasma membrane both in growing colonies and in the reconstituted epidermis. Conversely, alpha v beta 5 integrin was expressed at the basal surface in growing and migrating but not in stationary keratinocytes. The integrin alpha 6 beta 4 was organized in patches and spots corresponding to F-actin-free submembraneous areas and did not colocalize to focal contacts; moreover, alpha 6 beta 4 colocalized with patches of laminin deposited underneath the ventral membrane of individual cells. The two beta 1 laminin receptor integrins (alpha 2 beta 1 and alpha 3 beta 1) were never found in the basal domain but matched the lateral position of vinculin (but not talin), cingulin and desmoplakins. Only the integrin complex alpha v beta 5 was associated with talin- and vinculin-containing focal adhesions mostly in the peripheral cells of expanding keratinocyte colonies and in coincidence with fibronectin strands. The topography of beta 1 and beta 4 integrins reflects a functional role in adhesion and in the maintenance of the state of aggregation of cultured keratinocytes since lateral aggregation was impaired by antibodies to beta 1 whereas antibodies to beta 4 prevented cell-matrix adhesion.