The receptor tyrosine kinase Axl regulates cell-cell adhesion and stemness in cutaneous squamous cell carcinoma.

Axl is a receptor tyrosine kinase (RTK) upregulated in various tumors including cutaneous squamous cell carcinoma (SCC). Axl expression correlates with poor prognosis and induction of epithelial-mesenchymal transition (EMT), hence we hypothesized that Axl is involved in the disruption of cell-cell adhesion to allow invasion and chemotherapy resistance of the cancer stem cell population. Cutaneous SCC cell lines with stable knockdown of Axl were generated using retroviral vectors. Axl depletion altered expression of intercellular junction molecules increasing cell-cell adhesion with downregulation of Wnt and TGFßR signaling. Furthermore, Axl expression correlated with the expression of putative cancer stem cell markers, CD44 and ALDH1, increased resistance to chemotherapy drugs, enhanced sphere formation ability and expression of EMT features by cancer stem cells. Axl depletion resulted in loss of tumor formation in an in vivo zebrafish xenograft model. In conclusion, these data suggest that abrogation of Axl results in loss of cancer stem cell properties indicating a role for Axl as a therapeutic target in chemotherapy-resistant cancer.

Modulation of gingival epithelial phenotypes by interactions with regionally defined populations of fibroblasts.

BACKGROUND AND OBJECTIVE: The unusual structure and functions of junctional epithelium, together with its pattern of migration in periodontal disease, raise interesting questions about the factors associated with the maintenance of its unique phenotype. To explore the effects of regionally differing fibroblast populations on the growth and patterns of differentiation of oral epithelia, this study used an organotypical in vitro model in an attempt to detect interactions occurring between populations of human oral fibroblasts and keratinocytes. MATERIAL AND METHODS: Keratinocytes and fibroblasts, isolated from the gingival region and periodontal ligament, were characterized by their patterns of growth and by their expression of known differentiation markers. Changes in cell behaviour and phenotypic marker expression were examined during in vitro passage as an indication of the maintenance of in vivo phenotypic traits. Using early passage cells, organotypical cultures were generated and patterns of epithelial growth and expression of phenotypic markers were examined. RESULTS: Phenotypically different populations of junctional and oral-gingival keratinocytes, and of oral-gingival and periodontal ligament fibroblasts, were successfully isolated, cultured and characterized. In the organotypic culture system, oral-gingival fibroblasts were found to have a markedly greater ability than periodontal ligament fibroblasts to support and maintain the growth of either type of epithelium. Shifts of epithelial phenotype were induced by different fibroblasts. CONCLUSION: Periodontal and gingival fibroblast subpopulations have differential effects on the growth and patterns of differentiation of oral and junctional epithelia. By modulating the epithelial phenotype, regionally differing fibroblasts can influence the stability and behaviour of the gingival attachment apparatus in health and disease.

The in vitro behaviour and patterns of colony formation of murine epithelial stem cells.

OBJECTIVE: The mechanisms of renewal of skin and mucosal epithelia in vivo are associated with hierarchies of stem and amplifying cells organized in distinct spatial patterns. Stem and amplifying characteristics persist after isolation and growth of human keratinocytes in vitro but the pattern for murine keratinocytes has been less clear. MATERIALS AND METHODS: Murine keratinocytes were grown in low calcium media and examined for their patterns of colony morphologies. RESULTS: We consistently identified three types of colonies, one of which contains concentric zones of amplifying and differentiated cells surrounding a central zone of cells that have patterns of expression and behavioural characteristic of stem cells. This zonal organization facilitated analysis of stem cell formation and loss. Cells in the central stem cell zone undergo rapid symmetric divisions but expansion of this population is partially limited by their peripheral transition into amplifying cells. A striking feature of central zone cells is their enhanced apoptotic susceptibility and stem cell expansion limited by consistently high background rates of apoptosis. This occasionally reaches catastrophic levels with elimination of the entire central zone. CONCLUSION: In vitro amplification of stem cells for the generation of engineered tissue has tended to focus on control of asymmetric division but these findings suggest that development of mechanisms protecting stem cells from apoptotic changes are also likely to be of particular value.

Cancer stem cells - new and potentially important targets for the therapy of oral squamous cell carcinoma.

There is increasing evidence that the growth and spread of cancers is driven by a small subpopulation of cancer stem cells (CSCs) - the only cells that are capable of long-term self-renewal and generation of the phenotypically diverse tumour cell population. Current failure of cancer therapies may be due to their lesser effect on potentially quiescent CSCs which remain vital and retain their full capacity to repopulate the tumour. Treatment strategies for the elimination of cancer therefore need to consider the consequences of the presence of CSCs. However, the development of new CSC-targeted strategies is currently hindered by the lack of reliable markers for the identification of CSCs and the poor understanding of their behaviour and fate determinants. Recent studies of cell lines derived from oral squamous cell carcinoma (OSCC) indicate the presence of subpopulations of cells with phenotypic and behavioural characteristics corresponding to both normal epithelial stem cells and to cells capable of initiating tumours in vivo. The present review discusses the relevance to OSCC of current CSC concepts, the state of various methods for CSC identification, characterization and isolation (clonal functional assay, cell sorting based on surface markers or uptake of Hoechst dye), and possible new approaches to therapy.

Development of a quantitative method to analyse tumour cell invasion in organotypic culture.

Tumour invasion is a dynamic process occurring in three dimensions and involving interactions between both tumour and stromal cells. Experimental analysis of squamous carcinoma cell invasion has often used the organotypic gel culture system, which is generated by plating tumour cells on to a synthetic stroma composed of a collagen gel embedded with fibroblasts. Unfortunately, quantitation of invasion in these organotypic gels has relied largely on subjective pathological opinion, which may be influenced by different patterns of tumour cell infiltration. Therefore a computer-assisted digital image analysis system that assesses invasion objectively and provides a numerical 'Invasion Index' was developed. The Invasion Index, by combining depth and pattern of invasion in a single value, establishes a quantitative value that allows assessment of the influences of positive and negative regulation of tumour invasion. These data demonstrate that the organotypic gel system is a robust, accurate, and reproducible method for measuring tumour cell invasion. They also show that the Invasion Index can be used after organotypic gels have been implanted in mice for up to 6 weeks. Illustrative examples of how various factors influence the invasion of squamous carcinoma cells in three dimensions both in vitro and in vivo are provided.

Regulation of keratinocyte growth factor and scatter factor in cyclosporin-induced gingival overgrowth.

BACKGROUND: Epithelial proliferation is a histological characteristic of drug-induced gingival overgrowth. Keratinocyte growth factor (KGF) and scatter factor (SF) are fibroblast-derived growth factors with potent mitogenic and motogenic effects on epithelial cells, and, therefore, could be involved in the pathogenesis of gingival overgrowth. The aims of this study were to investigate: (i) the effects of cyclosporin on KGF and SF expression by gingival fibroblasts; and (ii) the expression levels of KGF and SF mRNA in normal and overgrown gingival tissue. METHODS: The KGF and SF protein production was determined by enzyme-linked immunosorbent assay. Relative levels of KGF and SF mRNA expression were determined using semi-quantitative reverse transcriptase polymerase chain reaction. Expression levels in biopsies of normal and overgrown gum were also determined. RESULTS: In overgrown fibroblasts, 500 ng/ml cyclosporin significantly inhibited KGF and SF mRNA and protein while 2000 ng/ml cyclosporin induced a stimulatory effect. In normal cells cyclosporin significantly increased both KGF and SF. KGF and SF mRNA was detected in both normal and overgrown tissues with a tendency towards increased expression levels in overgrown tissue. CONCLUSION: These results suggest that KGF and SF may have an important role in cyclosporin-induced gingival overgrowth.

Construction of large area organotypical cultures of oral mucosa and skin.

BACKGROUND: Skin and oral mucosal keratinocytes grown in vitro usually lose their normal patterns of differentiation, unless grown as organotypical cultures that are constructed by allowing collagen gels containing fibroblasts to contract before they are plated with keratinocytes and raised to the air/medium interface. However, the contraction process tends to produce small irregular cultures. METHODS: To generate uniformly differentiating large cultures, we have investigated several aspects of the factors involved in the culture construction. By adjusting the number of fibroblasts used and by plating the matrices with keratinocytes prior to contraction, cultures of up to 72 cm2 were constructed. RESULTS: The cultures retained almost the full surface areas of the original matrices and showed uniform patterns of epithelial plating and differentiation. Immunostaining for cytokeratins and integrins indicated restoration of in vitro phenotypes similar to those of the epithelial tissues of origin. CONCLUSIONS: These methods successfully generate cultures required for certain types of investigations and tissues that are suitable for clinical use as grafts.

Keratinocyte growth factor expression in human gingival fibroblasts and stimulation of in vitro gene expression by retinoic acid.

BACKGROUND: Keratinocyte growth factor (KGF) is a stromally derived growth factor of the fibroblast growth factor (FGF) family with paracrine effects targeted to influence the growth and differentiation of epithelia. Regional and temporal changes in KGF expression play important roles in the development and maintenance of epithelial structures and in epithelial wound healing. Differing patterns of expression of KGF by fibroblasts in the gingival region could therefore be related to the observed regional variation in the differentiation and behavior of gingival epithelia. METHODS: The in vitro and in vivo patterns of expression of KGF mRNA in human gingival and periodontal fibroblasts were examined using reverse transcription polymerase chain reactions (RT-PCR) and in situ hybridization with digoxigenin-labeled riboprobes. The patterns observed for human gingiva were compared with those for human skin and for murine tissues. RESULTS: Gingival and periodontal fibroblasts showed expression of KGF transcripts in vitro, and the degree of expression was markedly influenced by the presence of retinoic acid, an agent known to influence patterns of epithelial differentiation. Sections of human and murine gingiva and skin showed regionally variable expression of transcripts with the cells expressing KGF in the subepithelial, rather than the deeper, connective tissues and periodontium. CONCLUSIONS: The results point to a role of KGF in the maintenance of normal growth and differentiation of gingival epithelia. A lack of KGF expression by periodontal fibroblasts in vivo is expected to hinder apical epithelial migration and thus stabilize the epithelial attachment. The effects of retinoic acid (RA) on KGF expression in vitro provide an indirect mechanism by which RA may regulate the growth and differentiation of gingival epithelia.

Transforming growth factor-beta response and expression in junctional and oral gingival epithelial cells.

The junctional (JE) and oral gingival (OGE) epithelium show distinct morphological phenotypes and express different cell surface and keratin markers. Transforming growth factor-beta (TGF-beta) has been shown to stimulate extracellular matrix formation and inhibit proteolytic matrix degradation in periodontal wound healing. To elucidate potential roles of TGF-beta in gingival epithelial regeneration and reattachment, the present study examined the effects of TGF-beta on JE and OGE cell growth and determined the patterns of expression of mRNAs for the TGF-beta isotypes beta 1, beta 2 and beta 3 and TGF-beta receptor types I, II and III. Primary cell cultures were initiated from JE and OGE and the cell phenotypes confirmed using monoclonal antibodies to specific keratins. TGF-beta induced a significant growth inhibition in OGE cells derived from 6 different patients with a mean inhibition of 46% and a range of 16-70% (p = 0.031). Although responses varied between patients, in general maximum inhibition occurred at 10 ng/ml TGF-beta. JE cells from 5 patients showed no significant growth inhibition by TGF-beta (p = 0.125). Greater expression of TGF-beta 2 and receptor type I mRNA was found in OGE than JE cells and thus appeared to be associated with differentiating epithelial cells. JE cells expressed more TGF-beta type II receptor specific mRNA than did OGE cells, but TGF-beta 1 mRNA expression was similar in JE and OGE cells. JE or OGE cultures derived from 2 of 3 patients showed expression of mRNA for the TGF-beta type III receptor. TGF-beta 3 mRNA was not detected in any of the JE or OGE samples examined. The greater sensitivity of OGE than JE to the growth inhibiting effects of TGF-beta correlated with higher expression of receptor type I mRNA which, together with the type II receptor, is required for sensitivity to growth inhibition by TGF-beta. The results suggest that, in addition to structural differences, the development of functional differences in the responses of JE and OGE to TGF-beta may be associated with the formation of JE from OGE cells and the reformation of attachment after periodontal surgery.

Retroviral transduction of murine epidermal stem cells demonstrates clonal units of epidermal structure.

It has been suggested that the number and position of epidermal stem cells are related to the units of columnar structure in the upper epidermal strata and that the cells of each unit are derived from a single stem cell. Studies of cell lineage in developing tissues have been facilitated by the use of retroviral transduction to provide inherited expression of a histochemically demonstrable foreign gene product. To provide direct evidence about the clonal nature of epidermal units, murine epidermal keratinocytes were transduced with a replication-deficient retroviral vector carrying the beta-galactosidase gene. Subepidermal injection of virus in vivo led to infrequent transduction with only transient presence of beta-gal-staining keratinocytes within the epidermis. Transduction of keratinocytes in vitro and transplantation back to in vivo sites permitted demonstration of the transduced gene in clusters of cells within the reformed epidermis throughout a 12-wk period. The epidermis redeveloped an ordered columnar structure with restriction of transduced cells to individual columnar units. This clonal appearance is compatible with derivation of each epidermal unit from a single stem cell but is not compatible with a random pattern of cell proliferation. Transduced epidermal sheets that were recombined with oral mucosal connective tissue also redeveloped normal columnar structure with restriction of beta-gal staining to individual columnar units. These data suggest that the establishment of an epidermal stem cell pattern related to units of structure is an intrinsic property of the epithelium and is not dependent on regionally-specific connective tissue influences.

Expression of keratinocyte growth factor in periapical lesions.

The epithelial proliferation associated with inflammatory periapical lesions and with periapical cyst formation represents an interesting but poorly understood pathological change. Keratinocyte growth factor (KGF) is a recently identified growth factor that is produced by stromal fibroblasts and acts specifically to stimulate epithelial growth and differentiation. To investigate its possible role in the activation of the normally quiescent rests of Malassez, we examined the expression of KGF by in situ hybridization of sections of normal periodontal ligament (PDL) and of 12 periapical granulomas or cysts. Normal PDL and periapical granulomas with scant inflammatory infiltration showed few cells expressing message for KGF. However, KGF-expressing cells were found in the connective tissue stroma close to dense foci of inflammatory cells and to proliferating epithelial elements and cystic epithelial linings. Examination of tissues by the reverse-transcription polymerase chain reaction (RT-PCR) showed KGF expression in 4 specimens of periapical lesions but low or undetectable levels in normal PDL. These observations suggest that the induction of KGF expression in the stromal cells of periapical lesions may play an important role in stimulating the epithelial proliferation associated with cyst formation.

Influence of retinoic acid on the expression of cytokeratins, vimentin and ICAM-1 in human gingival epithelia in vitro.

Phenotypic differences exist in vivo between junctional (JE) and oral gingival (OGE) epithelia and an in vitro system has been developed that maintains phenotypic differences. This system, which permits in vitro studies of factors that may influence the epithelial phenotype, was used to investigate the effects of retinoic acid (RA) on epithelial expression of various markers known to distinguish JE from OGE. Primary cultures of JE and OGE were initiated from defined gingival regions and were subcultured and grown for 48 h in 96-well plates or on multiple-well slides. Control cultures were grown in medium supplemented with delipidized serum and all-trans RA was added to experimental groups. Other cultures were grown in a defined RA-free medium. Cultures were examined using monoclonal antibodies against cytokeratins, vimentin, and ICAM-1 and binding displayed by indirect immunocytochemical staining. Staining reactions were assessed by direct microscopic observation and assayed by spectrophotometric quantitation. The results showed that RA had minor effects on the marker expression of JE but markedly enhanced expression of cytokeratins 8, 18, 19, vimentin and ICAM-1 in OGE. These markers, which normally distinguish JE from OGE, were expressed at levels approaching or exceeding those of control JE cultures. These observations indicate that RA responsive mechanisms affect the phenotypes expressed by epithelia in vitro and suggest that such mechanisms may be related to the different phenotypic patterns expressed by gingival epithelia in vivo.

Effect of corticosteroids on rabbits corneal keratocytes after photorefractive keratectomy.

BACKGROUND: To determine the corticosteroid effect on the activity and repopulation of keratocytes after photorefractive keratectomy (PRK). METHODS: A 193-nm excimer laser (VISX Twenty/Twenty) created a central ablation depth of 22 microns (diameter:5 nm) on 22 corneas of 16 albino rabbits. Two ablated eyes were examined 6 hours following PRK. Twelve eyes received no postoperative corticosteroids and eight were treated with topical fluoromethalone for 3 months. Corneas were examined 1, 3, 6, and 12 months after PRK by immunofluorescence and transmission electron microscopy. RESULTS: Corticosteroids reduced haze (p=0.02), but all corneas (treated or untreated) cleared 6 months after PRK. Keratocytes were absent from the anterior 100 microns of the stroma 6 hours after PRK. However, the number and activity of keratocytes were significantly greater in this area in untreated corneas at 1 month and then gradually decreased. By 6 and 12 months, the number of keratocytes approached controls. Treated corneas had fewer keratocytes than either controls or untreated eyes (p<0.01) and by 3 months, a subepithelial acellular zone of 30 to 50 microns thickness appeared and persisted until at last 12 months after PRK. CONCLUSIONS: Corticosteroids have a transient effect in reducing haze and seem to inhibit keratocyte movement, leading to an acellular subepithelial region beneath the ablated area.

Formation of normal gingival epithelial phenotypes around osseo-integrated oral implants in humans.

The oral, oral sulcular, and junctional epithelia of the natural gingiva each possess distinct patterns of differentiation that are demonstrable both ultrastructurally and by their individual patterns of macromolecular synthesis. The supracrestal tissues reformed around oral implants structurally resemble those of natural gingiva but little is known about phenotype changes occurring in the epithelia. To investigate whether peri-implant epithelia acquire similar patterns of differentiation to those of natural gingiva, biopsies from the supracrestal regions of five oral implants were examined by immunofluorescent methods using a panel of monoclonal antibodies with specificities for individual cytokeratins and ICAM-1, macromolecules which act as markers of the three gingival epithelial phenotypes. The observed staining patterns indicated the formation of oral, oral sulcular, and junctional epithelia which were phenotypically indistinguishable from those of natural gingival epithelia. This degree of reprogramming of epithelial gene expression is a surprising observation and the potential mechanisms leading to the development of those new epithelial phenotypes are discussed in the context of what is known about the development of natural gingiva, in terms of the possible effects of inflammation, and in relation to the known connective tissue influences on epithelial differentiation.

Expression of blood group-related glycoconjugates in the junctional and other oral epithelia of rodents.

BACKGROUND: The junctional epithelium (JE) attaches the gingiva to the non-vital tooth surface and has other unusual properties which protect the underlying periodontal tissues. The JE differs from other gingival and oral epithelia in its unusual expression of cytokeratins typical of both stratifying and of simple epithelia, a phenotypic pattern possibly related to its specialized functions. METHODS: The patterns of differentiation of rodent gingival and other epithelia were examined using monoclonal antibodies against various glycoconjugates which are expressed on epithelial cell surfaces and provide an alternative marker system for regionally-differing patterns of cell maturation. RESULTS: Markers that are typical of basal cells in other stratifying epithelia were expressed by all cell strata of JE. JE lacked differentiation markers typical of other stratifying oral epithelial but showed suprabasal expression of markers typically expressed by simple epithelia and specialized epithelia, such as taste buds. CONCLUSIONS: The phenotype of rodent JE differs from that of other oral epithelia and the pattern of differentiation assessed by its expression of glycoconjugates parallels that for other phenotypic markers, such as cytokeratins. Differentiation of rodent JE is similar to that of human JE. The functional significance of these patterns of expression is not yet clear but the markers characterizing this unusual epithelium in rodents may be associated with its behavior in periodontal disease and of value to experimental studies of its development.

Rates of clearance of the epithelial surfaces of mouse oral mucosa and skin.

Cell desquamation, by removing material adherent to the epithelial surface, appears to play an important role in limiting bacterial colonization of epithelia. Data are available concerning rates of epidermal surface clearance but similar data for oral mucosal epithelia, which exist in an environment more conducive to colonization, have not been reported. The rates of surface clearance of six regions of murine oral mucosa and skin were assessed by two different methods: (a) EM autoradiography to examine the rate of passage of a 3H-histidine label through the stratum corneum, and (b) measurement of rates of cell proliferation and of the size of the superficial cells of the stratum corneum. The autoradiographic (direct) method indicated regionally-varying epidermal clearance rates (6.3-15.9 h) which compared well with published data. This method indicated considerably faster (2.3-3.4 h) clearance rates for the three oral mucosal regions. The rates of surface clearance that were calculated indirectly, from rates of cell proliferation and of cell size, similarly showed faster clearance rates for mucosa (2.0-3.9 h) than for skin (6.0-17.9 h) but differed in the relative rates calculated for the individual regions of mucosa and skin.

Keratin expression in taste bud cells of the circumvallate and foliate papillae of adult mice.

The patterns of keratin expression of taste buds in murine oral mucosa were examined using a panel of antibodies with various specificities for cytokeratins. The patterns for taste buds differed markedly from those of the surrounding epithelium but no regional differences in the staining patterns of the taste buds themselves were detected. The taste buds and the ducts of von Ebner's glands were strongly stained by monoclonal antibodies (mAbs) against cytokeratins (K) 8, 18 and 19, those typically expressed by simple epithelia. Merkel cells, which are also present in the oral epithelium and may correspond to the type III cells in taste buds, stained for K8 and K18 but not K19. None of the mAbs against simple epithelial keratins stained the stratifying epithelium of the trench wall or of the oral mucosa. Antibodies with specificity for keratins that are expressed as differentiation products of the mucosal epithelium did not stain taste buds. The staining pattern of the epithelium of the trench wall indicated that it expressed some keratins typical of stratifying epithelia but lacked the full pattern of differentiation of the adjacent mucosal epithelia. Staining within the taste buds was not homogeneous but no clear differences of keratin staining could be directly related to the subtypes of cells constituting them. The markedly differing patterns of keratin expression between taste buds and the adjacent epithelium raises questions about the type of inductive signals that produce and maintain these patterns.

Patterns of cytokeratin expression in the epithelia of inflamed human gingiva and periodontal pockets.

Fourteen specimens of periodontal pockets and the associated marginal gingiva were collected and either frozen for examination using antibodies against various defined cytokeratin specificities or processed for 2-dimensional gel electrophoresis. The epithelium forming the pocket lining typically extended into the connective tissue of the pocket wall in the form of a network of finger-like strips. Immunocytological staining indicated that keratins (K) 5, 6, 14 and 19 were expressed by almost all cells of the pocket lining and K13 and K16 by the suprabasal cells. The coronal region of the pocket lining showed some cells staining for K4. Staining for K8 and K18 was seen in the apical region of the pocket lining and in the finger-like extensions of epithelium into the connective tissue. Compared with normal gingiva, the sulcular and the oral gingival epithelia showed a marked increase in staining for K19. Surprisingly, the pattern of keratin expression of the epithelium of the pocket lining was found to be essentially similar to that of normal junctional epithelium and the anatomical position of the boundaries between each epithelial phenotype were not significantly altered. These patterns of keratin expression were confirmed by the 2D electrophoretic analyses of microdissected regions of epithelium. The potential significance of inflammation to the epithelial changes associated with pocket formation is discussed.