Targeted inhibition of wound-induced PAI-1 expression alters migration and differentiation in human epidermal keratinocytes.

In the adult epidermis, keratinocytes do not normally express the type-1 inhibitor of plasminogen activator (PAI-1). Basal epithelial cell-specific PAI-1 synthesis, however, accompanies epidermal wound repair in vivo in which PAI-1 transcripts and immunoreactive protein are confined to epithelial cells in the migrating tongue and the hyperproliferative zone. A model system using human keratinocytes (HaCaT cells) was developed to assess functional relationships between epithelial growth state transitions and PAI-1 expression. PAI-1 synthesis was maximal in low population density, exponentially growing HaCaT cultures; relative PAI-1 mRNA and protein levels progressively declined as cells attained, and were maintained in, a postconfluent condition. While the fraction of PAI-1(+) keratinocytes remained stable (at approximately 85-90% of the population) throughout the culture period, both PAI-1 mRNA abundance and mean cell-associated PAI-1 protein declined by >90% during prolonged (i.e., 8-day) growth arrest. Similar to epidermal trauma in vivo, scrape wounding of HaCaT monolayers resulted in the rapid and location-specific induction of PAI-1 protein (an increase of 11- to 16-fold relative to unwounded cultures) in cells immediately bordering the injury site. PAI-1 expression was evident in keratinocytes that comprised the opposed migrating fronts and remained elevated until wound closure. Down-regulation of PAI-1 synthesis in HaCaT cells transfected with an inducible LacSwitch-based antisense vector system markedly impaired both the rate and the extent of wound closure. All injuries created in antisense PAI-1 monolayers remained unhealed at day 8 postinjury compared to the 3-day complete repair typical of control cultures. Vector-driven modulation of PAI-1 synthesis was also associated with an increase in the percentage of suprabasal-type keratinocytes within the wound field. PAI-1 expression by migrating HaCaT cells appears necessary to maintain the basal epidermal phenotype and/or appropriate cell-to-substrate adhesion during injury repair.

PAI-1 gene expression is growth state-regulated in cultured human epidermal keratinocytes during progression to confluence and postwounding.

Growth of human keratinocytes (NHKs) in submerged cultures approximates a "wound" response generally considered equivalent to regenerative maturation. Within this context, PAI-1 expression in cultured NHKs appears to be growth state-regulated and is associated with specific NHK subpopulations undergoing migration and proliferation in response to wounding; NHKs transit through specific phases during growth to confluence. Basal layer keratinocytes comprise several classes of nucleated epidermal cells (designated "A," "B," and "C") which are distinguishable on the basis of RNA content, population generation time, and expression of basal cell marker proteins. "A" substrate cells initially give rise to expanding proliferating keratinocyte colonies in vitro, but are rapidly replaced (at the stage of 50-75% culture confluence) by transient amplifying "B" cells and, eventually, the larger "C" subpopulation which subsequently differentiates into suprabasal spinous cells during the postconfluent growth period. Expression of plasminogen activator inhibitor type-1 (PAI-1), a serine protease inhibitor and member of the serpin gene family which is synthesized by "activated" or migrating keratinocytes in vivo, was restricted to the preconfluent stages of cell growth in vitro, a condition equivalent to wound regeneration in situ. PAI-1 mRNA and protein expression was maximal in 50-75% confluent cultures correlating, thereby, with the emergence of the transient amplifying "B" cell population. Flow cytometric analysis revealed that PAI-1 is detected early in expanding NHK colonies, during the initial recruitment of basal cells from the "A" state into the "B" compartment in the absence of significant proliferation, suggesting that PAI-1 may be active in regulating early NHK migratory events, independent of cell proliferation. Thereafter, these PAI-1-expressing "A" cells are recruited into the "B" compartment, where they continue to migrate, proliferate, and enlarge, eventually giving rise to PAI-1-expressing "C" cells. By the time NHK cultures reach exponential growth, virtually no small "A" cells contain immunoreactive PAI-1. PAI-1 expression peaks during preconfluent growth and remains confined to larger basal cell phenotypes. Cellular accumulation of both PAI-1 mRNA and protein appeared to be cell cycle phase-specific and characteristic of progression through an activated G1 growth phase. Induced expression, moreover, was restricted to a "window" in G1 with PAI-1 mRNA evident within 2 h after serum addition to growth-arrested cells and attaining maximal levels (50-fold) at 10 h poststimulation. Induced PAI-1 expression, thus, appears to be a general characteristic of the activated epidermal phenotype in vitro as well as in vivo.

Growth of melanocytes in human epidermal cell cultures.

Epidermal cell cultures were grown in keratinocyte-conditioned medium for use as burn wound grafts; the melanocyte composition of the grafts was studied under a variety of conditions. Melanocytes were identified by immunohistochemistry based on a monoclonal antibody (MEL-5) that has previously been shown to react specifically with melanocytes. During the first 7 days of growth in primary culture, the total number of melanocytes in the epidermal cultures decreased to 10% of the number present in normal skin. Beginning on day 2 of culture, bipolar melanocytes were present at a mean cell density of 116 +/- 2/mm2; the keratinocyte to melanocyte ratio was preserved during further primary culture and through three subpassages. Moreover, exposure of cultures to mild UVB irradiation stimulated the melanocytes to proliferate, suggesting that the melanocytes growing in culture maintained their responsiveness to external stimuli. When the sheets of cultured cells were enzymatically detached from the plastic culture flasks before grafting, melanocytes remained in the basal layer of cells as part of the graft applied to the patient.

Sodium-N-butyrate induces cytoskeletal rearrangements and formation of cornified envelopes in cultured adult human keratinocytes.

The technique developed in our laboratory allows us to culture multilayered, stratified sheets of human keratinocytes, which can be used to cover the burn wounds of patients. Organization of cells in these cultures resembles stratum germinativum and stratum spinosum but there are only a few fully keratinized cells and the stratum corneum is not developed. Since the fully differentiated sheets may offer additional advantages as epidermal transplants, attempts were made to enhance the degree of differentiation in vitro. In the present study sodium-N-butyrate (NaB) was used as a differentiating agent and its effect on the cell cycle and cytoarchitecture of epidermal cells was investigated. Incubation of keratinocytes in the presence of 2.5 mM NaB induced the appearance of enucleated cornified envelopes, covering approximately 70-80% of the surface of the cultures. Their appearance correlated with a decrease in expression of keratin K13, previously shown to be inhibited during terminal differentiation of human keratinocytes. An increase in transglutaminase transferase activity was also observed. The induction of cornified layers also correlated with an increase in the amount of microfilament (MF)-associated actin. NaB also induced changes in the cell cycle distribution of the keratinocyte cultures. A decrease in the proportion of S and G1B phase cells was paralleled by an increase in G1A cells, maximally expressed 30-48 h following addition of the inducer. Interestingly, NaB also induced a cell arrest in G2 phase. These cell cycle perturbations preceded the onset of keratinocyte differentiation. The results indicate that the enhanced differentiation of human keratinocytes in the presence of NaB may serve as a means to produce epidermal sheets with improved properties for transplantation in a clinical setting. It also serves as an in vitro model system to study the interrelationships between biochemical events and cell cycle changes accompanying differentiation.

Human keratinocyte culture. Identification and staging of epidermal cell subpopulations.

Stratification of human epidermal cells into multilayered sheets composed of basal and suprabasal layers (resembling the stratum germinativum and stratum spinosum of the epidermis) was studied in a dermal component-free culture system. Although no stratum corneum developed in vitro, this culture system provided a method to study early events in human keratinocyte differentiation. Multiparameter flow cytometric analysis of acridine orange-stained epidermal cells from these cultures revealed three distinct subpopulations differing in cell size, RNA content, and cell cycle kinetics. The first subpopulation was composed of small basal keratinocytes with low RNA content and a long generation time. The second subpopulation consisted of larger keratinocytes, having higher RNA content and a significantly shorter generation time. Finally, the third subpopulation contained the largest cells, which did not divide, and represent the more terminally differentiated keratinocytes. This in vitro approach provides discriminating cytochemical parameters by which the maturity of the epidermal cell sheets can be assessed prior to grafting onto human burn patients.

High rate of sister chromatid exchanges of Bloom's syndrome chromosomes is corrected in rodent human somatic cell hybrids.

The high rate of sister chromatid exchange (SCE) characteristic of cultured somatic cells from patients with Bloom's syndrome (BS) was found to be fully corrected in BS chromosomes retained by somatic cell hybrids between Chinese hamster cells (CHO-YH 21) and BS fibroblasts (GM 1492), independent of the type and the number of human chromosomes retained. On the contrary, the average rate of SCE per Chinese hamster chromosome remained unaffected by hybridization with both BS and normal human cells. A partial suppression of SCE of about 30% was observed in the BS cells themselves when these were co-cultivated with Chinese hamster/Bloom's syndrome hybrid cells. In these hybrids, the rate of SCE per chromosome (Chinese hamster or human) was unaffected by co-cultivation. The data reported indicate that the high rate of SCE in BS cells must be considered to be the consequence of a lost normal function, rather than the acquisition of a new abnormal one, and that several independent genetic systems may be involved in the control of SCE during the replication of mammalian cells. Accordingly, the high rate of SCE in a cultured cell line or an individual should be looked upon as the common phenotype resulting from mutation(s) at any one of these systems. The occurrence of genetic complementation for SCE across the species barrier suggests that at least some of these genetic systems are homologous in different mammalian species and emphasizes the potential(s) of somatic cell hybridization for studying the biology of SCE, in general, and the genetics of Bloom's syndrome, in particular.