Characterization and functionality of cell surface molecules on human mesenchymal stem cells

We have characterized adhesion molecules on the surface of multipotential human mesenchymal stem cells (hMSCs) and identified molecules whose ligands are present on mature hematopoietic cells. Flow cytometric analysis of hMSCs identified the expression of integrins: alpha1, alpha2, alpha3, alpha5, alpha6, alphav, beta1, beta3, and beta4, in addition to ICAM-1, ICAM-2, VCAM-1, CD72, and LFA-3. Exposure of hMSCs to IL-1alpha, TNFalpha or IFNgamma up-modulated ICAM-1 surface expression, whereas only IFNgamma increased both HLA-class I and -class II molecules on the cell surface. Whole cell-binding assays between the hMSCs and hematopoietic cell lines showed that T lymphocytic lines bound hMSCs with higher affinity than lines of either B lymphocytes or those of myeloid lineage. Experiments using autologous T lymphocytes isolated from peripheral blood mononuclear cells showed that hMSCs exhibited increased affinity for activated T-lymphocytes compared to resting T cells by quantitative whole cell binding and rosetting assays. Flow cytometric analysis of rosetted cells demonstrated that both CD4+ and CD8+ cells bound to hMSCs. To determine the functional significance of these findings, we tested the ability of hMSCs to present antigen to T lymphocytes. hMSCs pulsed with tetanus toxoid stimulated proliferation and cytokine production (IL-4, IL-10, and IFNgamma) in a tetanus-toxoid-specific T cell line. Maximal cytokine production correlated with maximal antigen-dependent proliferation. These data demonstrate physiological outcome as a consequence of interactions between hMSCs and human hematopoietic lineage cells, suggesting a role for hMSCs in vivo to influence both hematopoietic and immune function(s).     

Mediated exodus of L-dopa from human epidermal Langerhans cells

Summary. L-3,4-dihydroxyphenylalanine (L-dopa) is not metabolized within human epidermal Langerhans cells (LC); yet they can take up substantial amounts of this amino acid which subsequently can be released into the extracellular space. We recently reported that human epidermal energy metabolism is predominantly anaerobic and that the influx mechanism is a unidirectional L-dopa/proton counter-transport system and now we describe conditions for the mediated transport of L-dopa out of the LC. It is demonstrated that certain amino acids and one dipeptide can effectively trigger the efflux of L-dopa taken up by the LC.Thus, -methyl-dopa (-m-dopa), D-dopa and the dipeptide, met–ala at the outside of the plasma membrane stimulated the efflux of L-dopa from L-dopa loaded LC. Similar effects were achieved by a variety of other amino acids in the extracellular fluid while some other amino acids were inactive. The time required for 50% D-methionine-induced exodus of L-dopa from L-dopa loaded LC was in the range of 5–7min and a complete exodus of L-dopa was attained at about 20min of incubation. This dislocation of L-dopa to the extracellular fluid is interpreted as an expression of trans-stimulation. In the case of -m-dopa, D-dopa and met–ala, which admittedly were not able to penetrate the plasma membrane of LC, the concept of trans-stimulation was given a new purport, since none of them were able to participate in an exchange reaction. Finally, it could be concluded that L-dopa escaped by a route different from the one responsible for L-dopa uptake in LC.Thus, while the influx of L-dopa supports extrusion of protons deriving from anaerobic glycolysis in the LC, L-dopa efflux can provide the cells with useful amino acids in an energy-saving way, altogether a remarkable biological process. From this follows that L-dopa has a biological function of its own, besides being a precursor in the catecholamine and pigment syntheses.     

Differential modulation of human epidermal Langerhans cell maturation by ultraviolet B radiation.

UVB irradiation of the skin causes immunosuppression and Ag-specific tolerance in which Langerhans cells (LC) are involved. We tested the effect of UVB on LC that had migrated out of cultured epidermal sheets derived from the skin that was irradiated ex vivo (200, 400, 800, or 1600 J/m2). Two separate subpopulations of LC were distinguished: large-sized LC with high HLA-DR expression, and HLA-DR-low, small LC. UVB stimulated the maturation of the former LC subset as demonstrated by enhanced up-regulation of CD80, CD86, CD54, CD40, and CD83 and reduced CD1a expression in comparison with unirradiated controls. In contrast, the latter LC exhibited little or no up-regulation of these molecules except for high CD1a expression and high binding of annexin V, indicating that they were apoptotic, although their CD95 expression was relatively low. Stimulation of enriched LC with CD40 ligand-transfected cells and IFN-gamma revealed that the release of IL-1beta, IL-6, IL-8, and TNF-alpha was enhanced by UVB. In comparison with HLA-DR-low LC, HLA-DR-high LC were the principal IL-8 producers as demonstrated by intracellular cytokine staining, and they retained more accessory function. There was no detectable secretion of IL-12 p70, and IL-18 production was neither affected by any stimulus nor by UVB. These results suggest a dual action of UVB on LC when irradiated in situ: 1) immunosuppression by preventing maturation and inducing apoptotic cell death in part of LC, and 2) immunopotentiation by enhancing the up-regulation of costimulatory molecules and the production of proinflammatory cytokines in another part.     

Isolation of subpopulations of murine epidermal cells using monoclonal antibodies against differentiation-related cell surface molecules

Abstract. Monoclonal antibodies (mAbs) against epithelial cells were prepared by immunization of rats with lyophilized murine epithelia. Screening against tissue sections and epithelial cell suspensions permitted identification of mAbs against surface molecules that are expressed early in cell differentiation. Staining with These mAbs followed by fluorescence-activated cell sorting enabled isolation of subpopulations of basal epithelial cells. Staining these subpopulations with antibodies against known differentiation markers (cytokeratins and bullous pemphigoid antigen) and measurements of cell size indicated that they represented fractions of the basal cell population in sequential stages of early differentiation. Labeling mice with bromodeoxyuridine at various limes prior to cell isolation showed that the least-differentiated basal cells cycle more slowly than those at later stages, data which support the concept of a differentiation-related, hierarchical pattern of organization of the proliferative compartment.     

Glycosylation of human leukocyte locus A molecules is dependent on the cell type

Peripheral blood monocytes and B cells were isolated from a normal donor, and a portion of the B cells was transformed by the Epstein-Barr virus (EBV). Human leukocyte locus A (HLA) class-I and class-II molecules were immunoprecipitated by specific monoclonal antibodies after cell labeling with [3H]mannose. Glycopeptides of HLA molecules were obtained by pronase digestion and were analysed by lectin-affinity chromatography. Complex structures were hydrazinolysed, and their sialic acid content was analysed by ion-exchange chromatography, whereas the high-mannose structures were separated by HPLC. In normal cells, class-I antigens bear principally fucosylated biantennary structures while HLA-DR class-II antigens bear bi-, tri- and tetra-antennary structures and high-mannose structures. HLA antigens are more sialylated on normal B cells than on normal monocytes. An EBV cell line had a very different pattern of HLA-antigen glycosylation when compared with the original B cells. In the transformed cells, the fractions containing biantennary structures are largely decreased. In contrast, an increase of the tri- and tetra-antennary structure fractions is noticed, particularly in class-II molecules, while both triantennary and high-mannose structures are increased in class-I molecules. Moreover, when compared to normal B cells, the complex structures of class-I antigens in the EBV-transformed B-cell line are undersialylated while they are oversialylated in the case of the class-II molecules.     

Ontogeny of Langerhans cells in human embryonic and fetal skin: cell densities and phenotypic expression relative to epidermal growth

Langerhans cells (LCs) positive for HLA-DR antigens were present in developing human epidermis by at least 7 weeks estimated gestational age (EGA). Most were negative for CD1 (T6) until 12-13 weeks EGA when they underwent a dramatic increase in CD1 reactivity. To gain insight into the density of LCs during ontogeny and to assess whether their distribution was coordinated with epidermal growth, the number of cells positive for both HLA-DR and CD1 antigens was determined relative to surface area and to volume of developing, interfollicular epidermis. LCs differed in their phenotype, distribution (follicular vs. interfollicular), size, and shape between 7 and 21 weeks EGA; however, during this period they maintained a statistically equivalent (P greater than .25) density (65 cells/mm2 and 1,750/mm3) even though the epidermis increased in thickness and the fetus rapidly expanded its surface area. While LCs were evenly distributed within the epidermal sheets at all gestational ages, those in embryonic skin were much smaller and less dendritic than the older cells. The density, size, and shape of LCs in developing skin seemed to be independent of epidermal status (e.g., thickness of keratinization, and number of cell layers) but rather were correlated with gestational age. The number of fetal LCs, through at least 23 weeks EGA, was only 10-20% of the adult LC density. Thus, we can conclude that the increase in LC density to adult levels must occur either during the third trimester or after birth.     

Labelling of murine epidermal Langerhans cells with H3-thymidine.

Epidermal Langerhans cells may be identified by light microscopy by their strongly positive reaction following incubation for ATPase activity. Intact sheets of epidermis from mice killed at various time intervals following a single pulse label of H3-thymidine were incubated to demonstrate ATPase activity and subsequently processed for autoradiography. In specimens taken one hour after labelling, many basal keratinocytes were labelled but very few ATPase-positive dendritic cells. At subsequent time periods a few pairs of labelled ATPase-positive cells were found but individually labelled cells were not observed. The findings suggest that epidermal Langerhans cells form a very stable (labelling index less than 0.01%) self-replicating population which divides to maintain cell spacing during growth. No evidence was found for migration and interchange of Langerhans cells with the connective tissue, or for an origin of Langerhans cells by transformation of another cell type.     

Lactoferrin: influences on Langerhans cells, epidermal cytokines, and cutaneous inflammation.

It has been suggested previously that, in addition to other biological roles, lactoferrin (LF) may display antiinflammatory properties secondary to the regulation of cytokine expression. To explore this concept further, we have here examined in human volunteers the influence of recombinant homologous LF on the migration of epidermal Langerhans cells (LC), a process that is known to be dependent upon the local availability of certain proinflammatory cytokines including tumor necrosis factor alpha (TNF-alpha) and interleukin 1beta (IL-1beta). In common with previous studies in mice, it was found that topical administration of LF prior to exposure at the same site to the contact sensitizer diphenylcyclopropenone resulted in a significant reduction of allergen-induced LC migration from the epidermis (measured as a function of the frequency of CD1a+ or HLA-DR+ LC found in epidermal sheets prepared from punch biopsies of the treated skin sites). However, under the same conditions of exposure, LF was unable to influence migration of LC induced by the intradermal administration of TNF-alpha data consistent with the hypothesis that one action of LF in the skin is to regulate the local production of this cytokine. Further support for this hypothesis was derived from experiments conducted with IL-1beta. This cytokine is also able to induce the mobilization of LC following intradermal injection, although in this case, migration is known to be dependent upon the de novo production of TNF-alpha. We observed that prior exposure to LF resulted in a substantial inhibition of IL-1beta-induced LC migration, data again consistent with the regulation of TNF-alpha production by LF. Collectively, these results support the view that LF is able to influence cutaneous immune and inflammatory processes secondary to regulation of the production of TNF-alpha and possibly other cytokines.     

Immunological and histochemical analysis of regional variations of epidermal Langerhans cells in normal human skin

Summary Epidermal Langerhans' cells (LC) were enumerated in normal human skin from various anatomical sites using a monoclonal antibody (NA1/34) to human thymocyte antigen (HTA-1) and the standard ATPase reaction on frozen sections. The same population of cells was identified with each technique. LC densities were found to be significantly higher in hair bearing skin than in skin from the palm and sole. LC were also identified in hair follicles (where the numbers decreased from the superficial to the deep portions) and sebaceous glands but in no other adnexal structure. Normal numbers were encountered in patients who had received radiotherapy or systemic chemotherapy for malignant disease for periods of greater than two months before death. As LC are important antigen presenting cells, the variation in their density suggests that the immunological properties of normal skin may not be uniform throughout the body. This may be related to the varying anatomical distribution of some skin disorders with an immunological basis.     

Multiple keratins of cultured human epidermal cells are translated from different mRNA molecules.

The keratins of human epidermis consist of several distinct proteins of different molecular weight that can be separated by gel electrophoresis in the presence of sodium dodecylsulfate. These proteins are very similar in structure, as determined by amino acid composition, polypeptide mapping and immunological reactivity. At least five such keratins are found in cultured human epidermal cells. We have examined the mode of synthesis of these keratins by isolating the poly(A)⁺ mRNA from the cultured cells and translating it in a reticulocyte system. All the keratins characteristic of the cultured cells were synthesized in vitro from the mRNA; they were identified by their molecular weight and by polypeptide mapping. No evidence was found for any precursor of substantially greater molecular weight. A study of the kinetics of synthesis showed that all the keratins were labeled within 2 min after the addition of ³⁵S-methionine to a translation system preincubated with epidermal mRNA, and the relative intensities of labeling did not change upon further incubation. It was therefore improbable that one keratin could be the precursor of another. The mRNAs of the large keratins could be completely separated from those of the small keratins by gel electrophoresis under either native or denaturing conditions. Within the group of small mRNAs, each had a different mobility although resolution was incomplete. Upon translation, the mRNA fractions yielded different groups of keratins corresponding in molecular weight to their counterparts in the cells. Consequently, most if not all keratins of different size are translated from different messages. The approximate sizes of the mRNA molecules for different keratins were determined from their mobility under denaturing conditions. The size of the mRNA was not always proportional to the size of the encoded keratin, demonstrating the existence of noncoding segments of different length in the different mRNA molecules.     

Influence of cell density and receptor number on the binding and distribution of cell surface epidermal growth factor receptors

Summary Previous studies have shown that cell density influences the expression of receptors for at least four growth factors. The data presented in this report demonstrate that epidermal growth factor receptors are regulated differently on cells expressing over a million receptors as opposed to cells expressing approximately fivefold fewer receptors. Specifically, we show that BT-20, MDA-468, and A-431-R1 cells, which exhibit a large number of epidermal growth factor receptors, preferentially down-regulate the high affinity class of these receptors as cell density increases. In addition, we show that these cells express cell surface epidermal growth factor receptors that are localized predominately to the periphery of the cells. In contrast, A-549 and BSC-1 cells, which exhibit fewer cell surface epidermal growth factor receptors and which reduce all affinity classes of epidermal growth factor receptors as cell density increases, exhibit a diffuse cell surface distribution of these receptors at both low and high densities.     

Inhibition of the L-dopa transport system in human epidermal Langerhans cells by omeprazole and its analogues

Summary. L-3,4-dihydroxyphenylalanine (L-dopa) transport into human Langerhans cells (LC) occurs by a saturable mediation. This plasma membrane agency is, due to its characteristics, distinguishable from systems transporting other neutral, cationic and anionic amino acids into other cells and serves to catalyze the flow of L-dopa, only, into LC. The uphill operation of this L-dopa transport system is believed to occur by down-gradient countermigration of H⁺. Due to the uniqueness of the L-dopa transport system, the widely used analogue inhibition approach was not applicable. Instead we studied omeprazole and its analogues in our search for suitable inhibitory candidates. Omeprazole and most of its analogues were indeed inhibitory in the concentration range 1–100mol/L. Conspicuously, the compounds with strongest polarity were least inhibitory. The inhibitory pattern displayed by omeprazole and the other analogues on L-dopa uptake in LC corresponded to some extent to what has been observed previously for purified H⁺,K⁺-ATPase from tubulovesicles of the stomach. No effects of the inhibitors were registered on energy charge and lactate production of epidermal biopsies, nor were any gross alterations of ultrastructure of LC noticed.     

Effect of age on surface molecules and cytokine expression in human dendritic cells

Dendritic cells (DCs) are central in regulating both innate and acquired immunity, but their possible age-related functional modifications are still unclear. Here we have analyzed the effect of age on LPS-treated monocyte-derived DCs (MDDCs). A negative correlation between age and cell expression of ICAM-1, CD25 and IL-10 was observed in a group of healthy donors. This has been confirmed by a significantly reduced expression of the same molecules in cells of subgrouped elderly versus younger individuals. On the contrary, a positive correlation between age and cell expression of IL-6 and IL-18 has been reported in all the subjects and further supported by a significant increase of the two pro-inflammatory cytokines in cells of elderly versus young subjects.Our data indicate that aging can impair the expression of ICAM-1 and CD25 and skew the production of cytokines, including IL-18, which concur to make a pro-inflammatory milieu. Altogether, the present results point to additional molecules whose role might be relevant in immunosenescence of human DCs, confirming that these cells undergo functional changes during aging.     

Maturational steps of bone marrow-derived dendritic murine epidermal cells. Phenotypic and functional studies on Langerhans cells and Thy-1+ dendritic epidermal cells in the perinatal period

The adult murine epidermis harbors two separate CD45+ bone marrow (BM)-derived dendritic cell systems, i.e., Ia+, ADPase+, Thy-1-, CD3- Langerhans cells (LC) and Ia-, ADPase-, Thy-1+, CD3+ dendritic epidermal T cells (DETC). To clarify whether the maturation of these cells from their ill-defined precursors is already accomplished before their entry into the epidermis or, alternatively, whether a specific epidermal milieu is required for the expression of their antigenic determinants, we studied the ontogeny of CD45+ epidermal cells (EC). In the fetal life, there exists a considerable number of CD45+, Ia-, ADPase+ dendritic epidermal cells. When cultured, these cells become Ia+ and, in parallel, acquire the potential of stimulating allogeneic T cell proliferation. These results imply that CD45+, Ia-, ADPase+ fetal dendritic epidermal cells are immature LC precursors and suggest that the epidermis plays a decisive role in LC maturation. The day 17 fetal epidermis also contains a small population of CD45+, Thy-1+, ADPase-, CD3- round cells. Over the course of 2 to 3 wk, they are slowly replaced by an ever increasing number of round and, finally, dendritic CD45+, Thy-1+, CD3+ EC. Thus, CD45+, Thy-1+, ADPase-, CD3- fetal EC may either be DETC precursors or, alternatively, may represent a distinctive cell system of unknown maturation potential. According to this latter theory, these cells would be eventually outnumbered by newly immigrating CD45+, Thy-1+, CD3+ T cells--the actual DETC.     

The distribution of cell surface proteins on spreading cells. Comparison of theory with experiment.

Bretscher (1983) has shown that on uniformly spread giant HeLa cells, the receptors for low density lipoprotein (LDL) and transferrin are concentrated toward the periphery of the cells. To explain these nonuniform distributions, he proposed that on giant HeLa cells, recycling receptors return to the cell surface at the cell's leading edge. Since the distribution of coated pits on these cells is uniform, Bretscher and Thomson (1983) proposed that there is a bulk membrane flow toward the cell centers. Here we present a mathematical model that allows us to predict the distribution of cell surface proteins on a thin circular cell, when exocytosis occurs at the cell periphery and endocytosis occurs uniformly over the cell surface. We show that on such a cell, a bulk membrane flow will be generated, whose average velocity is zero at the cell center and increases linearly with the distance from the cell center. Our model predicts that proteins that aggregate in coated pits will have concentrations that are maximal at the cell periphery. We fit our theory to the data of Bretscher and Thomson (1983) on the distribution of ferritin receptors for the following cases: the receptors move by diffusion alone; they move by bulk membrane flow alone; they move by a combination of diffusion and bulk membrane flow. From our fits we show that tau m greater than 3.5 tau p, where tau m and tau p are the lifetimes of the membrane and the ferritin receptor on the cell surface, and that tau pD less than 6.9 X 10(-7) cm2, where D is the ferritin receptor diffusion coefficient. Surprisingly, we obtain the best fits to the data when we neglect membrane flow. Our model predicts that for proteins that are excluded from coated pits, the protein concentration will be Gaussian, being maximal at the cell center and decreasing with the distance from the cell center. If on giant HeLa cells a protein with such a distribution could be found, it would strongly support Bretcher's proposal that there is an inward membrane flow.