|
|||||||||
|
|
|
|
|
|
|
|
|
|
|
IMMUNOBIOLOGY
From the Vienna International Research Cooperation
Center/Novartis Research Institute and the Departments of Dermatology I
and Internal Medicine III, Institute of Immunology, University of
Vienna, Vienna, Austria.
Epithelial tissues of various organs contain immature Langerhans
cell (LC)-type dendritic cells, which play key roles in immunity. LCs
reside for long time periods at an immature stage in epithelia before
migrating to T-cell-rich areas of regional lymph nodes to become
mature interdigitating dendritic cells (DCs). LCs express the
epithelial adhesion molecule E-cadherin and undergo homophilic E-cadherin adhesion with surrounding epithelial cells. Using a defined
serum-free differentiation model of human CD34+
hematopoietic progenitor cells, it was demonstrated that LCs generated
in vitro in the presence of transforming growth factor Dendritic cells (DCs) represent a developmentally
heterogeneous class of leukocytes that are highly specialized in
antigen uptake, processing, and presentation. DCs differ in migration pathways, tissue location, and functional abilities. Mature DCs in
T-cell areas of lymph nodes are known as interdigitating DCs and are in
part derived from a peripheral DC migration pathway that involves
immature DCs in peripheral organs such as epithelial Langerhans cells
(LCs).1-3 After homing to T-cell areas of secondary lymphoid organs, DCs rapidly undergo apoptosis unless they receive a
survival signal from antigen-specific T cells.4
Immature epidermal LCs fulfill a sentinel role by filtering the
surrounding tissue for foreign antigens and pathogens. They form a
3-dimensional network in suprabasal epidermal layers and spend long
time periods at an immature or a nonactivated differentiation stage in
the epidermis before they migrate to the lymph nodes.5 LCs
express high levels of the homophilic adhesion molecule E-cadherin and
undergo E-cadherin-dependent adhesion with epidermal
keratinocytes.6 E-cadherin expression is markedly
down-regulated upon the migration and maturation of epidermal LCs, and
lower expression levels of E-cadherin on the surfaces of cultured LCs
correlate with decreased cell adhesiveness.6-8 LC
migration can be induced in vivo by the topical application of
allergens or the intradermal or systemic injection of tumor necrosis
factor- LC migration from the epidermis to the draining lymph nodes is preceded
by the activation and maturation of immature LCs locally in the
epidermis. Because immature epidermal LCs spontaneously undergo
maturation on in vitro culture,14 inhibitory signals that
are provided by the epidermal microenvironment might likely counteract
the maturation of LCs in vivo.
We recently established defined serum-free culture conditions for the
in vitro generation of immature LCs from human CD34+ cord
blood progenitor cells. We demonstrated that LC differentiation from
CD34+ cells is dependent on the addition of transforming
growth factor The availability of a differentiation model that allows the maintenance
of LCs in vitro in defined serum-free cultures at an immature stage
provided the opportunity to study signaling mechanisms leading to the
maturation of immature LCs. Our data suggest that functional activation
of the TGF- Antibodies
Immunofluorescence staining procedures
For suspension staining of intracellular antigens, we used the commercially available reagent combination Fix&Perm (An der Grub, Kaumberg, Austria) according to the manufacturer's procedure. Briefly, cells were first fixed for 15 minutes at room temperature (50 µL cells plus 100 µL formaldehyde-based fixation medium). After one washing with phosphate-buffered saline, pH 7.2, cells were resuspended in 50 µL phosphate-buffered saline and mixed with 100 µL permeabilization medium plus 20 µL fluorochrome-labeled antibody. After further incubation for 15 minutes at room temperature, cells were washed again and analyzed by flow cytometry. Indirect suspension staining for the intracellular Birbeck granule marker molecule Lag was performed as described previously.15 Flow cytometry and cell sorting Flow cytometric analyses were performed using a FACScan flow cytometer (Becton Dickinson Immunocytometry Systems) equipped with a single laser emitting at 488 nm. Cell sorting was performed using a FACS Vantage flow cytometer (Becton Dickinson). The purity of the CD1a+ cell fractions obtained by sorting was determined by re-analysis and was found to exceed 95%.Cord blood cells Cord blood (CB) samples were collected during healthy full-term deliveries. Mononuclear cells were isolated within 10 hours of collection by discontinuous Ficoll/Hypaque (Pharmacia, Uppsala, Sweden) density gradient centrifugation. CD34+ cells were isolated from CB mononuclear cells using the MACS CD34 Progenitor Cell Isolation Kit (Miltenyi Biotec, Bergisch Gladbach, Germany) as described previously,15 according to the instructions of the manufacturer. The purity of the CD34+ population exceeded 90%.Primary cultures of CD34+ CB cells Primary (1°) cultures of purified CD34+ CB cells were grown in 24-well plates (Costar, Cambridge, MA) (5 × 103 cells in 1 mL/well) at 37°C in a humidified atmosphere and in the presence of 5% CO2, as previously described by us.16 The serum-free medium X-VIVO 15 (BioWhittaker, Walkersville, MD) contained L-glutamine (2.5 mmol/L), penicillin (125 U/mL), and streptomycin (125 µg/mL). Cultures were supplemented with optimized concentrations of the following human cytokines: FL (100 ng/mL; kindly provided by Immunex, Seattle, WA), TGF- 1 (0.5 ng/mL; purified from platelets; British
Biotechnology, Abington, UK), rhTNF (50 U/mL; Bender, Vienna, Austria), rhGM-CSF (100 ng/mL; Novartis, Basel, Switzerland), and rhSCF (20 ng/mL; Amgen, Thousand Oaks, CA). A 50% medium exchange was regularly performed at culture day 7.
Secondary cultures of in vitro-generated immature LC Secondary (2°) culture experiments were set up using cells generated in 1° LC generation cultures as described above. Cells from 1° cultures were harvested at days 10 to 14. Single-cell suspensions were prepared by pipetting, and resuspended cells were plated in the above-described growth medium X-VIVO 15 in 24-well plates at a cell density of 1 × 105 cells per well (1 mL). Cultures were supplemented with GM-CSF, TNF, or TGF-1, or any combination of
them, at concentrations described above. CD40 ligand trimer (CD40L; 200 ng/mL; kindly provided by Immunex) was added when indicated.
Aggregation cultures of LC in the presence of antibodies Cells from 1° cultures (see above) were harvested, carefully resuspended by pipetting, and submitted to 2° cultures (ie, aggregation cultures) of 5 × 104 cells/well in flat-bottomed 96-well plates (Costar) under exactly the same growth conditions as for 1° cultures. These reaggregation cultures were supplemented with 20 µg/mL mAb when indicated, preincubated with mAb for 20 minutes at 4°C, and transferred to 37°C. After additional culture for 24 hours, cell cluster formation was analyzed by phase-contrast microscopy using a scoring system previously established in our laboratory18 with slight modifications. Scores ranged from 0 to 4, as follows: 0, less than 10% of the cells were in aggregates; 1, 10% to 50% of the cells were in aggregates; 2, approximately 50% to 75% of the cells were in aggregates; 3, 75% to 90% of the cells were in aggregates; 4, 90% to 100% of the cells were in aggregates. In the first series of experiments, these cultures were initiated with total cells generated in primary cultures at day 8. To further study direct homotypic LC-to-LC interaction, we initiated identical cultures (see above) with flow sorted CD1a+ cells generated after 10 to 12 days in 1° cultures. To assess the differentiation stage of LC, cells from these reaggregation cultures were stained for CD1a or were double-stained for CD1a versus CD86 or CD83 expression and analyzed by flow cytometry as indicated in "Results."Mixed leukocyte reaction Graded numbers of irradiated (30 Gy; cesium Cs 137 source) stimulator cells (generated LC) were added to constant numbers (5 × 104/well) of purified (greater than 98%) allogeneic T cells in round-bottom 96-well tissue culture plates (Costar). Triplicate analyses were performed. Stimulation of responding T cells was monitored by measuring 3H-thymidine (Amersham Life Science, Buckingham, UK) incorporation on day 5 of culture. Incorporated radioactivity was measured using a Top-Count microscintillation counter (Packard Instrument, Meriden, CT). Allogeneic T cells used in these experiments were purified from peripheral blood mononuclear cells by negative immunomagnetic depletion (MACS beads; Miltenyi Biotec) using mAbs specific for CD14 (clone VIM13; generated in our laboratory), CD16 (clone 3G8; Caltag), CD19 (clone HD37; kindly provided by Dr G. Moldenhauer, Heidelberg, Germany), HLA-DR (clone L243; ATCC, Rockville, MD), and CD33 (4D3; generated in our laboratory) as previously described by us.16
TGF- 1 to serum-free
cultures of CD34+ cells, supplemented with the cytokines
GM-CSF plus TNF, SCF, and FL, results in the generation, after a
culture period of 7 to 10 days, of a large cell fraction that strongly
expresses CD1a (approximately 60%).16 Indistinguishable
from epidermal LC in vivo, most CD1a+ cells generated in
these cultures are predominantly CD86 (B7.2)dim/ and lack
expression of the mature DC marker molecule CD83.16 This
immature DC phenotype of most of the generated cells remains preserved
on extension of the culture period to 14 to 16 days (Figure
1A,B). Furthermore, most
CD1a+ cells generated in these TGF-1-supplemented
cultures express high levels of the homophilic epithelial adhesion
molecule E-cadherin (Figure 1A, upper right diagram), and a large
proportion of cells expresses the Birbeck granule-associated molecule
Lag (Figure 2, upper panel; mean, 52% at
day 14 to 16; n = 7). Cells generated in identical culture medium in
the absence of TGF-1 lack expression of Lag (Figure 2, lower panel),
and most of these cells remains CD1a/E-cadherin (Figure 1A).
The most striking morphologic feature of TGF- LCs acquire mature DC features when resuspended and plated in defined serum-free secondary (2°) cultures LCs generated in the above-described primary cultures of CD34+ cells for 10 to 14 days were harvested, resuspended to obtain single-cell suspensions, and plated in 2° serum-free cultures. Stimulation of LC in these 2° cultures in the presence of GM-CSF plus TNF for 48 hours resulted in the acquisition of mature
DC features by most generated CD1a+ cells. As shown in
Figure 3, panel A (top diagrams), most
immature CD1abright cells from primary (1°) cultures
acquired increased expression of CD86 and were found to express the
mature DC marker molecule CD83 on stimulation with GM-CSF plus TNF.
These phenotypic changes were correlated with a marked reduction of
mean CD1a expression density of cultured cells (Figure 3A). Morphologic
examination of GM-CSF plus TNF-supplemented 2° cultures revealed
numerous loosely plastic adherent and floating single cells with highly dendritic processes (Figure 4). GM-CSF
plus CD40L-supplemented parallel cultures showed a similar maturation
pattern, though these cells were found to be consistently brighter
CD86+ and CD83+ than those cultured in GM-CSF
plus TNF, and they showed a more mature DC morphology (Figures 3A,
4). Interestingly, even in the presence of GM-CSF alone (Figure 3A,
third panel), 2° cultures contained higher percentages of
CD86+ cells than 1° control cultures in which cell
clusters remained undisturbed (Figure 3A, bottom). Parallel staining of
cultured cells for additional marker molecules revealed a marked
up-regulation of cell surface expression of HLA-DR, CD80, and CD54
molecules in parallel with the above-described DC maturation-linked
phenotypic changes (CD83 induction, CD86 up-regulation, and decreased
CD1a expression). Conversely, most CD1a+ cells expressed
CD11c, and CD11c expression levels did not clearly change in 2°
cultures (data not shown).
Figure 3, panel B shows a phenotypic analysis of cells stimulated in
the presence of fresh medium supplemented with the initial cytokine
combination (TGF- These experiments demonstrate that cytokines present in 1° cultures
(ie, GM-CSF plus TNF TGF- 1-supplemented cultures remain
immature (Figure 3A, bottom). Because TGF-1 is mainly described as a
suppressive cytokine that inhibits immune responses, we next asked
whether TGF-1 might suppress the maturation of LC. To study this
possibility, we added TGF-1 to the 2° cultures supplemented with
either GM-CSF plus TNF or GM-CSF plus CD40L and phenotypically characterized cultured cells. We observed similar CD86 and CD83 expression patterns by cultured cells in the presence or absence of
TGF-1 (Figure 5), demonstrating that
TGF-1 does not inhibit LC maturation in 2° cultures.
Replating immature LC in 2° cultures at a high cell density induces rapid LC re-aggregation The experiments described above revealed that the observed LC maturation arrest in 1° cultures cannot be explained by a lack of maturation-inducing cytokines or by a direct maturation suppressive effect of TGF-1. Therefore, we investigated an alternative
possibility and analyzed whether LC clustering may provide a
suppressive signal that prevents the maturation of LC in this
differentiation model.
We observed that the mechanical disruption of LC clusters and replating
of generated cells to short-term, serum-free 2° cultures at a high
cell density (5 × 104 cells in flat-bottom, 96-well
plates; see "Materials and methods"), supplemented with identical
cytokines as described for 1° cultures (see above), resulted in
secondary cell cluster formation. This permitted us to analyze LC
clustering in more detail. As can be seen from Figure
6, single-cell suspension of LCs (Figure
6B) prepared by the enforced pipetting of clustered cells from 1° cultures gave rise within 24 hours of additional culture in identical growth medium (cytokines GM-CSF plus TNF
Involvement of cytoadhesion molecules in LC cluster formation in 2° cultures Double staining revealed that most generated CD1a+ cells in primary cultures express the homophilic adhesion molecule E-cadherin (Figure 1). We analyzed cells generated under the various culture conditions for E-cadherin expression. As can be seen in Table 1, expression of E-cadherin slightly decreases on subculturing of day 10 generated cells in 2° cultures supplemented with GM-CSF, GM-CSF plus TNF, or GM-CSF plus CD40L. Conversely, expression intensity of
E-cadherin even increases in parallel 2° cultures containing the
initial TGF-1-supplemented cytokine combination.
We further analyzed cells generated in 1° cultures for the expression
of additional cytoadhesion molecules. As can be seen in Table
2, substantial proportions of generated
CD1a+ LC express the
The short-term re-aggregation characteristics of generated LC at a
high-cell density allowed us to study whether the addition of blocking
mAbs to candidate cytoadhesion molecules might influence secondary LC
cluster formation. For quantification of secondary LC cluster formation
in these 2° cultures, we applied the scoring system (see "Materials
and methods"). Additionally, we analyzed cells after stimulation in
these 2° cultures for expression of CD1a (immature LC,
CD1abright; mature LC, marked loss of CD1a expression
density; see Figure 3). As can be seen from Figure 6, panel C and
Figure 7, panel A, the addition of an mAb
to E-cadherin in these 2° cultures led to the inhibition of cell
cluster formation (score 1), and this effect was associated with
increased CD1a expression by stimulated cells compared to control
cultures (Figure 7B). Monoclonal antibodies to CD11a or CD31 also
strongly inhibited secondary cell cluster formation (score 0 or 1, respectively; Figure 7A), but this effect was associated with a
decrease of mean CD1a expression levels (Figure 7B) of cultured cells
in 2° cultures. Conversely, mAb ligation of CD43 led to the
enhancement of cell clustering (Figures 6D, 7A; score 4 versus 3), and
mean CD1a expression by cultured cells decreased only slightly (Figure
7B).
Ligation of E-cadherin on immature LC inhibits acquisition of mature DC phenotype CD1a+ LC generated in 1° cultures make up approximately 60% of all cultured cells. To eliminate a possible influence of contaminating CD1a cells on cell clustering
and maturation induction of LC, we performed cell sorting experiments
of generated immature CD1a+ LC from 1° cultures (see
"Materials and methods"). In these experiments (Figure
8), we purified CD1a+ LC from
day 10 to 12 cultures by flow sorting and replated them in secondary
cultures in the presence of mAbs as described above. As can be seen
from Figure 8, most purified immature CD1a+ LC acquire,
after further stimulation in control cultures, high expression levels
of CD86 and decreased CD1a expression densities indicative of DC
maturation. This maturation pattern of LC is associated with secondary
cell cluster formation (cluster score 3; data not shown). Parallel
cultures supplemented with mAb HECD-1 specific for E-cadherin contained
substantially higher percentages of immature CD86 to
dim/CD1abright LCs compared with control cultures
(64% vs 37% in the representative experiment shown in Figure 8) and
showed only minimal cell clustering (score 1). These observations were
confirmed in 3 additional experiments using flow sorted
CD1a+ LCs (mean ± SE, 66% ± 5% vs 38% ± 2%,
P = .008; n = 4). In comparison, control cultures
supplemented with a mAb to LFA-1 (CD11a) contained reduced percentages
of immature LC (29% vs 37%, Figure 8) and, in turn, higher
percentages of phenotypically mature
CD1adim/CD86bright DCs. As observed for
anti-E-cadherin mAb, mAb to LFA-1 inhibited secondary cluster
formation of purified LCs (score 1). Parallel analyses for CD1a versus
CD83 expression in 2 experiments revealed virtually identical results
(CD83dim//CD1abright LCs, 60% vs 32% and
53% vs 36%, respectively;
CD86dim//CD1abright LCs, 61% vs 32% and
53% vs 36%, respectively). We further analyzed the capacity of
generated LC to induce allogeneic T-cell proliferation and found that
anti-E-cadherin mAb-pretreated cells appear to be less potent inducers
of allogeneic T-cell proliferation at low stimulator cell numbers than
cells from control cultures supplemented with a nonbinding control mAb
or cells from cultures without antibody supplementation (Figure
9).
Our data suggest an active involvement of E-cadherin in the regulation of LC maturation. We demonstrate here that immature LCs generated in cultures of CD34+ cells undergo profound E-cadherin-dependent homotypic cell clustering similar to what has been described recently for in vitro-generated murine LC.26 We also observed that mechanical disruption of these E-cadherin-dependent LC clusters rapidly induces the acquisition of mature DC features (CD86bright, CD83+) by immature LCs. Furthermore, mAb ligation of E-cadherin on the surface of immature LCs after mechanical cluster disaggregation inhibits the maturation of LC. This effect was found to be specific for E-cadherin engagement because mAbs to other adhesion molecule systems that were found to be similarly involved in LC clustering (CD31, LFA-1) failed to inhibit DC maturation when added to parallel cultures. This suggests that anti-E-cadherin mAb induces functional activation of E-cadherin, resulting in the inhibition of DC maturation in this differentiation model. This effect of anti-E-cadherin mAb mimics E-cadherin-dependent homotypic LC clustering because we observed that LCs stay immature if cell clusters remain undisturbed. Thus anti-E-cadherin mAb binding to E-cadherin seems to functionally mimic E-cadherin binding by its homophilic ligand. Our observations that anti-E-cadherin mAb inhibits the maturation of
LC despite the continuos presence of TNF According to our differentiation model, the following mechanism may
allow maintenance of immature LCs in vivo. TGF- In support of a direct functional involvement of E-cadherin in
regulating the maturation of epidermal LCs, E-cadherin is capable of
inducing a variety of cellular responses, including the regulation of
epithelial37 and hematopoietic38,39 precursor
cell differentiation. LCs express the intracellular E-cadherin-binding
signaling molecule armadillo We demonstrate that mAb binding to E-cadherin on the surfaces of
immature LCs inhibits the acquisition of mature DC characteristics. This is in line with the concept that lateral clustering of cadherin molecules is critical for cadherin-mediated functional responses, such
as tight cell adhesion.48,49 E-cadherin binding by mAb may
induce cross-linkage of E-cadherin dimers on the surfaces of LCs, which
may mimic E-cadherin clustering in response to the homophilic ligation
of E-cadherin. Recent observations with cell lines support this
concept. First, E-cadherin adhesion leads to a rapid increase in
tyrosine phosphorylation at sites of cell contact formation, and this
effect can be mimicked by anti-E-cadherin mAb.50 Second,
anti-N-cadherin mAb-coated beads functionally mimic beads coated with
the extracellular domain of N-cadherin in inducing tyrosine
phosphorylation, accumulating junction-associated Our data suggest that the loss of E-cadherin adhesion rapidly induces the maturation of LC. Despite rapid secondary E-cadherin-mediated cluster formation after the disruption of primary LC clusters, most LCs have already undergone maturation (within 24 hours). This indicates that DC maturation in response to loss of E-cadherin adhesion may represent a rapid and irreversible process. These observations may mimic the in vivo situation. In vivo, the down-regulation of E-cadherin expression by LCs seems to precede the emigration of LCs from the epidermis.8 We performed a side-by-side comparison of the effects of soluble CD40L
and TNF As mentioned, our data suggest that E-cadherin-mediated suppression of
LC maturation may prevent the uncontrolled maturation of LCs induced by
pro-inflammatory mediators. This seems to be in conflict with murine
studies, which demonstrated that the same pro-inflammatory mediators
that induce LC maturation simultaneously down-regulate E-cadherin
expression and E-cadherin-mediated adhesion and, thus, seem to reduce
the proposed maturation inhibition by E-cadherin.58
Although several parameters vary between this study and ours, it
remains to be analyzed whether exogenous TGF- Although most cells generated in primary
TGF- In conclusion, active maturation suppression of LCs by
TGF-
We thank A. Renner for his invaluable contribution in cell separation and flow sorting, M. Merad and R. Smith for critically reading the manuscript, and all the collaborating nurses and doctors of the gynecology departments at Sozialmedizinisches Zentrum Ost and Kaiser Franz Josef Spital for providing cord blood samples.
Submitted June 9, 1999; accepted August 29, 2000.
Supported by the ICP Program of the Austrian Ministry for Research and Transport.
The publication costs of this article were defrayed in part by page charge payment. Therefore, and solely to indicate this fact, this article is hereby marked "advertisement" in accordance with 18 U.S.C. section 1734.
Reprints: Herbert Strobl, Institute of Immunology, University of Vienna, Borschkegasse 8A, A-1090 Vienna, Austria; e-mail: elisabeth.riedl{at}univie.ac.at.
|
Top
Abstract
Introduction
specific phenotypic features of
mature DCs. Antibody ligation of E-cadherin on the surfaces of immature
LCs after mechanical cluster disruption strongly reduces the
percentages of mature DCs. The addition of mAbs to the adhesion
molecules LFA-1 or CD31 to parallel cultures similarly inhibits
homotypic LC cluster formation, but, in contrast to anti-E-cadherin,
these mAbs fail to inhibit DC maturation. Thus, E-cadherin engagement
on immature LCs specifically inhibits the acquisition of mature DC
features. E-cadherin-mediated LC maturation suppression may represent
a constitutive active epithelial mechanism that prevents the
uncontrolled maturation of immature LCs.
(Blood. 2000;96:4276-4284)
) or identical medium containing
anti-E-cadherin mAb (E-cad, 
) or control mAb (clone VIAP, 
), as
described in "Materials and methods" (and as also shown in Figure