Department of Dermatology, Charité, Humboldt-Universität zu Berlin, Germany
Correspondence: Magda Babina, Department of Dermatology, Charité, Campus Virchow, Humboldt-Universität zu Berlin, Augustenburger Platz 1, D-13344 Berlin, Germany. E-mail: magda.babina{at}charite.de
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, RAR
, RXRß, and RXR transcripts in all cell
lines studied, and HMC-1 cells were the only line lacking RXR.
RARß, not expressed at baseline, was induced by RA in a fashion
obviously correlating with ICAM-3 up-regulation. Increased ICAM-3
expression was of functional significance, such that processes
stimulated or co-stimulated via ICAM-3 (homotypic aggregation, IL-8
secretion) were clearly enhanced upon RA pretreatment, suggesting that
RA may contribute via hitherto unrecognized pathways to immune function
and host defense.
Key Words: vitamin A metabolites • adhesion molecules • homotypic aggregation • interleukin-8
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
dß2, as well as for the recently
discovered lectin, DC-SIGN [1
2
3
4
5
6
]. Although highly
homologous to ICAM-1 and ICAM-2, its expression profile and regulatory
potential are quite distinct from that of the other two ICAM receptors.
ICAM-1 is present on a broad spectrum of cells and up-regulated by a
variety of exogenous factors, especially by pro-inflammatory cytokines
such as interleukin-1 (IL-1), interferon- (IFN-), and tumor
necrosis factor (TNF-) [7
, 8
] as
well as by retinoic acid (RA) [9
, 10
], and
ICAM-2 expression is constitutive and found predominantly on
endothelial cells and certain leukocyte subsets [11
,
12
]. Conversely, ICAM-3 is expressed exclusively among
cells of hematopoietic origin under nonpathological conditions
[1
2
3
, 13
]. Although already high at
baseline, enhanced ICAM-3 expression was found to be associated with
lymphoid differentiation and/or activation [14
,
15
]. ICAM-3 is suggested to mediate various processes in addition to its function as adhesion receptor. Cross-linking of ICAM-3 by monoclonal antibodies (mAbs) directed against certain epitopes triggers intracellular signaling involving calcium mobilization and tyrosine phosphorylation [15 16 17 18 19 ]. The cellular consequences thereof vary from cell type to cell type but involve co-activation of lymphocytes, changes in cell morphology, chemokine production, homotypic and heterotypic aggregation, as well as cellular adhesion to immobilized integrin ligands [18 19 20 21 22 23 24 25 ]. Furthermore, the ability of ICAM-3 to enhance T-cell activation by antigen-presenting cells corroborates its importance for the initial phase of an immune response [6 , 26 27 28 ]. Thus, ICAM-3 is equipped with the property of mediating and regulating cellular adhesion as well as other physiological processes and therefore appears to be of key relevance to immune function.
RA, active metabolites of vitamin A, are implicated in a broad array of
biological processes such as development, cell growth, and
differentiation [reviewed in 29]. However, it becomes increasingly
clear that RA also possess potent immunomodulatory capacity, although
the underlying processes at the cellular level are only incompletely
understood [29
30
31
]. The predominant mode of RA action
at the molecular level is based on its activation of nuclear receptors,
which belong to the family of steroid/thyroid/retinoid-activated
transcriptional regulators [32
, 33
]. Two
classes of retinoid receptors have been described, the RA receptors
(RAR, RARß, RAR) and the retinoid X receptors (RXR, RXRß,
RXR), with each receptor subtype being encoded by a separate gene.
Genes responsive to RA contain RA response elements (RARE) within their
promoter/enhancer regions consisting of two half-sites of a PuG(G/T)TCA
or related motif [32
, 33
]. In most cases,
an efficient transduction of the RA signal requires heterodimerization
between RAR and RXR [34
].
Although ICAM-3 up-regulation by RA has been shown before for the acute promyelocytic leukemia (APL) cell line NB-4 by flow cytometry, neither the functional consequences thereof nor the cell-type dependence of the process has been addressed [35 ]. In addition, a strong distinction between individual cell types with regard to retinoid-induced processes is widely appreciated. In contrast to most other myeloid cell lines, effects of RA on cells of the mast cell lineage have been poorly investigated so far. We have, however, demonstrated previously that expression of mast cell adhesion receptors is modulated by RA treatment in a highly selective manner [36 , 37 ]. We show herein that several RA isoforms also affect ICAM-3 expression and function by immature HMC-1 cells and certain other cell types, thereby supporting and enlarging the previous observation that pathways regulating ICAM-3 expression prior to translation also exist for myeloid cells [35 ]. Furthermore, the presence of several retinoid receptors in the mast cell line, shown for the first time by the present study, offers a rational explanation for the susceptibility of these cells to RA action.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, IgG1), 3E8
(anti-LFA-1ß, IgG1), and D2.10 (anti-glycophorin A, IgG1) were
supplied by Immunotech (Marseille, France). mAb RR1/1, recognizing
human ICAM-1 and mAb CBR-IC3/1 against ICAM-3 (binding to IgSF domain I
[20
]), were from Bender MedSystems (Vienna, Austria).
Streptavidin-peroxidase was from Boehringer-Mannheim (Mannheim,
Germany). Fluorescein isothiocyanate (FITC)-conjugated, affinity-pure
F(ab')2 fragment goat anti-mouse was from Jackson Immuno
Research Laboratories (West Grove, PA). The calcium ionophore A23187
was supplied by Calbiochem (San Diego, CA). Phorbol 12-myristate
13-acetate (PMA), 9-cis, 13-cis, and all-trans RA were purchased from
Sigma (München, Germany). The RAR-specific agonists CD336, CD437,
and CD2019 were a kind gift of Dr. U. Reichert (CIRD, Galderma,
France).
Cell culture and treatment
The human mast cell line HMC-1 [38
] was kindly
provided by Dr. Butterfield (Rochester, MN). Cells of the HMC-1
subclone 5C6 were used throughout [39
]. They were grown
in basal Iscove’s medium and supplemented with 10% heat-inactivated
fetal calf serum (FCS) and 10 µM monothioglycerol (Sigma). The human
basophilic precursor cell line KU812 [40
] was kindly
provided by the Borstel Research Institute (Germany) and cultured in
RPMI 1640 medium containing 15% FCS and 50 µM ß-mercaptoethanol
(Sigma). The promyeloid cell line HL-60 [41
] and the
promonocytic cell line U-937 [42
] were obtained from the
American Type Culture Collection (ATCC, Rockville, MD) and maintained
in RPMI 1640 medium/10% FCS. All cell culture media contained 4 mM
L-glutamine, streptomycin, and penicillin (all from Seromed, Berlin,
Germany). Monocytes were isolated from buffy coats of healthy donors by
Ficoll-Hypaque (Seromed) density-gradient centrifugation, followed by
one step of adherence to plastic culture dishes at 37°C for 30–40
min. Adherent cells were >95% positive for CD14, as determined by
flow cytometry. Isolated monocytes were kept in RPMI 1640 medium/10%
FCS, 4 mM L-glutamine, and antibiotics.
Isolation of skin mast cells was performed as described previously with several modifications [37 ]. Briefly, foreskins were cut into stripes and treated with dispase (Boehringer-Mannheim) at 0.5 mg/ml and 4°C overnight. The epidermis was then removed from the dermis, the latter chopped in small pieces and treated with collagenase at 5 mg/ml (type 4, Worthington, Lakewood, NJ) for 1 h at 37°C. Cells were separated from remaining tissue by three steps of filtration. Mast cells were further purified by positive selection using anti-c-kit mAb YB5.B8 (kindly provided by Dr. L. K. Ashman), goat anti-mouse Ig-coated magnetic beads (Miltenyi Biotec, Bergisch Gladbach, Germany), and a MACS separation device. Mast cell purity in these preparations ranged from 85% up to 95% typically, as assessed by toluidine blue staining.
Cells were treated in standard growth medium containing 9-cis, 13-cis, all-trans RA, or RAR-specific agonists at 37°C for the periods and at the concentrations indicated. Control cells were kept in medium alone for the same time.
Flow cytometry
Approximately 5 x 105 cells were washed twice
in Ca2+/Mg2+-free
phosphate-buffered saline (PBS), reacted with saturating concentrations
of primary mAb diluted in human AB-serum (Biotest, Dreieich, Germany)
for 30 min at 4°C, washed, and stained with FITC-conjugated
F(ab')2 fragment of goat anti-mouse Ig for 20 min at 4°C.
After being washed, cells were analyzed by an EPICS XL flow cytometer
(Coulter Electronics, Krefeld, Germany). A minimum of 10,000 cells were
investigated/sample. Mean channel fluorescence (MCF) was considered to
correlate with surface-Ag density. The MCF value of untreated cells was
considered as 100%. Cell-surface molecule density of differently
treated cells was determined as MCF (treated cells)/MCF (medium
control) x 100. An isotype-matched, nonbinding mAb (R&D Systems)
was used as negative control.
In case of skin mast cells, 2 x 105 cells were used per assay and stained with FITC-coupled, anti-ICAM-3 mAb CBR-IC3/1 or isotype-matched, nonbinding mAb.
Immunoprecipitation, SDS-PAGE, and Western blot
Proteins at the exterior cell surface of HMC-1 cells were
biotinylated specifically using the BM cellular labeling and
immunoprecipitation kit (Boehringer-Mannheim), following the
manufacturer’s instructions. After labeling, cells were lysed at
1 x 107/ml for 1 h at 0°C in 1% Nonidet P-40,
0.5% sodium deoxycholate, 50 mM sodium borate, pH 8.0, 150 mM NaCl,
100 µg/ml phenylmethylsulfonyl fluoride (PMSF), 1 µg/ml aprotinin,
and 1 µg/ml leupeptin. Protein concentration was determined using the
bicinchoninic acid (BCA) protein determination kit (Pierce, Rockford,
IL), and samples were adjusted to the same protein concentration (600
µg/ml) prior to being precleared overnight using the irrelevant mAb
D2.10 and 50 µl anti-mouse IgG1 (heavy chain-specific) agarose beads
(Sigma) per 1 ml lysate. The supernatants were reacted with 3 µg mAb
followed by addition of 100 µl beads, and mixtures were incubated
overnight under gentle rocking at 4°C. Beads were collected after
brief centrifugation and washed, and proteins were eluted with 2 x sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)
sample buffer (100 mM Tris-HCl, pH 6.8, 4% SDS, 20% glycerol, 200 mM
dithiothreitol, and 0.2% bromphenol blue) prior to being heated at
95°C for 5 min. Immunoprecipitated proteins were then resolved by a
discontinuous SDS-PAGE using a 6% separating gel and a 5% stacking
gel and subsequently blotted onto polyvinylidene difluoride transfer
membrane (Du Pont de Nemours, Bad Homburg, Germany) using a semi-dry
transfer cell (Bio-Rad Laboratories, Richmond, CA). The membrane was
saturated with 5% nonfat dried milk and incubated with
streptavidin-peroxidase (Boehringer-Mannheim) at 50 mU/ml for 30 min at
room temperature. After extensive washing, immunoprecipitated proteins
were visualized with 3,3',5,5'-tetramethylbenzidine (TMB, insoluble;
Calbiochem).
Enzyme-linked immunosorbent assay (ELISA)
A commercially available ELISA kit specific for human soluble
ICAM-3 (R&D Systems) was used for quantification of ICAM-3 expression.
Differently treated HMC-1 cells were lysed at 1 x
107/ml in 1% Triton X-100, 50 mM Tris-HCl, pH 8.0, 150 mM
NaCl, 100 µg/ml PMSF, 1 µg/ml aprotinin, and 1 µg/ml leupeptin.
Protein concentrations were determined with the BCA protein
determination kit as above. The lysates were diluted 1:5 with assay
diluent during the assay. ICAM-3 concentration was determined relative
to a standard curve and is expressed as ng ICAM-3/mg total protein.
For quantification of IL-8 in the supernatants of HMC-1 cells, cells were pretreated with or without RA for three days and then washed three times in serum- and RA-free medium. Cells from the different groups were plated in culture dishes at 1 x 106/ml under serum-free conditions, preincubated or not with different anti-ICAM-3 mAbs (30 min at 37dgC), and then stimulated by PMA (at 25 ng/ml) alone or in combination with the calcium ionophore A23187 (at 500 nM). After 24 h, supernatants were removed and kept at -80°C until use. IL-8 content was detected by a specific ELISA kit (R&D Systems) following the manufacturer’s instructions.
Northern blot
For comparison of ICAM-3-specific mRNA content, total cellular
RNA was extracted using the RNeasy Midi Kit (Qiagen, Hilden, Germany),
according to the manufacturer’s protocol. RNA concentration was
assessed by the RiboGreen RNA quantitation kit (Molecular Probes,
Göttingen, Germany). Samples of 15 µg RNA were electrophoresed
through a 1% agarose/2.2 M formaldehyde gel, transferred to GeneScreen
Plus nylon membrane (NEN, Zaventem, Belgium) by capillary transfer, and
fixed by ultraviolet light.
Hybridization with peroxidase-labeled DNA probe (generated with the primer pairs given in Table 1 ) at 10 ng/ml and specific detection of messenger RNAs were performed with the North2South labeling and detection system, according to the manufacturer’s protocol (Pierce). Specific transcripts were visualized by enhanced chemiluminescence. After drying, the same membrane was rehybridized with a peroxidase-labeled glyceraldehyde 3-phosphate dehydrogenase (GAPDH) probe as control of equal RNA loading.
|
View this table: [in a new window] |
Table 1. Primer Pairs Used for PCR Amplifications
|
, 61°C; RARß and RXR, 63°C;
RXRß and RXR, 70°C; and RAR, 72°C. RNA at the same dilution
as cDNA was used as negative control and run under the same conditions.
Aggregation assay
Quantitative aggregation assays were performed in 96-well
microtiter plates (Greiner, Frickenhausen, Germany) essentially as
described [37
]. In brief, cells treated without (medium
control) or with 9-cis RA, 13-cis RA, or all-trans RA (all at 10 µM)
for 4 days were plated in culture dishes at 5 x
105/ml and stimulated with anti-ICAM-3 mAbs (each at 1
µg/ml) in a total reaction vol of 100 µl at 37°C for the times
indicated. After incubation, suspensions of each sample were mixed by
pipetting up and down 20 times and were transferred to a hemocytometer.
The number of free cells and total cell number were assessed, and
percent aggregation was determined as (1 number of free cells/total
cell number) x 100.
Statistics
Statistical analysis was carried out with the unpaired
Student’s t-test.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
 View larger version (29K): [in a new window] |
Figure 1. Cell-specific ICAM-3 regulation in response to all-trans RA. Different
myeloid cell types were cultured in the presence or absence of
all-trans RA (all at 1 µM) for 3 days and investigated concerning
surface expression of ICAM-3 by flow cytometry. Representative
histograms of at least three separate tests are shown. Black line,
cells kept in medium only; gray line, all-trans RA-treated cells. Very
similar results were obtained when 9-cis or 13-cis RA were used instead
of all-trans-RA.
|
|
View this table: [in a new window] |
Table 2. Modification of ICAM-3 and ICAM-1 Expression on Different Myeloid Cells
Following RA Treatment
|
Conversely, ICAM-1 expression was increased substantially on KU812, HL-60 cells, and blood monocytes but not on HMC-1 or U-937 cells (Table 2B) . In contrast to ICAM-3, RA-mediated effects on ICAM-1 obviously depended on the precise RA isoform, with all-trans RA being lowest throughout in potency. Also, the maximal response was lower for ICAM-1 than for ICAM-3.
Dose-response studies revealed concentration dependency of ICAM-3 up-regulation over a very wide concentration range from 0.0001 up to 10 µM (Fig. 2A ). Slight differences among the three RA isoforms could be detected: Although all-trans RA appeared most potent at very low doses, 9-cis RA was most effective in the range of 0.1–1 µM. At 10 µM, there were no significant differences among the RA isoforms. Because 9-cis RA represents the only RA isoform interacting efficiently with RXR in addition to its binding to the RAR [47 48 49 ], this suggests that ICAM-3 up-modulation occurs independently from transactivation through RXR, fitting the concept that RXR does not necessarily need to be ligand-activated on its own and functions as a silent partner in RXR-RAR heterodimers [32 , 33 ]. The anticipated EC50 values (corresponding to the RA concentration that exerts half-maximal ICAM-3 induction) were approximately 0.03 µM (9-cis RA), 0.08 µM (13-cis RA), and 0.01 µM (all-trans RA). Time-course studies revealed that ICAM-3 induction was slow in onset, first detectable after 2 days, and further increasing up to day 5 after treatment (Fig. 2B) .
 View larger version (28K): [in a new window] |
Figure 2. Concentration and time dependency of ICAM-3 up-regulation on HMC-1
cells by 9-cis, 13-cis, and all-trans RA. (A) Cells were treated with
various concentrations of 9-cis, 13-cis, and all-trans RA for 72 h
at 37°C or (B) kept in the presence or absence (control) of either RA
for the indicated time periods. ICAM-3 surface density was quantitated
by flow cytometry using CAL3.10 as anti-ICAM-3 mAb. Data are expressed
in relation to the medium control, which was normalized at 100%.
Results are means ± SD of three (A) or four (B)
independent assays.
|
 View larger version (30K): [in a new window] |
Figure 3. Up-modulation of ICAM-3 expression by RA as detected by Western blot
analysis (A) and ELISA (B). HMC-1 cells were cultured in the presence
or absence of 9-cis, 13-cis, or all-trans RA (all at 10 µM, 72 h), after which time, cells were harvested. (A) The cell surface was
biotinylated, the cells were lysed, and ICAM-3 was immunoprecipitated
using the mAbs CAL3.34 (lanes 1–4), ICAM-3.3 (lanes 5–8), or D2.10
(negative control; lanes 9–12). Untreated cells: lanes 1, 5, and 9;
9-cis RA: lanes 2, 6, and 10; 13-cis RA: lanes 3, 7, and 11; all-trans
RA: lanes 4, 8, and 12. (B) ICAM-3 content in lysates of untreated or
RA-treated (10 µM, 72 h) HMC-1 cells was determined by an
ICAM-3-specific ELISA. Protein concentration was quantitated by the BCA
protein determination kit. Data are presented as ng ICAM-3/mg total
protein and are the mean ± SD of three independent
experiments.
|
 View larger version (57K): [in a new window] |
Figure 4. RA treatment results in an augmented steady-state level of
ICAM-3-specific transcript in HMC-1 but not in KU812 cells. Total RNA
was extracted from untreated or RA-treated (all-trans and 9-cis RA,
each at 1 µM, 72 h) HMC-1 and KU812 cells, and equal amounts
thereof were subjected to Northern blot analysis as described in
Materials and Methods. Lanes 1 and 4: Cells kept in medium only; lanes
2 and 5: cells treated with 9-cis RA; lanes 3 and 6: cells treated with
all-trans RA. The Northern blot was hybridized sequentially with ICAM-3
and GAPDH probes, as indicated in the figure. Blots are representative
of four independent tests.
|
As shown in Table 3 , significant differences between RA-pretreated and control cells could be detected after 2 h and 24 h with the inducing anti-ICAM-3 mAbs CAL3.10 or CBR-IC3/1, and spontaneous aggregation and aggregation in the presence of inert mAb CAL3.34 were unaffected by RA treatment. Very similar results were obtained when 9-cis and 13-cis RA were used instead of all-trans RA (unpublished results).
|
View this table: [in a new window] |
Table 3. Up Regulation of ICAM-3 by RA Leads to Enhanced Homotypic Aggregation
Induced via ICAM-3
|
 View larger version (38K): [in a new window] |
Figure 5. Effects of RA pretreatment on IL-8 release from HMC-1 cells
co-stimulated via ICAM-3. HMC-1 cells were cultured in the presence or
absence of all-trans RA (10 µM, 72 h), after which time cells
from both groups were washed in serum-free medium and plated in culture
dishes at 1 x 106/ml. Preincubation (30 min, 37°C) was
done in the presence or absence of the anti-ICAM-3 mAbs CBR-IC3/1 or
CAL3.34 (each at 2 µg/ml) prior to stimulation by PMA (at 25 ng/ml)
or PMA + A23187 (25 ng/ml and 500 nM, respectively) for 24 h.
IL-8 concentration in the cell-free supernatants was measured by an
IL-8-specific ELISA (R&D Systems). Data from one single experiment
performed in duplicate are shown. Similar results were obtained in two
further tests.
|
As determined by RT-PCR technique and summarized in Table 4
, HMC-1, KU812, U-937, and HL-60 cells all expressed RAR,
RAR, RXRß, and RXR at baseline and following treatment with
either RA isoform. In contrast to the other cell lines, it is
interesting that HMC-1 cells did not exhibit appreciable RXR mRNA
levels, irrespective of treatment. The only receptor whose expression
was strongly influenced by RA was RARß. None of the cell lines
expressed RARß at baseline, but its transcript was clearly detected
in RA-treated HMC-1 and HL-60 cells, very weakly in KU812, and not at
all in U-937 cells. Expression patterns of RAR and RXR in HMC-1 and
KU812 cells are presented in Figure 6
.
|
View this table: [in a new window] |
Table 4. Expression Profiles of RAR and RXR in HMC-1, KU812, HL-60, and U-937
Cells
|
 View larger version (61K): [in a new window] |
Figure 6. Expression pattern of RAR and RXR in HMC-1 and KU812 cells.
Representative RT-PCR amplifications using RNA of untreated cells (lane
1) and cells treated with 9-cis RA (lane 2), 13-cis RA (lane 3), and
all-trans RA (lane 4) are shown. Lanes 5 and 6: negative controls using
RNA from untreated and RA-treated cells instead of cDNA. RT-PCR
reactions were carried out as described in Materials and Methods using
a minimum of three different cell preparations. Note that RARß
expression in RA-treated KU812 cells is undetectable in this form of
presentation, although a faint band was visible in the original gel in
three out of three experiments.
|
, RARß, and RAR agonists (all at 1 µM) for three days,
after which time ICAM-3 expression was investigated by flow cytometry
and by ELISA. As illustrated in Figure 7
, the highest effect on ICAM-3 surface density was obtained with
the RAR-selective compound CD437, followed by the RAR-specific
retinoid CD336. Conversely, the RARß agonist CD2019 produced a weak
effect only. CD437, however, also possessed potent influence on
morphological features and viability of HMC-1 cells (unpublished
results), and CD336 and CD2019 displayed no such effects and were thus
better comparable to all-trans RA.
 View larger version (26K): [in a new window] |
Figure 7. ICAM-3 modulation by RAR-selective agonists. HMC-1 cells were cultured
in the presence of the retinoids all-trans RA, CD336, CD2019, CD437, or
CD336 + CD2019 (all at 1 µM for 72 h) or kept in medium
alone. Cells were then investigated concerning their ICAM-3 expression
by the use of flow cytometry (A) or ELISA (B). Data are the mean ± SD of three independent experiments. *, P <
0.05; **, P < 0.01; ***, P < 0.001 versus the
medium control.
|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In addition to HMC-1 cells, ICAM-3 regulation by RA was found to be a more generalized phenomenon, with strong differences between individual cell types, however. Although the highest effect was observed for HL-60 cells, responsiveness of KU812 and U-937 cells was low or absent, and strong donor-dependence was revealed for blood monocytes. Although HL-60 and HMC-1 cells vary remarkably with regard to morphology and phenotype, the positive influence of RA on their ICAM-3 expression suggests that a certain degree of myeloid immaturity (HL-60) or mast cell commitment (HMC-1) might be necessary prerequisites for the process.
Our data are in line with previous observations made by Larson et al. [35 ], who showed up-regulation of ICAM-3 expression in response to RA for the APL cell line NB-4, further proving that ICAM-3 is an RA-inducible molecule. Conversely, our study shows that the process of ICAM-3 up-regulation is clearly cell-type-dependent affecting some but not all myeloid cells. Based on the correlation between steady-state level of ICAM-3 transcript (clearly shown to increase in HMC-1 but not in unresponsive KU812 cells) and ICAM-3 protein in our study, the regulatory process is likely to occur pretranslationally by enhanced transcription or selective mRNA stabilization. Whether additional mechanisms (such as enhanced translation and/or reduced protein degradation) also contribute to ICAM-3 up-regulation cannot be completely ruled out, however.
It is interesting that ICAM-3 expression of fully differentiated skin mast cells was not modulated by RA, suggesting that some degree of cellular immaturity is a necessary prerequisite for the response. ICAM-3 expression on mast cells most likely decreases in a differentiation-dependent manner, as discussed previously [45 ]. Therefore, terminally differentiated mast cells express ICAM-3 to a much lower degree, with the dominant percentage of cells negative for ICAM-3 expression (Fig. 1) , and mature cells are obviously unable to re-express it following administration of exogenous stimuli like RA. Also, the HMC-1 line was established from a patient with mast cell leukemia [38 ], and there are differences with regard to responsiveness of leukemic cells and their normal counterparts toward RA also in other cellular systems such as monocytes [50 ]. Because committed precursors of the mast cell lineage have not been identified so far in the circulation, these cells thus cannot be isolated for study [51 ]. Immature mast cells have, however, been shown to increase in various types of inflammatory conditions [52 , 53 ], including psoriasis, an RA-responsive disease; ICAM-3 may well be involved in the increased influx into the tissue and subsequent increased mediator production of these cells.
The expression of the homologous ICAM-1 receptor was also cell-specifically affected by RA in a fashion that did not correlate to ICAM-3 modulation. HMC-1 cells, for example, did not up-modulate their surface ICAM-1 after a 72-h incubation (Table 2B) , and an increase was detected for monocytes, KU812, and HL-60 cells, with the strongest effect being, however, lower for ICAM-1 than for ICAM-3. ICAM-1 expression has been shown in different studies to be up-regulated by RA [9 , 54 ], and the ICAM-1 promoter is known to harbor a degenerate RARE involved in its transcriptional regulation [10 ]. Given the high homology between ICAM-1 and ICAM-3, it is tempting to speculate that the ICAM-3 promoter may also possess potential RARE(s), but clarification of the precise mechanism by which RA up-regulates ICAM-3-specific transcript will await characterization of the organization and regulatory sequences of the ICAM-3 gene.
The selective up-regulation of ICAM-3 versus that of ICAM-1 as observed for the HMC-1 cell line (and given it also happens in vivo) could lead to a shift in a cell’s preference to bind to LFA-1 by these two counter-receptors, thereby favoring ICAM-3-mediated signal transduction leading to aggregation and cytokine production. These processes appear selective for ICAM-3 because we have not been able to induce any of them by mAbs directed against various epitopes of ICAM-1 (unpublished results), which is in accordance with the hypothesis that ICAM-3 is the predominant signal transducer among the ICAM subfamily.
Because our previous [36
, 37
] and present
data strongly implied the presence of retinoid receptors in mast and
basophilic cells, this issue was addressed in greater detail, revealing
expression of RAR-, RAR-, RXRß-, and RXR-specific mRNA in
HMC-1 and KU812 cells (in addition to HL-60 and U-937 cells). Our data
for HL-60 cells are largely in accordance with previous studies:
Although expression of RAR and RXR in these cells was shown by
several groups, slight differences were obtained for the other receptor
subtypes, possibly because of differences in the precise subline used
or the detection limit of the method applied [55
56
57
58
59
60
61
62
].
It is of interest, however, that HMC-1 cells represented the only cell
line lacking detectable RXR expression. Because RA induction of
ICAM-1 at the transcriptional level has been demonstrated to be
mediated by RARß in combination with RXR [63
], the
inefficiency of RA to modulate ICAM-1 expression in HMC-1 cells might
be a result of the lack of RXR expression. Because RXR is
expressed by all other hematopoietic cell types thus far investigated
[60–62, 64, 65, and this study], the absence of RXR in HMC-1
cells makes them an attractive model with which to delineate processes
selectively dependent on RXR action.
With regard to RARß, none of the cell lines under investigation expressed detectable levels of its transcript prior to RA treatment. However, its induction by RA was observed in HMC-1 and HL-60 and very weakly in KU812 cells. RARß has been shown previously to be induced vigorously by retinoids in different cells [56 , 66 67 68 ], and the RARß gene is known to harbor a functional RARE in its P2 promoter involved in this process [69 ].
The fact that RARß induction, as shown here, occurred in a
cell-specific manner, obviously correlating with RA-mediated modulation
of ICAM-3, suggested a functional role for this receptor in myeloid
ICAM-3 regulation or the requirement for additional factors responsible
for RARß and ICAM-3 induction. We tested the first hypothesis by the
use of selective RAR, RARß, and RAR agonists. Although
RAR-specific CD437 produced the highest response with regard to
ICAM-3, it additionally possessed very pronounced effects on
proliferation and viability of HMC-1 cells (in contrast to RA and the
other retinoids used), making it difficult to ascertain whether its
positive influence on ICAM-3 expression was a direct effect of the
compound or secondary to other changes. Further, CD437 has been shown
previously to act by other mechanisms in addition to those mediated by
RAR [70
, 71
]. Because RARß-selective
CD2019 was lowest in potency by far, not enhancing ICAM-3 expression
further even when combined with CD336, these experiments, however,
largely ruled out the possibility that RARß might itself be involved
in ICAM-3 regulation. Therefore, it appears more likely that RARß
induction and ICAM-3 up-regulation in myeloid cells depend on the same
prerequisites, one of which may be presence of COUP-TF, which has been
shown very recently to be involved in RARß expression
[72
]. Studies aimed at assessing expression of this
transcription factor in the cell lines we used are underway currently.
ICAM-3 is known to mediate and regulate cellular adhesion and to be capable of signal transduction. Therefore, two experimental approaches were used to elucidate the functional significance of enhanced ICAM-3 expression. First, ICAM-3 surface density was found to correlate with the level of HMC-1 aggregation triggered through ICAM-3. Second, ICAM-3 up-regulation resulted in enhanced IL-8 production by HMC-1 cells co-stimulated via ICAM-3. It is of interest that RA pretreatment of HMC-1 cells exerted by itself a significant effect on stimulated secretion of the chemokine, implying that RA is possibly able to regulate other mast cell functions in addition to those involving adhesion receptors. However, the enhancing effect of anti-ICAM-3 mAb was clearly more striking for RA-pretreated cells (i.e., cells expressing a higher level of ICAM-3) than for control cells, and this was detected with both activation regimens used, namely PMA alone and PMA in combination with A23187, further substantiating a role for newly synthesized ICAM-3 for mast cell cytokine production. It remains to be established whether enhanced ICAM-3 expression on leukocytes may also be viewed in the context of their clearance by macrophages following apoptosis. ICAM-3 expressed by apoptotic cells was demonstrated recently to function as ligand for a macrophage counter-structure, possibly CD14, and to be involved in intercellular interactions relevant to removal of apoptotic cells [73 ]. It should be noted that CD437, producing the highest effect with regard to ICAM-3, was (in contrast to RA) also a potent inducer of HMC-1 apoptosis (unpublished results). Thus, increased ICAM-3 expression by RA may also be exploited experimentally to delineate further functional aspects of the molecule.
ICAM-3 appears to be especially important for mediation (by binding to
LFA-1, dß2, and the very recently
described counter-receptor, DC-SIGN) and regulation of adhesion (by its
property to induce homotypic and heterotypic adhesion events and
adhesion of leukocytes to components of the extracellular matrix, the
latter event being mediated by ß1 integrins).
RA is used therapeutically in different disease states, most importantly for the treatment of APL. An undesired, adverse effect in those settings is the RA syndrome, which is characterized by fever, progressive hypoxemia, and multiorgan failure [74 ]. The basic mechanisms underlying this syndrome are poorly understood but appear to involve extravasation of leukemic cells and other leukocytes into tissues and organs, and cytokine release. Therefore, RA-mediated alterations in expression and/or function of adhesion molecules are likely involved. Thus, the high levels of RA achievable in serum during RA therapy could in fact up-regulate ICAM-3 expression by leukemic and normal myeloid cells, thereby not only strengthening cellular binding to LFA-1-expressing cells but also increasing the processes stimulable through this receptor, such as activation state of other adhesion receptors and, very importantly, production of inflammatory or immunomodulatory cytokines, as shown in this study (Fig. 5) and substantiated by previous studies [25 , 45 ]. Because ICAM-3 is such a versatile molecule, its up-modulation on different cells (including APL cells, as shown previously [35 ]) may be of relevance to the events observed during RA syndrome, such as migration of leukocytes into respiratory tissue and kidney.
Together with our very recent study of 1,25-dihydroxyvitamin D3-mediated, ICAM-3 up-modulation by HMC-1 cells [75 ], the present study thus established a regulatory pathway for ICAM-3 expression used by some but not all myeloid cells. Pronounced ICAM-3 expression on HMC-1 cells correlated with certain functional properties of the molecule, and others (such as clearance of apoptotic leukocytes) need to be addressed in the future. The detailed analysis of retinoid receptor expression patterns in mast (and basophilic) cells is thought to contribute to a better understanding of the observed effects exerted by RA. In addition to ICAM-3 up-regulation, primarily addressed in this study, RA indeed exerted other profound effects on HMC-1 cells such as RARß induction and enhanced IL-8 secretion. Thus, these cells may represent a further interesting model system for investigation of the basic principles of RA action on growth, differentiation, and immune function. Some of these studies are currently underway.
Received April 3, 2000; revised October 29, 2000; accepted November 20, 2000.
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
dß2, binds preferentially to ICAM-3 Immunity 3,683-690[Medline]: tissue distribution, biochemistry, and function of a natural adherence molecule (ICAM-1) J. Immunol. 137,245-254[Abstract] interferon J. Invest. Dermatol. 98,741-747[Medline] and ß genes EMBO J 8,429-433[Medline] and ß Proc. Natl. Acad. Sci. USA 86,3089-3093 and retinoid X receptor- RNA by HL-60 cells Blood 82,3592-3599,25-dihydroxyvitamin D3 co-operate to promote differentiation of the human promyeloid leukemia cell line HL60 to monocytes Leukemia 8,806-815[Medline] complexes stimulate expression of the intercellular adhesion molecule-1 gene Cell Growth Differ 6,515-521[Abstract],25-dihydroxyvitamin D3: selective promotion of ICAM-3 expression and constitutive presence of vitamin D3 receptor Biochem. Biophys. Res. Commun. 273,1104-1110[Medline]This article has been cited by other articles:
 |
|
 |
|
  J. Wang and A. Yen A Novel Retinoic Acid-Responsive Element Regulates Retinoic Acid-Induced BLR1 Expression Mol. Cell. Biol., March 15, 2004; 24(6): 2423 - 2443. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Babina, S. Guhl, A. Starke, L. Kirchhof, T. Zuberbier, and B. M. Henz Comparative cytokine profile of human skin mast cells from two compartments--strong resemblance with monocytes at baseline but induction of IL-5 by IL-4 priming J. Leukoc. Biol., February 1, 2004; 75(2): 244 - 252. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
