| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
*
Center for Experimental Therapeutics, University of Pennsylvania School of Medicine, Philadelphia, PA 19104;
Wistar Institute, Philadelphia, PA 19104;
Department of Pharmacology, Catholic University School of Medicine, Rome, Italy; and
§
Department of Medicine and Aging, University of Chieti "G. D’Annunzio" School of Medicine, Chieti, Italy
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
up-regulated expression and activity of COX-2 in
medullary cells, in which COX-2 was expressed constitutively. In
contrast, IFN- down-regulated COX-1 activity, but not expression, in
cortical cells. Stromal cells support T cell development in the thymus,
although the mediators of this effect are unknown. Selective inhibition
of COX-2, but not COX-1, blocked the adhesion of
CD4+CD8+ and CD4+CD8-
thymocytes to medullary cell lines. No effect of the inhibitors was
observed on the interactions of thymocytes with cortical epithelial
lines. These data further support the differential regulation of COX-1
and COX-2 expression and function in thymic stromal cells. PGs produced
by COX-2 in the medullary thymic stroma may regulate the development of
thymocytes by modulating their interaction with stromal
cells. |
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Increasing evidence supports the hypothesis that eicosanoids modulate the immune response mediated by mature T cells, as well as T cell development within the thymus. For example, PGE2 modulates cytokine secretion by both human and murine mature CD4+ T cells. This eicosanoid, which is synthesized by APCs, such as macrophages and fibroblasts, inhibits Th1 cytokines, such as IL-2, whereas it enhances production of Th2 cytokines, such as IL-4 and IL-5 (10, 11, 12, 13, 14, 15). PGE2 also inhibits the adhesion and motility of cytotoxic T lymphocytes (16). Synthesis of PGE2 is increased in clinical syndromes of immune disfunction, such as AIDS (17, 18), autologous bone marrow and stem cell transplantation (19), atopic dermatitis, and the hyper IgE syndrome (20).
Aside from the effects of PGs within the mature T cell compartment, they may also influence maturation of the T cell lineage. Thus, thymic expression of various PG biosynthetic enzymes and receptors has been detected in human (21, 22, 23, 24), mouse (23, 24, 25), and rat (26) thymus. Furthermore, thymus and nonlymphoid thymic stromal cell lines have been shown to secrete PGs in vitro (27, 28, 29). While PGE2 protects a human thymocyte cell line from apoptosis (30), it has also been reported to initiate apoptosis in neonatal primary thymocytes (31). More recently, using COX-deficient mice and pharmacological inhibitors, we have found that COX-1-dependent PGE2 formation facilitates the transmission of CD4-CD8- double-negative (DN) thymocytes to CD4+CD8+ double-positive (DP) cells, acting via the EP2 subtype of the PGE2 receptors (32). By contrast, COX-2-dependent PGE2 formation supports CD4+ single-positive (SP) lymphocyte selection and earlier stages of thymocyte differentiation, in this case acting via the EP1 subtype of the PGE2 receptor (32).
Differentiation and migration of thymocytes occurs within a complex cellular network known as the thymic stroma (33). Among the constituents of the stroma are nurse cells, cortical reticular epithelial cells, medullary epithelial cells, nonepithelial dendritic cells, and macrophages (33, 34). It is thought that each of these cell types subserves distinct functions in thymocyte development, but the molecular bases of these cellular interactions are poorly understood. Cortical epithelial cells are thought to influence positive selection, while medullary epithelial cells appear to contribute to induction of tolerance and to the later stages of thymocyte development (33, 34).
Given our observations in the fetal thymus, where the spatial
expression and functional roles of the COX isozymes are distinct (32),
we wished to examine the potential role of PGs in mediating
interactions between thymocytes and stromal cells. We have used the
role of IFN- to regulate the expression of COX isozymes in cell
lines representative of cortical nurse (427.1), cortical reticular
(1308.1), medullary interdigitating-like (6.1.1), and medullary
epithelial (1307) cells in the thymic stroma (35). We report
differential expression of the COX isozymes in cortical vs medullary
cells. COX-1 expression is greatest in cortical cell lines. COX-2, by
contrast, is constitutively expressed predominantly in medullary cells.
IFN- up-regulates COX-2 and consequent production of
PGE2 in medullary cells, while it reduces COX-2 protein
expression and PGE2 production in cortical cell lines.
Selective inhibition of COX-2, but not COX-1, prevents adhesion of DP
and CD4 SP thymocytes to medullary, but not to cortical stromal cells,
an effect which is reversed by PGE2. These results suggest
that COX-2-dependent PGE2 formation in the medulla may
influence selective interactions between developing thymocytes and
constituents of the thymic stroma.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Recombinant mouse IFN- was obtained from PharMingen (San
Diego, CA). AA and PGE2 were purchased from Cayman Chemical
(Ann Arbor, MI). NS-398 was purchased from Biomol (Plymouth Meeting,
PA). L-759,700 was a generous gift of Dr. W. Tanaka (Merck).
Nonidet P-40, indomethacin, BSA, and deoxycholic acid were bought from
Sigma (St. Louis, MO). Tween 20 was obtained from J. T. Baker
(Phillipsburg, NJ). Enhanced chemiluminescence (ECL) substrates and
Hyperfilm ECL were obtained from Amersham (Arlington Heights, IL).
Donkey polyclonal anti-mouse or anti-rabbit IgGs coupled to
fluorescein or peroxidase were obtained from Jackson ImmunoResearch
Laboratories (West Grove, PA). Electrophoresis reagents were obtained
from Bio-Rad (Hercules, CA). Anti-CD4-FITC and
anti-CD8-Red-613 were obtained from Life Technologies
(Gaithersburg, MD). C57BL and RAG-1-/- mice were obtained
from Jackson Laboratories (Bar Harbor, MA) and maintained under
standard conditions. Anti-mouse E-cadherin Ab was obtained from
Transduction Laboratories (Lexington, KY). Anti-mouse-ICAM-1, CD40,
CD44 Abs, and hyaluronate-fluorescein-conjugated Abs (36, 37) were a
generous gift from Dr. Ellen Puré (The Wistar Institute,
Philadelphia, PA). Anti-mouse ß actin was obtained from Sigma.
Cell biology
Cells were cultured in DMEM, supplemented with 10% FCS, 2 mM
L-glutamine, 5 x 10-5 M 2-ME, and 1 mM
HEPES buffer. Confluent cells were washed in PBS and incubated with
different concentrations of rIFN- added to the culture medium. Cells
were collected at selected times and proteins were extracted as
described below. Confluent cells were washed twice with HBSS containing
1 mg/ml BSA and incubated for 40 min in 4 ml of the same buffer with 10
µM AA to evaluate COX activity. Supernatants were collected for
measurements of PGs. Cells were harvested by trypsinization and
counted. The experiments were always performed in duplicate. Confluent
cells were incubated with increasing concentrations of IFN- to study
COX activity for 18–20 h, washed with HBSS, and stimulated with AA.
In experiments performed with COX inhibitors, cells were incubated for
40 min with L-759,700, NS-398, or indomethacin. Indomethacin was
solubilized in absolute ethanol; NS-398 and L-759,700 were prepared in
DMSO. Aliquots from each stock solution were added to fresh medium, and
each preparation contained 
0.1% ethanol or DMSO (v/v, final
concentration), as did the control medium. Cells were then washed with
HBSS/BSA and incubated for 40 min in the same buffer with 10 µM AA as
already described.
Western blot analyses
Adherent cells were washed twice with PBS and lysed in ice-cold lysis buffer (20 mM Tris-HCl, pH 7.5, 5 mM EDTA, 5 mM EGTA, 10 µg/ml leupeptin, 20 µg/ml aprotinin, and 1 mM PMSF) for 15 min. Cells were scraped with a rubber policeman and centrifuged at 4°C for 10 min at 10,000 x g. The protein content was determined using a microbicinchinoninic acid assay (Pierce, Rockford, IL) using BSA as standard. Cell lysates (40–50 µg of protein) were mixed with Laemmli reagent under reducing conditions. SDS-PAGE was performed according to standard techniques using 10% bis-acrylamide for the separation gel and transfer of proteins (38, 39). Nitrocellulose membranes were saturated for 2 h at room temperature in 5% fat-free dry milk-Tris buffer saline (50 mM Tris-HCl, pH 7.5, 250 mM NaCl, and 0.1% Tween 20) (TBTT). Membranes were further incubated at room temperature for 1 h with specific mAbs for COX-1 (1:1000), COX-2 (1:2000) (kindly provided by the late Dr. J. Maclouf, Institut National de la Santé et de la Recherche Médicale Unit 348, Paris, France), or ß actin (1:5000), or with polyclonal Abs directed against E-cadherin (1:1000). Blots were washed three times with TBTT buffer and then incubated with anti-rabbit or anti-mouse horseradish peroxidase-linked secondary Ab in TBTT and 5% fat-free dry milk. Chemiluminescent substrates were used to reveal positive bands visualized after exposure for 1–2 min to Hyperfilm ECL. In some samples, a shorter incubation period was employed to evaluate a large increase in COX-2 expression, resulting in apparent heterogeneity in density between figures. Protein bands were quantified using an LKB Ultrascan XL laser densitometer (Pharmacia). Immunodetection was linear between 15 and 75 µg of COX-1 and -2.
Measurement of PG production
PGs were determined in the supernatants of the cell cultures using previously validated RIA (40).
Preparation of thymocytes and coculture experiments
Thymocytes were isolated for coculture experiments from thymi of 4- to 5-wk-old mice by mechanical dissociation of the whole organ in culture medium. Thymocytes were then washed twice with HBSS, resuspended in a serum-free culture medium, and counted with trypan blue. Their viability was always >95%. Thymocytes were stained for CD4 and CD8 and analyzed by flow cytometry to assess the distribution of the different subsets.
Stromal cells were cultured to confluence on 12-well culture plates incubated overnight with indomethacin, L-759,700, NS-398, (0.1 µM), and increasing concentrations of PGE2, or vehicle. They were then washed twice with PBS, and 1 x 106 thymocytes were added to each well in 1 ml of serum-free culture medium and incubated for 3 h at 37°C under 5% CO2. Similarly, single cultures of thymocytes or stromal cells were incubated with the same inhibitors for the same length of time for use as controls. After 3 h of incubation, the nonadherent thymocytes were harvested by four sequential washes with serum-free culture medium (1 ml each wash), counted, and stained for flow cytometry analysis. Experiments were always performed in duplicate and were repeated at least four times.
Stromal cells and adherent thymocytes were harvested from the wells incubated with Versene (Life Technologies) for 10 min at 37°C. The cells were then collected and stained for CD4 and CD8 for qualitative analyses of thymocyte subsets adherent to stromal cell lines in coculture experiments. The number of thymocytes attached to the stromal cells was calculated by counting and analyzing the nonadherent cells by flow cytometry. To study the adhesion of DN cells to 427.1 and 1308.1 cell lines, we isolated thymocytes from RAG-1-/- mice. In these animals, T cell development is blocked at the DN stage because of a failure to rearrange the TCR-ß locus (41).
Flow cytometric analysis and cell sorting
Thymocytes were stained using anti-CD4 FITC and anti-CD8 Red 613 mAbs. Briefly, thymocytes were spun down at 1000 rpm for 5 min at 4°C and resuspended in staining medium (HBSS/5% FCS/0.01% sodium azide) with diluted Abs for CD4 and CD8 (1:200). After 30 min incubation at 4°C, cells were then washed and resuspended in PBS and analyzed with a FACScan (Becton Dickinson, San Jose, CA). Ungated data (30,000 events per sample) were collected and analyzed using FACScan CellQuest software (Becton Dickinson). Before analysis of Ab staining, samples were gated on live cells based on forward and side scatter parameters. The relative percentages of various thymocyte subpopulations based on CD4 and CD8 expression were obtained using CellQuest on the dot-plots of flow cytometry analyses. The absolute numbers of each thymocyte subset were calculated by multiplying the total cell number by the percentage of each thymocyte population. Cell sorting was performed on thymocytes from embryonic or adult thymi stained for CD4 and CD8 as indicated. The purity of the sorted populations of thymocytes was always >99%.
Adherent cells were harvested with Versene as described previously. Cells were washed twice with staining medium and incubated for 30 min with the primary Ab, diluted in staining medium, at 4°C (ICAM-1, CD40, and CD44 were used at 1:50 dilution, FITC-conjugated hyaluronate was used at 1:75 dilution). After two washes, the cells were incubated for 30 min with FITC-conjugated secondary Abs diluted in staining medium (1:100). The flow cytometry analyses were performed as previously described.
RT-PCR analyses
Total RNA from thymocytes isolated by cell sorting was extracted
with an RNAeasy minikit (Qiagen, Chatsworth, CA). First-strand
cDNAs were synthesized from 1 µg of total RNA using Moloney murine
leukemia virus reverse transcriptase and oligo(dT) primers (Boehringer
Mannheim Biochemicals, Indianapolis, IN). The mixtures were subjected
to PCRs with primers specific for COX-2 (sense,
5'-CTGTACAAGCAGTGGCAA-3', antisense, 5'-TTACAGCTCAGTTGAACGCCT-3')
spanning several intron-exon junctions. The expected DNA fragment was
530 bp. The PCR products were separated by electrophoresis through a
1.5% agarose gel containing ethidium bromide. The identity of the PCR
product was confirmed by Southern blotting; the filters were hybridized
using specific oligonucleotide (5'-CATTAACCCTACAGTACTAAT-3') after 5'
end labeling with [32P]ATP.
Immunocytochemistry
Cells were fixed in ice-cold 70% methanol/30% acetone, rehydrated, equilibrated in PBS buffer, pH 7.2, and stained. Briefly, specimens were incubated with 10% blocking serum diluted in PBS/1%BSA/0.02% sodium azide for 20 min. The primary Ab, either COX-1 antiserum (1:200) or COX-2 antiserum (1:400), were applied for 60 min. After washing (3 x 5 min) with PBS, the secondary FITC-labeled Ab was applied for 45 min. Sections were thoroughly washed, mounted in glycerol-gelatin, and inspected for green fluorescence under a Zeiss fluorescence microscope. Specificity of labeling was checked by appropriate controls, including single staining, omitting specific Ab, or replacing the first Ab with an irrelevant Ab. All controls gave negative results.
Statistical analysis
The results were evaluated by using ANOVA with subsequent comparisons by Student’s t test for paired or nonpaired data, as appropriate. Statistical significance was defined as p < 0.05. Values are reported as means ± 1 SD. The IC50s were calculated using Biosoft-Dose software (Elsevier-Biosoft, Cambridge, U.K.).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Western blot analyses were performed to evaluate constitutive COX
isozyme expression in thymic stromal cell lines. COX-1 was detected in
all four cell lines, although the levels of expression, normalized for
ß-actin, appeared to differ. The highest level of COX-1 protein
expression was in the cortical cell lines 1308.1 and 427.1 (Fig. 1
a). COX-2 protein also
appeared to be constitutively expressed, although it was
differentially expressed across the four cell lines (Fig. 1b). In this case, the highest expression was in the
medullary line (6.1.1), while the lowest expression was in the cortical
1308.1 cells (Fig. 1b).
|
a), while
the reactivity of the 1308.1 line was very weak (data not shown). COX-2
immunoreactivity was particularly apparent in the perinuclear area
(Fig. 2a), consistent with observations in other cells (42).
By contrast, the 1308.1 and 427.1 lines, which were intensively
positive for COX-1 (Fig. 2b), exhibited a diffuse granular
pattern of positivity, suggesting predominant localization of this
isoform in the endoplasmic reticulum (42).
|
Expression of COX proteins in thymic stromal cell lines prompted
us to investigate COX activity under basal conditions and upon
stimulation with exogenous AA. We first measured the spontaneous
release of PGs from each cell line over 24 h. PGE2 was
the predominant PG secreted by all the cell lines (Table I), while thromboxane (Tx) B2
and PGF2
only accounted for <10% of the total
prostanoids. Additional to marked constitutive expression of COX-2,
both of the medullary 6.1.1 and 1307 lines released higher amounts of
PGE2 compared with 427.1 and 1308.1 cells (Table I).
Indomethacin (10 µM), a nonselective COX inhibitor, inhibited PG
secretion by >98%.
|
). Again, the
medullary cell lines (6.1.1 and 1307) generated greater amounts of
PGE2 from exogenous AA. Indomethacin (10 µM) again
inhibited >98% the production of all the PGs from each cell line.
To address the relative contribution of each isoform to
PGE2 production, cells were treated with increasing
concentrations of the selective COX-1 inhibitor, L-759,700 or of the
selective COX-2 inhibitor, NS-398 (43), and stimulated with 10 µM AA.
L-759,700 exhibits preference for COX-1 vs COX-2 in different cellular
systems (unpublished observations), while NS-398 is roughly
160-fold more potent as an inhibitor of COX-2 vs COX-1 (44). Both
L-759,700 and NS-398 dose-dependently inhibited PGE2
production, but with different potencies in each cell line. Thus,
L-759,700 was more potent in inhibiting PGE2 biosynthesis
from 1308.1 cells (IC50 3 nM), which predominantly express
COX-1, than in the 6.1.1 and 1307 lines (IC50s 185 nM and
146 nM, respectively), which predominantly express COX-2 (Fig. 3b). In contrast, NS-398 was
more potent in inhibiting PGE2 secretion from the cells
(6.1.1 and 1307) that predominantly express COX-2 (Fig. 3a).
427.1 cells, which expressed both isozymes, are inhibited with
intermediate potency by the selective inhibitors (Fig. 3, a
and b).
|
Differential regulation of COX expression and activity by IFN-
IFN- is an immunoregulatory lymphokine that is primarily
produced by T cells and affects T cell differentiation and maturation
(45). Therefore, in the thymus, stromal cells are physiologically
exposed to this cytokine. Indeed 1307, 6.1.1, 1308.1, and 427.1 cell
lines are IFN- responsive (35). For example, we have found that
IFN- (1000 IU/ml, for 24 h) up-regulates the expression of
ICAM-1 and CD44 in 1307 and 6.1.1 cell lines (data not shown), while
others have reported the up-regulation of many other different
molecules, including MHC class II Ags, in all these thymic stromal cell
lines (35). In the present experiments, each cell line was treated with
different concentrations of IFN-, and the levels of protein
expression for each COX isozyme were investigated. We also determined
whether changes in COX protein levels caused by IFN- were associated
with changes in COX activity, measured as PGE2 production
after addition of 10 µM AA.
IFN- up-regulated COX-2 in both 6.1.1 and 1307 medullary
stromal cell lines. COX-2 up-regulation was time-dependent,
starting 3 h after addition of the cytokine and reaching
maximal expression between 12 and 24 h later (Fig. 4). No changes were observed in the
levels of COX-1 protein expression under the same conditions (Fig. 4).
COX-2 up-regulation was also concentration-dependent (Fig. 5, a and b) and was
more prominent in the 6.1.1 compared with 1307 line (Fig. 5, a and b). PGE2 production was also
dose-dependently increased by IFN- (Fig. 5, a and
b). Consistent with COX-2 protein expression, the increase
in PGE2 production was more pronounced in the 6.1.1 line
than in the 1307 line (Fig. 5, a and b).
|
|
slightly down-regulated COX-2 (Fig. 5, c and d), while it did not modify COX-1 protein
expression (Fig. 6) in the 427.1 and
1308.1 cortical stromal cell lines. Interestingly, we observed a
significant reduction in PGE2 production (Fig. 5, c and d) that was greater than expected based on
COX protein expression. In fact, inhibitor studies (Fig. 3) suggest
that COX-2 isoform contributes roughly 50% and 30% to the
PGE2 production in the 427.1 and 1308.1 cells,
respectively, while the reduction of PGE2 production was
>50% in both cell lines. To rule out an effect of IFN- on specific
PG synthases, we measured also Tx production (as TxB2)
under the same experimental conditions. Inhibition of Tx was similar to
that of PGE2 (data not shown). This indicated that IFN-
acts indeed at the level of COX activity rather than at the level of
specific PG synthases.
|
up-regulated COX-2 protein expression and activity
in the 1307 and 6.1.1 medullary cell lines, although to variable
degrees. This was not observed in the cortical cell lines, where
IFN- appeared to reduce COX-2 expression. Impairment of adhesion of thymocyte to thymic stromal cells by COX-2 inhibitors
The interaction of thymocytes with stromal cells is believed to be
necessary for development of the T cell lineage (46). Thymocytes,
freshly isolated from young adult (4–5 wk) mouse thymus, were
cocultured with each of the four stromal cell lines. Preliminary
experiments were performed to identify qualitatively the thymocyte
subsets that adhered to each stromal cell line. More than 80% of the
thymocytes adherent to the cortical 427.1 cells were phenotipically
DNs, while 
80% of thymocytes adherent to the cortical 1308.1 line
were DPs and the remaining 20% were DNs. The vast majority of
thymocytes adherent to the medullary 1307 and 6.1.1 lines (80%)
were DPs. In addition, CD4 SP thymocytes bind to 6.1.1 cells. The
differential capacity of these stromal cell lines to bind different
thymocyte subsets is consistent with their origins. Both 427.1 and
1308.1 lines derive from the cortex, where the majority of resident
thymocytes are DN or DPs, while medullary cells, such as 6.1.1 and
1307, would be expected to reflect interactions with DPs (at the
corticomedullary junction) and CD4 or CD8 SP lymphocytes, which reside
in the medulla.
Next, we studied whether COX-1 or -2 modulated the adhesiveness of thymocyte subpopulations to stromal cells. Neither indomethacin, NS-398, nor L-759,700 influenced the adhesion of DN thymocytes to the 1308.1 and 427.1 cortical cell lines (data not shown). The adherence of DP cells to 1308.1 was also unaffected by the COX inhibitors (data not shown).
In contrast to our observations with the cortical stromal cell lines,
NS-398 dose-dependently inhibited the adhesion of DP cells to the
medullary 1307 and 6.1.1 cells (Fig. 7).
The selective COX-2 inhibitor, NS-398, also impaired the adhesion of
CD4 SP cells to 6.1.1 cells (Fig. 7). Indomethacin had the same effect
(data not shown), while the selective COX-1 inhibitor L759,700 at doses
up to 0.1 µM was ineffective (data not shown). Inhibition of
adherence by NS-398 was dose-dependently rescued by adding exogenous
PGE2 to the cocultures (Fig. 7). NS-398 and indomethacin,
under these experimental conditions, did not affect the counts of the
adherent stromal cells and did not have any effect on either the counts
or the phenotypic distribution of the thymocytes alone. Therefore, the
reduction of adherent thymocytes cannot be attributed to a possible
reduction in the number of stromal cells. Moreover, expression of COX-2
m-RNA in thymocyte subsets (DN, DP, 4SP) was excluded by RT-PCR
(data not shown). Thus, the effects of NS-398 are attributable to
inhibition of COX-2 in stromal cells (Fig. 1b).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Our previous studies have shown the presence of both COX isozymes in the embryonic murine thymus. Both MHC class II-positive stromal cells in embryonic thymus and medullary stroma in adult thymus stain positively for COX-2 (32), raising the possibility that PG formation might represent a signaling mechanism(s) between stroma and developing thymocytes.
The thymic stroma is a complex network of phenotypically distinct cells, such as cortical and medullary epithelial cells, nurse cells, nonepithelial dendritic cells, macrophages, and fibroblasts (34). This heterogeneity seems likely to reflect functional specialization (33, 34). For example, cortical epithelial cells appear to support positive selection of DP thymocytes, and the medullary stroma is involved in further nurturing, maturation, induction of self-tolerance, and circulation of CD4 and CD8 SP lymphocytes (33, 34).
Given these observations, we investigated the expression, activity, regulation, and function of COX isozymes in four different thymic stromal cell lines that are representative of thymic subcapsular cortex nurse cells (427.1), cortical reticular cells (1308.1), medullary interdigitating-like cells (6.1.1), and medullary epithelial cells (1307) (35).
COX-2 appears to be the predominant COX isoform, in terms of protein
expression and activity in both the medullary thymic stromal cell lines
6.1.1 and 1307. Selective COX inhibitors were used to demonstrate that
production of PGE2, the predominant PG formed, derives
largely from COX-2 in these medullary stromal cells. These findings are
consistent with our observation that COX-2 is the predominant isozyme
in the medullary stroma of mouse adult thymus by immunohistochemistry
(32). Both the expression and activity of COX-2 appear to be regulated
in the medullary cell lines. Thus, IFN- dose-dependently increased
COX-2 protein expression and PGE2 production in the 6.1.1
and 1307 cells, albeit with different potencies (Fig. 5). Furthermore,
pharmacological inhibition of COX-2 impaired the adhesion of DP and CD4
SP thymocytes to these medullary stromal cell lines (Fig. 7). Addition
of exogenous PGE2 to the cocultures rescued this effect of
the selective COX-2 inhibitors. Interestingly, transgenic mice that
overexpress IFN- exhibit an increase in the number of CD4 and CD8 SP
lymphocytes in their thymi. This is associated with a reduction in
circulating lymphocytes, reflective of a disruption in the normal
processes that govern release of mature lymphocytes into the
bloodstream (48). In the present study, IFN- up-regulated COX-2 in
medullary cells and COX-2-dependent PGE2 production
enhanced the adhesion of CD4 SP lymphocytes to them.
COX-2-dependent PG formation has previously been implicated in the regulation of adhesive processes involving nonthymic cells. For example, COX-2 overexpression increases the adhesion of epithelial cells to extracellular matrix molecules (49). It has also been reported that COX-2 induction (50) or overexpression are associated with up-regulation of adhesion molecules in epithelial cells (49). Because of these observations, we examined the effects of selective COX-2 inhibition on expression of some adhesion molecules and extracellular matrix receptors by the thymic stromal medullary 6.1.1 and 1307 cell lines. Despite the role of COX-2-dependent PGE2 formation in modulating the adhesive interactions of these cells with thymocytes, we failed to detect any variation in the stromal expression of ICAM-1, E-cadherin, CD40, CD44, or hyaluronate binding proteins in the presence of the selective COX-2 inhibitor, NS-398, or of the nonselective inhibitor, indomethacin. Nevertheless, we and others have found that various subtypes of PGE2 receptors are expressed on isolated thymocytes at different stages of their development (23 and our unpublished observations). Therefore, it is possible that COX-2-dependent PGE2 production from stromal cells acts on neighboring thymocytes to enhance their adhesiveness. It has been already shown that PGE2 and cAMP both increase the binding of immature DN and DP thymocytes to extracellular matrix proteins (51). Similar mechanism(s) may be involved in the adhesion of thymocytes to medullary stromal cells. It is also possible that PGE2 modulates the stromal synthesis of soluble chemoattractant factors that are known to be involved in the interaction of thymocytes with stromal cells (33). Therefore, it will be important to determine whether the spectrum of chemokines produced by stromal cells is affected by COX inhibitors. We have recently shown that COX-2-dependent PGE2 formation acting via the EP1 subtype appears important in the positive selection of the CD4 SP lymphocytes (32). Whether the role of PGE2 in modulating thymocyte interactions with medullary stromal cells is relevant to this process remains to be determined.
Expression and activity of the COX isozymes was markedly different in
the two cell lines that originated from the cortical stroma (1308.1 and
427.1 lines). Both expression and activity of COX-1 was more prominent
than COX-2 in these cells (Figs. 1 and 3). Furthermore, IFN- caused
a slight, dose-dependent down-regulation of COX-2 protein expression,
with a marked (>50%) reduction in COX activity in the cortical
stromal cell lines, while COX-1 protein expression was unchanged (
Figs. 4–6). Selective inhibition of COX-1 suggests that it accounts for
roughly 70% and 50% of the total COX activity in the 1308.1 and 427.1
cell lines, respectively. These observations raise the possibility that
IFN- down-regulates COX-1 activity without altering its expression
in cortical cells. A discrepancy between the effects of cytokines on
COX protein expression and activity has been previously reported in
other cells. For example, TNF- increased AA-induced PGE2
formation by human airway smooth muscle cells without inducing
expression of either COX-1 or -2 proteins (52). Similarly, nitric oxide
donors suppress COX activity, despite increasing COX-2 protein
expression in human endothelial cells (53). Again, selective
inhibitors of either COX isozyme failed to influence the attachment of
DN or DP thymocytes to the cortical cell lines, in contrast to our
observations with medullary stromal cells. Thus, the differential
regulation of the COX isozymes in cortical vs medullary cells suggests
that PG formation may serve to differentiate functionally these two
components in the thymic stroma.
We measured PG synthesis in stromal cells in the absence or presence of
exogenous AA. We found no differences in the profile or the relative
amounts (PGE2 > PGF2 > TxA2)
of PG synthesized under the two experimental conditions in the four
cell lines. Furthermore, selective COX-1 or COX-2 inhibitors decreased
PGE2 production from exogenous or endogenous AA with
potencies that reflected the relative predominance of isoform
expression. Thus, the isozyme available, rather than the source of the
substrate, seems to determine which isozyme catalyzes PG formation in
stromal cells. Similar observations have been made in human endothelial
cells (54).
PGE2 is the most abundant PG synthesized by both medullary
and cortical thymic stromal cell lines. Lesser amounts of
PGF2 and TxB2 are synthesized by thymic
stromal cell lines. Indeed, both Tx synthase and thromboxane receptors
are present in human and murine thymus and thymocytes (21, 23, 24).
Mouse thymi depleted of hemopoietic cells with 2'-deoxyguanosine
produce PGE2, PGF2, and TxB2,
presumably from the residual thymic epithelium (29).
These observations prompt interest in how deletion of genes encoding COX isozymes or EPs might affect central (thymus) and/or peripheral immune function. Although little information is presently available, fetal thymocyte maturation is altered in COX isozyme deficiency, and adult COX-2-deficient mice are susceptible to spontaneous peritonitis (32, 55). Furthermore, there are reports that isozyme nonspecific COX inhibitors interfere with the induction of thymic tolerance in vivo in some rodent models (56, 57).
In summary, we report that COX-1 expression and activity appears to
predominate in cells that originate from cortical thymic stroma. By
contrast, COX-2 predominates in thymic medullary stromal cell lines.
The lymphokine, IFN-, differentially modulates the expression and
activity of COX-2, rather than COX-1, in medullary stromal cell lines,
and COX-2-dependent PGE2 modulates adhesion of thymocytes
to those cells. PG formation by stromal cells may contribute to the
compartmentalized regulation of T cell development and function in the
thymus.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Abbreviations used in this paper: AA, arachidonic acid; COX, cyclooxygenase; DN, CD4-CD8- double-negative thymocytes; DP, CD4+CD8+ double-positive thymocytes; SP, single-positive thymocytes; TX, thromboxane; EP, prostaglandin E receptors.
Received for publication November 13, 1998. Accepted for publication January 21, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
gene expression at multiple levels. J. Immunol. 148:2845.[Abstract]
and IL-2, but not IL-4 and IL-5. J. Immunol. 155:4604.[Abstract]
-induced ICAM-1 and VCAM-1 expression and T lymphocyte adhesion. J. Immunol. 154:2358.[Abstract]
in human urine. J. Endocrinol. Invest. 2:173.[Medline]
results in severe B-cell lineage reduction, T-cell lineage alterations, and hematopoietic progenitor deficiencies. Blood 89:583. and interferon- on the induction of cyclo-oxygenase-2 in cultured human airway smooth muscle cells. Br. J. Pharmacol. 121:579.[Medline]
This article has been cited by other articles:
|
|
 |
|
  E. A. Jaimes, M.-S. Zhou, D. D. Pearse, L. Puzis, and L. Raij Upregulation of cortical COX-2 in salt-sensitive hypertension: role of angiotensin II and reactive oxygen species Am J Physiol Renal Physiol, February 1, 2008; 294(2): F385 - F392. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Testa, B. Rocca, L. Spath, F. O. Ranelletti, G. Petrucci, G. Ciabattoni, F. Naro, S. Schiaffino, M. Volpe, and C. Reggiani Expression and activity of cyclooxygenase isoforms in skeletal muscles and myocardium of humans and rodents J Appl Physiol, October 1, 2007; 103(4): 1412 - 1418. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Braun, R. S. Longman, and M. L. Albert A two-step induction of indoleamine 2,3 dioxygenase (IDO) activity during dendritic-cell maturation Blood, October 1, 2005; 106(7): 2375 - 2381. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. N. Rocha, T. J. Plumb, L. A. Robinson, R. Spurney, D. Pisetsky, B. H. Koller, and T. M. Coffman Role of Thromboxane A2 in the Induction of Apoptosis of Immature Thymocytes by Lipopolysaccharide Clin. Vaccine Immunol., August 1, 2005; 12(8): 896 - 903. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Stokes Peebles Jr., K. Hashimoto, J. D. Morrow, R. Dworski, R. D. Collins, Y. Hashimoto, J. W. Christman, K.-H. Kang, K. Jarzecka, J. Furlong, et al. Selective Cyclooxygenase-1 and -2 Inhibitors Each Increase Allergic Inflammation and Airway Hyperresponsiveness in Mice Am. J. Respir. Crit. Care Med., April 15, 2002; 165(8): 1154 - 1160. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. B. Clark The role of PPARs in inflammation and immunity J. Leukoc. Biol., March 1, 2002; 71(3): 388 - 400. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. L. Clark, Z. Zeng, A. L. Langford, S. M. Bowen, and S. C. Todd PGRL Is a Major CD81-Associated Protein on Lymphocytes and Distinguishes a New Family of Cell Surface Proteins J. Immunol., November 1, 2001; 167(9): 5115 - 5121. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Solle, J. Labasi, D. G. Perregaux, E. Stam, N. Petrushova, B. H. Koller, R. J. Griffiths, and C. A. Gabel Altered Cytokine Production in Mice Lacking P2X7 Receptors J. Biol. Chem., January 5, 2001; 276(1): 125 - 132. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
), 1307 (
), and
6.1.1 (•) cell lines were incubated with the selective COX-2
inhibitor, NS-398, before AA incubations. PGE2 production
was measured, and the percent of inhibition of PGE2
production (expressed in ng/106 cells) is shown in the
plot. Results are means of two to four determinations for each
concentration. b, The same experiments were performed in
the presence of the selective COX-1 inhibitor, L-759,700.
