| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Research Communications |
* Department of Medical Nutrition, Karolinska Institutet, Huddinge University Hospital, Novum F-60, SE-141 86 Huddinge, Sweden; and
Microbiology and Tumor Biology Center (MTC) and
Department of Medical Biochemistry and Biophysics (MBB), Karolinska Institutet, SE-171 77 Stockholm, Sweden
1Correspondence: Department of Medical Nutrition, Karolinska Institutet, Huddinge University Hospital, Novum F-60, S-141 86 Huddinge, Sweden. E-mail:Sam.Okret{at}mednut.ki.se
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: steroid biosynthesis • thymocyte development • cytochrome P450 • reporter cells • paracrine signaling
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Double positive CD4/CD8 thymocytes
with nonreactive or nonfunctional TCRs undergo apoptosis due to the
lack of a TCR-mediated protective signal (default depletion due to
neglect). This process represents the most extensive loss of thymocytes
in the gland (2,
3)
. The thymus also eliminates self-reactive
thymocytes, which carry TCRs that strongly interact with endogenous
self-peptides presented by MHC (negative selection) (3)
. In contrast,
thymocytes bearing TCRs with low to moderate avidity for
self-MHC/peptide complexes generate a protective signal that rescues
the thymocytes from apoptosis (positive selection) (4,
5)
. The
TCR-dependent positive selection process occurs in the cortical thymus
through an intimate contact between developing thymocytes and TEC,
whereas the negative selection process occurs in both the cortex and
medulla, mostly on bone marrow-derived antigen-presenting cells (6
7
8)
.
It is generally thought that the outcome of the thymocyte
differentiation process (apoptosis or survival) is dependent on the
strength of TCR-self-MHC/peptide reaction. However, it is not clear how
these differences determine this outcome, although several
investigators have suggested that glucocorticoids (GCs) may play a role
in this selection process (9
10
11
12
13
14)
.
Thymocytes, particularly the immature double positive CD4/CD8 cells,
are known to be highly sensitive to GCs, which readily induce apoptosis
in these cells both in vitro and in vivo (15,
16)
. As the thymocytes are killed by apoptosis during the normal
thymocyte development (3)
, we previously investigated whether GCs may
be involved in the T cell differentiation process. We and others
(9
10
11,
17,
18)
have found evidence for this premise in studies of
thymic apoptosis induced by anti-CD3 monoclonal antibody in normal mice
(14)
and by peptide in TCR transgenic mice (12)
. Most important, we
have found that the glucocorticoid receptor (GR) hormone antagonist
RU486 inhibits thymic apoptosis in response to anti-CD3 and apoptosis
induced by injection of a MHC-II binding peptide to peptide specific
TCR transgenic mice (12)
. The protective effect by RU486 occurred both
in vivo in adrenalectomized mice and in isolated thymic
organ culture, excluding the possibility that RU486 antagonized GCs
secreted from the adrenal gland. The protective effect by RU486 could
be explained by the inactivation of GR activated by a GC-independent
way or, alternatively, by protection from GC(s) produced in the thymus,
as earlier suggested (17)
.
GC hormones are synthesized in the adrenal glands from the precursor
cholesterol through sequential conversions by members of the cytochrome
P450 superfamily of oxidases and 3ß-hydroxysteroid dehydrogenase
(3ßHSD) (19)
. Corticosterone is the main GC formed in rodents
(Fig. 1
). Its synthesis and release from the adrenals is under the control of
the hypothalamus-pituitary-adrenal (HPA) axis, adrenocorticotropic
hormone (ACTH) being the major mediator acting on the adrenals. When
GCs are released into the bloodstream after ACTH stimulation, they
distribute throughout the body, acting in an endocrine fashion on most
tissues controlling e.g., metabolism, immune responses, and
developmental processes. In addition to this endocrine signaling,
locally produced steroids have been implicated in paracrine signaling.
So-called `neurosteroids' formed in the brain have been suggested to
play a role both in cognitive functions and survival of cells in the
nervous system (20
21
22)
. Furthermore, all enzymes required for the
production of the major steroid hormones have been detected in the
brain (23,
24)
. Locally produced GCs may also play a role in T cell
development (10,
17)
. The thymus has endocrine features, e.g., it
expresses some mediators of the HPA axis like corticotropin-releasing
hormone (CRH) and ACTH (25
26
27)
.
|
GCs exert their effects via binding to the GR, a member of the nuclear
receptor superfamily of intracellular receptors (28,
29)
. On binding,
the hormone-receptor complex undergoes a process, called
transformation, which includes dissociation of associated proteins and
a conformational change. This permits the complex to interact with
specific DNA sequences, termed glucocorticoid response elements (GREs),
in target genes, which most often lie in the promoter region of
regulated genes. Once bound to the GREs, the GR modulates (induces or
represses) the activity of the target promoters. In some cases the GR
modulates the activity of target genes not by directly interacting with
GREs itself, but rather by binding to other transcription factors (28)
.
In the present work we have investigated the capacity of cells present in the thymus to produce GC hormone activity. We have found expression of several steroidogenic cytochrome P450's in TEC. Furthermore, coculturing of TEC and `reporter cells' containing GR and a GR-dependent reporter gene provided strong evidence that TEC produce a biologically active GC activity. This locally produced GC activity may contribute to the T cell differentiation at several different levels.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Irradiated thymi were cut into small pieces
by using extra-fine tip scissors, washed in phosphate-buffered saline
(PBS), and put into Eppendorf tubes. Adrenal and muscle tissues were
treated the same way. Isolated thymocyte single-cell suspensions were
prepared by passage through a steel net. The thymocytes were washed in
PBS and counted in trypan blue solution.
Reagents and plasmids
RU486 was kindly supplied by Roussel-UCLAF, Center for Research
(Romainville, France); rhodamine 123 was purchased from Sigma (St.
Louis, Mo.). Metyrapone (2-methyl-1, 2-di-3-pyridyl-1-propanone) was
bought from Aldrich Chem. Co. (Milwaukee, Wis.) and trilostane
(4,5-epoxy-17-hydroxy-3-oxoandrostane-2-carbonitrile) was donated by
Sanofi Winthrop (Newcastle, U.K.).
The reporter plasmid (GRE)2-tk-LUC, containing
two GREs upstream of the heterologous thymidine kinase (tk) promoter
and the luciferase reporter gene, has been described (30)
. The tk-LUC
reporter gene lacks the two GREs. The GR expression vector used was
SVGR1 (31)
.
RNA isolation and reverse transcription-polymerase chain reaction
(RT-PCR)
Total RNA was isolated from the tissues according to the method
of Chomczynski and Sacchi (32)
by homogenizing the tissue in 4 M
guanidinium thiocyanate containing 25 mM sodium citrate pH 7.0, 0.5%
(v/v) lauroyl sarcosine, and 0.1 M 2-mercaptoethanol, using a
glass-to-glass homogenizer. Extreme care was taken to avoid
cross-contamination between the samples. The quality of the RNA was
examined by agarose gel electrophoresis after staining with ethidium
bromide. RNA samples were stored as ethanol precipitates at -80°C.
First strand cDNA synthesis was carried out by incubating 2 µg of
total RNA with 15 µM of the 3'-oligo primer (see below) in the
presence of 50 mM Tris-HCl pH 8.3, 75 mM KCl, 3 mM
MgCl2, 10 mM DTT, 1 mM dNTPs, and 200 U of
reverse transcriptase (Superscript RT, Life Technologies, Inc.,
Paisley, U.K.) for 1.5 h at 42°C in a total volume of 20 µl.
The reaction was terminated by heating for 5 min at 95°C. Five
microliters of this reaction were taken to a 20 µl PCR amplification
reaction mixture containing 15 µM of the 5'-oligo primer (see below)
in the presence of 10 mM Tris-HCl pH 9.0, 50 mM KCl, 1.5 mM
MgCl2, 0.1% (v/v) Triton X-100, and 2 mM dNTPs.
After denaturing at 95°C for 2 min and the addition of 2 U of
Taq DNA polymerase (Promega, Madison, Wis.), PCR
amplification was performed by 30 cycles of 95°C for 1 min, 55°C
for 1 min, and 72°C for 1.5 min A parallel reaction was carried out
in the absence of reverse transcriptase to control for contamination
with DNA. The primers used for reverse transcription and PCR
amplification were as described previously by Stromstedt and Waterman
(24)
(all primers are given 5'
3'):
Cyp11A1
5'primer: CAA CAT CAC AGA GAT GCT GGC AGG
3'primer: CTC AGG CAT CAG GAT GAG GTT GAA
Length of product: 583bp
Cyp21:
5'primer: CTT CAC GAC TGT GTC CAG GAC TTG
3'primer: CAG CAG AGT GAA GGC CTG CAG CAG
Length of product: 775 bp
Cyp11B1:
5'primer: TCA CCA AAT GTA TCA AGA ATG TGT
3'primer. CCA TCT GCA CAT CCT CTT TCT CTT
Length of product: 666bp
Southern blot analysis
Agarose gel electrophoresis of the PCR products and Southern
blotting was performed as described (24)
. The oligonucleotides used as
probes for detection of PCR products were as follows:
Cyp11A1: GGT GGA GTC TCA GTG TCT CCT TGA TGC TGG CTT TGA G
Cyp21: AAG GCC AAG GGC ACC ACA GGC CGC AAA CGC AGC ACC T
Cyp11B1: CCT GCT GAA CAT CTG GGT TCC GAG CCA GCT CAA AAA G
The oligonucleotides were end-labeled by T4 polynucleotide kinase using
[ 
-32P]ATP (3000 Ci/mmol, Amersham, Little
Chalfont, U.K.).
Western blot analysis
The tissues were homogenized in PBS, 20% (v/v) glycerol, 1 mM
EDTA, and 0.1 mM phenylmethylsulfonyl fluoride by using a
glass-to-glass homogenizer. The mitochondrial fraction was enriched
according to Haaparanta et al. (33)
. Protein concentrations were
determined by the Bradford assay (34)
and the mitochondrial fraction
was mixed with one volume of 2 x sodium dodecyl sulfate (SDS)
buffer (100 mM Tris-HCl, pH 6.8, 0.2 M dithiothreitol, 4% SDS, 0.2%
bromphenol blue, and 20% (v/v) glycerol); the indicated amount of
protein was separated by 10% SDS-polyacrylamide gel electrophoresis
(SDS-PAGE) and electroblotted to a C-Extra Hybond membrane (Amersham).
Immunodetection was carried out by using the enhanced chemiluminescence
detection kit from Amersham after incubation with the anti-Cyp11B1
antibody (35)
at a dilution of 1:200, followed by a secondary
horseradish peroxidase-labeled donkey anti-rabbit antibody (Amersham)
according to the suggested protocols.
Cell culture and transactivation experiments
COS-7 cells (obtained from American Tissue Culture Collection,
Rockville, Md.) were grown in a 1:1 mixture of Dulbecco's modified
Eagle's medium and Ham's F-12 medium supplemented with 10% fetal
bovine serum, 2 mM L-glutamine, 100 IU/ml penicillin, and 100 µg/ml
streptomycin at 37°C in 5% CO2. Cells were
plated in 10 cm dishes at a density of 1 x
106 cells/plate 24 h before transfection.
Cells were transiently transfected by using the calcium phosphate
precipitation technique (36)
with 5 µg of the reporter gene
[(GRE)2-tk-LUC or tk-LUC] and 1 µg of the
GR-expressing vector SVGR1. After overnight exposure of the cells to
the DNA-calcium phosphate mixture, the cells were washed and allowed to
recover for 5 h. Transfected COS-7 cells (reporter cells) were
trypsinized, washed, and pooled from several plates. Two irradiated
thymi (TEC) or two adrenals in an Eppendorf tube containing 0.5 ml
culture medium were covered by 1 x 106 of
pooled reporter cells in 0.5 ml of cell culture medium by a 2 min spin
at 150 g. In some cases, an equivalent size of a piece of
intact thymus, spleen, or skeletal muscle was used instead of the TEC.
The addition of an equal number of transfected COS-7 cells from one
pool avoided alterations in luciferase activity due to variation in
transfection efficiency between the plates. If indicated, incubations
were performed in the presence of 30 nM ACTH, 1 µM RU486, 50 nM
trilostane, or 150 µg/ml metyrapone. In these cases, TEC were
pretreated with the drug for 1 h before the addition of the COS-7
cells. After overnight incubation at 37°C in 5%
CO2, cells were lysed and luciferase activity was
determined. Calculation of luciferase activity was based on the total
luciferase activity per tube, thus making it dependent only on added
COS-7 reporter cells.
TEC-induced thymocyte apoptosis
TEC and thymocyte single-cell suspension were prepared as
described above. TEC from one thymus or a piece of muscle of equivalent
size were added to an Eppendorf tube and covered by 1 x
106 thymocytes in 1 ml of culture medium. When
indicated, incubations were performed in the presence of 1 µM RU486,
50 nM trilostane, or 150 µg/ml metyrapone. In these cases, the drug
was added to the TEC 1 h before the addition of the thymocytes.
After 16 h of incubation at 37°C, thymocytes were stained with
2.5 mM rhodamine 123 for 60 min (37)
. Cells were then washed with cold
PBS and 10,000 gated cells were directly analyzed using a FACScan
(Becton Dickinson, Rutherford, N.J.).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The first and rate-limiting step in
corticosterone synthesis is the conversion of cholesterol to
pregnenolone (19)
. This reaction is catalyzed by the cytochrome P450
side chain cleavage enzyme Cyp11A1 (19)
. Using RT-PCR, we detected
Cyp11A1 mRNA in total RNA extracts prepared from intact thymi
(Fig. 2
A). Although not quantitative, the Cyp11A1 signal from thymus
was less intense than the Cyp11A1 signal from adrenal, indicating that
the Cyp11A1 expression level is lower in the thymus. Cyp11A1 expression
was also found in the lung (Fig. 2A
). When the thymus was
split into the thymocyte and TEC compartments, it was clear that
Cyp11A1 expression was confined to the TEC compartment (Fig. 2A
). In contrast, no Cyp11A1 expression was found in a TEC
cell line (38)
. Also, the cytochrome P450 21-hydroxylase (Cyp21)
and the P450 11ß-hydroxylase (Cyp11B1) mRNAs were detected in the
mouse thymus (Fig. 2B
and C, respectively). In
both cases, expression was restricted to the TEC compartment, with no
evidence for expression in the thymocytes. In contrast to Cyp11A1, both
Cyp21 and Cyp11B1 were detected in the TEC cell line. Cyp21 was also
detected in the lung, whereas Cyp11B1 was detected in the liver but not
in the lung or in the spleen.
|
To investigate whether the mRNA expression also resulted in protein
expression, a Western blot analysis of Cyp11B1 expression was
performed. As can be seen from Fig. 3
, a weak Cyp11B1 expression was detected in the thymus. In contrast, no
Cyp11B1 was detected in the muscle. The Cyp11B1 signal in the thymus
was at least 50-fold less intense than the signal from the adrenal.
Taking into account that the amount of thymic mitochondrial protein
applied was 25-fold more, the expression of Cyp11B1 in the thymus is at
least 1000-fold lower than the expression in the adrenal.
|
TEC secrete a factor that activates a GC-regulated reporter gene in
cocultured cells
To test whether TEC not only express the required enzymes for
corticosterone production but indeed produce a biologically active GC
activity, TEC were cocultured with COS-7 cells that had been
transfected with a GR expression vector and GRE-luciferase reporter
gene. This coculturing resulted in a 2.3 ± 0.3-fold induction
(n=9) of luciferase activity (Fig. 4
). This induction could be compared with the sevenfold increase observed
when the reporter cells were incubated with adrenals (Fig. 4
, insert).
The luciferase induction induced by the coculturing with TEC was
significant (P<0.001). No induction of luciferase activity
was observed when the reporter cells were incubated with muscle,
spleen, or intact thymus (Fig. 4)
. We observed a slight but not
significant enhancement of luciferase activity when ACTH was added to
the TEC during the coculturing.
|
The luciferase induction was totally blocked when the GC antagonist
RU486 was added to the incubation (Fig. 4)
. Furthermore, the induction
of the reporter gene activity required the presence of the GR in the
reporter cells, since no induction was observed in experiments in which
the receptor was omitted from the COS-7 cells. In addition, the
presence of the GREs in the reporter gene was also required, because no
induction was seen when the reporter gene lacked the GREs. Finally,
inclusion in the coculture of the 3ßHSD inhibitor trilostane or the
Cyp11B1 inhibitor metyrapone prevented the induction of the reporter
gene (Fig. 4)
. The reason for a low basal luciferase activity generated
when COS-7 cells were incubated with trilostane or metyrapone is
unclear, but is not due to a toxic effect on the COS-7 reporter cells.
In addition, activation of the reporter gene in the COS-7 cells by
dexamethasone is unaffected by the concentrations of trilostane or
metyrapone used (data not shown). In summary, the results demonstrated
that the induction is a true GR-mediated response due to the production
of a GC hormone activity, most likely a GC, in the TEC.
GC activity produced in the TEC contributes to thymocyte apoptosis
We have previously demonstrated that the GC antagonist RU486 can
partially prevent anti-CD3 or peptide-induced apoptosis in thymic organ
cultures of TCR transgenic mice (12)
. This may suggest that an
endogenously produced GC hormone activity cooperates with the anti-CD3
and peptide effect. Furthermore, fetal thymic organ cultures (39)
or
experiments in which GC sensitivity in the thymocytes has been changed
by repressing endogenous GR expression or overexpressing the GR have
demonstrated that GCs play a role in T cell development (18,
40
; A.
Pazirandeh et al., unpublished results). To test whether the GC
activity locally produced in the thymus can influence thymocyte
apoptosis, TEC were cocultured with freshly isolated thymocytes. As can
be seen in Fig. 5
, 23% of the thymocytes were alive after 16 h of coculturing with
TEC as compared with the number of living thymocytes incubated for the
same time in medium alone. In contrast, ~90% of the thymocytes were
alive when cocultured with a piece of skeletal muscle unable to produce
the GC activity (c.f. Fig. 4
above). The thymocyte population mainly
affected by the coculturing with the TEC was the
CD4+/CD8+ double positive
cells, as analyzed by FACS (data not shown). The inclusion of the GC
antagonist RU486 or the steroid synthesis inhibitors, trilostane and
metyrapone, all partially protected (P<0.05) against
TEC-induced apoptosis of the thymocytes (Fig. 5)
. The percentage of
surviving thymocytes doubled from 23% to ~41–48% after treatment
with any of these agents. In the absence of TEC, RU486, trilostane, or
metyrapone by themselves in the concentrations used did not influence
thymocyte survival (data not shown). This demonstrates that the GC
activity produced locally in the thymus affect thymocyte survival.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Furthermore, the activity is
blocked by the GR antagonist RU486, which competes with GC agonists for
binding to the same ligand binding pocket in the GR. Although we have
not so far established the exact chemical nature of the effector
molecule, several experimental results suggest that the TEC produce and
secrete GC hormone(s) that can influence the surrounding thymocytes.
The RT-PCR demonstrated the expression, albeit in low levels, of
mRNAs for the required hydroxylases involved in corticosterone
synthesis. The presence of 3ßHSD activity, the additional enzyme
activity required for corticosterone biosynthesis (Fig. 1)
, has been
demonstrated in separate studies. Thus, TEC, but not thymocytes, were
found to generate 20
-hydroxy-4-pregnen-3-one from added
pregnenolone, a conversion requiring participation of a 3ßHSD (J.
Zhang, Y. Xue, Yang, S. Okret, M. Jondal, and J. Sjövall,
unpublished results). Activation of the target gene in the coculture
assay is blocked by inhibitors of enzymes involved in steroidogenesis.
In addition, incubation of TEC with pregnenolone generated supernatants
from which material was extracted and separated by high-performance
liquid chromatography into an expected corticosterone fraction. This
fraction was found to induce apoptosis of thymocytes, inhibitable by
the GC antagonist RU486 (data not shown). Similar to Vaccio et al.
(17)
, we could detect some immunoreactivity using a radioimmunoassay
for deoxycorticosterone. However, the specificity of the assay was
interfered by the addition of the high concentration of the precursor.
Furthermore, the administration in vitro of a high
concentration of a precursor does not prove the production of GCs
in vivo when physiological levels of precursors are present.
Instead, we are trying to conclusively identify the active compound(s)
in the absence of precursor administration at the femtomol level by
using mass spectrometric methods.
Several experimental conditions used in the RT-PCR guaranteed that the
expression of the P450 steroid hydroxylases in the TEC population is a
true finding and not due to contamination of genomic DNA. First, no
expression was detected in the absence of reverse transcriptase;
second, primers used in the PCR were localized in different exons, thus
utilizing the expected cDNA size of the produced band as an internal
control. Finally, a probe independent of the PCR primers was used to
detect the expected product by Southern blotting. In addition, P450
11ß-hydroxylase (Cyp11B1), the final enzyme involved in the
biosynthesis of corticosterone, is detected by Western blotting. The
reason for the additional bands seen in the Southern blot after the
RT-PCR, particularly in the highly expressing tissues, is most likely a
conversion of PCR products to random-length, higher molecular weight
fragments when a high number of PCR cycles was used to detect the
products in the low-expressing tissues (41)
. It is important to note
that the expression of the cytochrome P450 steroid hydroxylases in TEC
was not a consequence of the irradiation of the mice, since expression
was also detected in intact thymi from nonirradiated mice. Furthermore,
despite the presence of enzymes involved in the classical biosynthetic
pathway of glucocorticoids, one also has to consider a possible
nonclassical biosynthesis pathway, as has been described for the
formation of some steroids in brain tumor cells (42)
.
The coculturing experiments between TEC and a reporter cell line
containing a GR-dependent reporter gene demonstrated that the TEC
secrete enough of GC hormone activity to influence a neighboring cell.
In contrast, other tissues like the spleen or muscle were unable to
induce the reporter gene. Thus, the induction of the reporter gene was
not a consequence of serum in the medium or blood contamination, since
all the tissues were in the same medium and some of the tissues (e.g.,
muscle and spleen) contained more blood than the TEC. When incubating a
piece of an intact thymus with the reporter cell line, no induction of
the reporter gene was seen. This may reflect that the TEC-secreted GC
activity only acts in a paracrine fashion and only cells in its close
proximity will be affected. This reflects a difficulty in general for
the analysis of paracrine systems where local concentrations can be
high but overall concentrations are very low. Thus, determination of
the concentration of the activity in medium or an extract will not
reveal the local concentration. As paracrine interactions are likely to
be important for the function of many tissues, a cocultured reporter
system, as used by us, is advantageous in revealing a physiologically
significant local hormone production. Such cocultured `reporter
systems' have become an important detection tool in studies of the
formation of low concentrations of bioactive molecules (43)
. That the
GC activity produced locally in the thymus indeed may have an important
role in T cell development is suggested by the demonstration that it
contributes to the apoptotic activity on thymocytes (Fig. 5)
. Of
interest is that incubation of a TEC cell line, in which we detected
some of the P450 steroid hydroxylases (Fig. 2)
, with RU486 protects
cocultured immature
CD4+8+3+
thymocytes from apoptosis (38)
. The reason for the incapability of
RU486, trilostane, or metyrapone to fully protect thymocyte apoptosis
in the TEC-thymocyte coculture assay is not clear, but could be due to
an inability of the metabolic inhibitors to fully block GC biosynthesis
(a nonclassical biosynthesis pathway, as discussed above), a partial
agonistic activity of RU486 (44)
, or other GC-independent mechanisms in
the TEC–thymocyte interaction may contribute to the degree of
thymocyte apoptosis (45)
.
Earlier reports support the idea that GCs are important for T cell
development in the thymus (9
10
11
12
13
14)
. A direct role for GCs in thymocyte
development has been formulated as the `mutual antagonism' model,
which states that GCs oppose apoptotic signals generated by medium
avidity TCR stimulation and lead to positive selection (9,
10)
.
Negative selection is explained as the inability of GCs to counteract a
strong TCR signal generated through high-avidity recognition of
self-MHC/peptide complexes. Evidence for this hypothesis originates
from experiments done in thymic tissue cultures and in GR antisense
transgenic mice (9,
10,
40)
. In GR antisense transgenic mice, which
have a reduced number of GRs in the thymocytes and thus an impaired GC
sensitivity, it was found that the size of the thymus was reduced,
mainly as a consequence of reduced number of CD4/CD8 double positive
thymocytes (40)
. However, experiments from another group as well as
ourselves using the same antisense GR transgene, under the control of a
different or the same promoter, found the opposite effect (18
; A.
Pazirandeh et al., unpublished results).
Our data support a model in which GCs, at a given concentration,
primarily have an apoptotic effect on immature thymocytes. Positive
selection would then be explained by protective TCR signals, possibly
associated with signals generated in the MAP kinase pathway, which
might down-regulate GR responsiveness (46,
47)
. Alternative signals,
generated by high-avidity TCR recognition of self-MHC/peptide
complexes, might have the opposite effect, e.g., to increase GC
responsiveness (13)
. To support this, we have found that the GC
antagonist RU486 inhibits thymic apoptosis induced by the OVA323–339
peptide in DO11.10 TCR transgenic mice (12)
. In the same DO11.10 mouse
strain, we also find a strong positive selection effect of
I-Ad, in combination with undefined endogenous
self-peptides, which is related to a relative GC resistance in the
transgenic thymocytes and to a reduced GR responsiveness in
vitro (unpublished results). In addition, when we increased GC
sensitivity in normal mouse thymus by overexpressing (twofold) the GR
under the control of the p56lck proximal
promoter, we found a reduced thymocyte number and a diminished number
of mature T cells in the periphery (unpublished results). The function
of GC activity in thymus could also be considered in terms of the
well-known regulatory effect of the HPA axis in the peripheral immune
system (11,
48)
. In thymic tissue, both CRH and ACTH are expressed (25,
26,
49)
and thus might be part of a local HPA axis potentially
important for regulating the size of the organ, as well as applying a
negative pressure during T cell differentiation. The strength of this
negative pressure might be important in shaping the T cell repertoire
and for the development of autoimmune diseases, as has been
demonstrated for the HPA axis in several animal autoimmune disease
models.
In summary, although the exact function of thymic GC exposure is equivocal, data clearly suggest that TEC do produce a biologically relevant GC activity that can affect surrounding thymocytes.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received for publication November 9, 1998.
Revision received January 4, 1999.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
5-
