| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
* Department of Pharmacology, University of Sassari Medical School, Sassari, Italy;
Neuropharmacology, OASI Institute for Research and Care on Mental Retardation and Brain Aging, Troina, Italy;
Molecular Cell Biology, Vrije Universiteit Medical Center, Van der Boechorststraat, Amsterdam, The Netherlands;
Centre Hospitalier de l’Université Laval Research Center and Department of Anatomy and Physiology, Laval University, Quebec, Canada;
¶ Department of Neuroinflammation, Division of Neurosciences, Imperial College School of Medicine, London, United Kingdom; and
|| Department of Immunobiology, Biomedical Primate Research Center, Rijswijk, The Netherlands
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Endogenous GCs are vital in determining EAE severity. Activation of the hypothalamic-pituitary-adrenocortical (HPA) axis during clinical disease is crucial for the recovery from EAE, since suppression of this response results in lethal EAE and renders resistant strains of animals susceptible to disease induction, while, conversely, treatment with exogenous GCs can block disease (6, 10, 11). Of major concern is that early life events, including neonatal exposure to GCs or pathogens, alter the developmental programming of the HPA axis and stress response as well as modulate the susceptibility to inflammatory and autoimmune diseases (12, 13, 14, 15, 16, 17).
GC receptors (GRs; type II), which are primarily involved in feedback
regulation of the HPA axis (1, 2), mediate the immune
regulatory effects of GCs at multiple levels (2). After
ligation, translocation of a GR complex to the nucleus and binding to
GC-responsive elements occurs, resulting in inhibition of the
transcription of a plethora of inflammatory mediators, such as IL-1,
IL-3, IL-4, IL-5, IL-6, IL-8, IL-12, TNF-
, and GM-CSF
(18). Among the many proteins regulated by GC is the
inducible NO synthase (iNOS or NOS2) gene. Suppression of this protein
decreases the production of NO, a key molecule in the inflammatory
response (19, 20, 21). An important molecular mechanism
underlying most GC anti-inflammatory and immunosuppressive activity
is inhibition of AP-1 (Jun-Fos heterodimers) and NF-
B (p65-p50
heterodimers) families of transcription factors (22, 23).
The cytokine-dependent stimulation of iNOS responsible for NO
production is mediated by NF-B activation and is suppressed by GCs
(24, 25).
Administration of synthetic GCs is used to control the symptoms of MS, although long-term therapy has failed to show a beneficial effect on either the progression of disability or the rate of relapse (26). The exact mechanisms of GC action in MS or EAE are unknown, although reduction of blood-brain barrier changes, decrease in inflammatory cells, suppression of cytokine production, altered adhesion molecule expression, and reduction of edema in MS lesions have been reported (26, 27).
Impaired GR function has been considered an important factor in the development and/or maintenance of various pathological conditions in humans, including (familial) GC resistance (28) and major depression (29). Long-lasting dysfunction/dysregulation of central GR expression are also known to accompany early neonatal experiences (12, 13) with potential consequences with respect to immune regulation (14, 15, 16, 17). A transgenic (Tg) mouse expressing antisense RNA directed against GR was created to serve as an animal model for the study of neuroendocrine changes occurring in stress-related disorders (30). These mice show reduced GR mRNA in the brain, pituitary, thymus, and spleen; reduced brain, thymic, and splenic GR binding; and reduced HPA axis sensitivity to GCs (30, 31, 32, 33). As a consequence of the impaired GR function, regulation of the HPA axis in these animals is disturbed (31, 32, 33, 34, 35). At the immunological level, GR-deficient mice show sex-dependent alterations in both trafficking and responsiveness of T lymphocytes during postnatal development of thymus and spleen, characterized by increased proliferative capacity in response to T- and B-dependent mitogens (31, 32, 36). In addition, a partial blockade of T cell differentiation and profound alterations of the thymic microenvironment, which persisted in adulthood, were observed, thus suggesting potential consequences for the programming of neuroendocrine-immune reactions in later life (36, 37).
We have used these Tg mice to verify the impact of a dysfunctional GR from early embryonic life on the susceptibility to EAE. Counter to the expected predictions, here we report that the long-lasting GR deficiency of Tg mice programs the resistance to EAE induction and that iNOS/NO has a chief role in this protection through NO-induced immunosuppression.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Female B6C3F1 (C57B female x C3H male) (H-2b) wild-type (Wt) and Tg (line 1.3) mice, in which a transgene, driven by a neurofilament promotor, was inserted in the genome constitutively expressing antisense RNA against the GR (30), were bred at the OASI Institute (Troina, Italy). The GR mRNA levels and GR binding capacity in the brain, pituitary, and immune organs of these Tg mice are decreased by 40–50% compared with those in Wt mice (30, 32).
Mice were housed five per cage in a temperature- (21–23°C), humidity- (60%), and light- (12-h light, 12-h dark cycle, lights on at 0600 h) controlled room. Food and tap water were available ad libitum. Studies were conducted in accordance with the Guide for the Care and Use of Laboratory Animals (National Institutes of Health) and were approved by the review boards of the OASI Institute (Troina, Italy).
Induction of EAE
EAE Wt and GR-deficient mice were immunized on days 0 and 7 s.c. in the flank with 200 µl emulsion containing 300 µg myelin oligodendrocyte glycoprotein (MOG) peptide aa 36–50 (EVGWYRSPFSRVVHL) (38, 39) added to PBS and emulsified in CFA supplemented with 500 µg heat-inactivated Mycobacterium tuberculosis (Difco, Detroit, MI) as reported previously (40). Wt and GR-deficient mice injected with vehicle (PBS) or CFA (Sigma-Aldrich, St. Louis, MO) alone served as sham-immunized and control immunized mice, respectively. Mice were also injected i.v. with 500 ng pertussis toxin derived from Bordetella pertussis (Sigma-Aldrich, Poole, U.K.) in 500 µl PBS in the tail vein immediately after and 48 h after Ag immunization. Mice were weighed before immunization and from day 8 were weighed and examined for neurological signs of EAE as follows: 0, no signs; 0.5, paresis (tail weakness) or partial paralysis; 1, complete tail paralysis; 2, paraparesis, hind limb weakness; 2.5, partial hind limb paralysis; 3, complete hind limb paralysis; 3.5, paraplegia; 4, quadriplegia; and 5, death.
Pathological examination
Five-micrometer wax sections of the CNS were stained with either H&E or Luxol Fast Blue/Cresyl Fast Violet for myelin evaluation (41).
Corticosterone (cort) measurement
Groups of 8–10 Wt and GR-deficient mice injected with PBS, CFA, or MOG36–50 in CFA were killed between 0800 and 1200 h at times after immunization. Great care was taken to keep the mice undisturbed the night before the experiment. Trunk blood was collected after decapitation, and plasma was stored at -80°C for hormones and nitrite assays. Cort was assayed using a specific RIA (ICN Biomedical, Costa Mesa, CA) (32). Results are represented as the mean ± SEM. No significant differences were observed between the control groups sacrificed at different time points after PBS injection, and thus results were pooled and expressed as preimmunization levels.
T cell proliferation assays
Splenocytes and lymph node cells (LNC) were isolated and processed as previously described (32, 42). Briefly, single-cell suspensions of splenocytes and LNC were collected from at least four individual mice per group, and 0.2 ml containing 2 x 106 cells/ml was cultured in 96-well microtiter plates (Costar, Cambridge, MA) in nutrient medium: RPMI 1640 medium (Life Technologies, Grand Island, NY) supplemented with L-glutamine, 2-ME, penicillin G (100 U/ml), streptomycin (100 µg/ml), and 10% heat-inactivated FCS. The cultures were incubated at 37°C in humidified 5% CO2/95% O2 with Con A (0.5–40 µg/ml), MOG36–50 (0.5–40 µg), or LPS (0.5–500 ng/ml). After 72 h, [3H]thymidine (25 Ci/mmol; 1 µCi/well; New England Nuclear, Boston, MA) was added, and the cultures were incubated for an additional 6 h. The cells were harvested, and the proliferation index was calculated as the mean counts per minute of triplicate wells containing Con A or MOG36–50 divided by the mean counts per minute of triplicate wells of cells only. For the time-course study, groups of four Wt and GR-deficient mice, immunized with MOG36–50 in CFA or CFA only, were sacrificed at 3, 7, 10, and 20 days, and spleen and LN cells were cultured in the presence or the absence of Con A or MOG36–50 (2.5 µg/ml). Wt and GR-deficient mice injected with PBS only were sacrificed at the same time points and used as controls.
Isolation and culture of peritoneal macrophages
On day 10 after immunization with
MOG36–50 peritoneal macrophages were isolated
from Wt and GR-deficient mice following i.p. injection of 10 ml RPMI
1640 medium, which was collected after gentle massage of the peritoneum
for 5 min (43, 44). The cell suspension was adjusted to
1 x 106 cells/well, and the adherent
macrophages were cultured in RPMI 1640 supplemented with 1 mM
L-glutamine, 0.5% HEPES, penicillin G (100 U/ml),
streptomycin (100 µg/ml), and 10% heat-inactivated FCS in 24-well
plates at 37°C. The macrophages were cultured for 24 h either
alone (basal level) or in the presence of 100 ng/ml LPS. The effect of
LPS activation was studied in the absence or the presence of the NO
synthase (NOS) inhibitors;
N
-nitro-L-arginine
methylester (L-NAME),
Ng-monomethyl-L-argine
(L-NMMA), the inactive enantiomer
Ng-monomethyl-D-arginine,
and the specific inhibitor of iNOS,
L-N6-(1-iminoethyl)-lysine
(L-NIL; 2 µg to 2.0 mg/ml; Sigma-Aldrich)
(45) or cort (Sigma-Aldrich) at
10-11–10-6 M. These
studies were also conducted in the absence or the presence of the GR
antagonist mifepristone, RU486
(17
-hydroxy-11-(4-dimethylaminophenyl)-17(propynil)-estra-4,9-dien-3-one),
at a concentration of 10-6 M (provided by
Roussel-UCLAF, Hoechst, Romainville, France). After 24 h
the supernatants were collected for nitrite analysis. The number of
macrophages was determined after harvesting by trypsinization in Ca-
and Mg-free PBS and washing in RPMI medium. Peritoneal macrophages from
sham-injected Wt and GR-deficient mice were used as controls.
Inhibition of iNOS
To evaluate the effects of macrophages on splenocyte or LNC proliferation, peritoneal macrophages from Wt and GR-deficient mice were isolated on day 10 following immunization with MOG36–50, and the cell suspension was added to lymphoid cell preparations according to the method described by Willenborg et al. (46). Briefly, cells were washed three times and added (1–4 x 104 cells in 100 µl) to LNC or spleen cell cultures (2 x 105cells in 100 µl) collected on day 10 of immunization. The cultures were incubated with 5 µg/ml MOG36–50, and cell proliferation was assessed as described above. The effect of antagonism of iNOS on the mixed cultures were conducted using L-NAME, L-NMMA, or L-NIL (2 µg to 2 mg/ml) in the presence of MOG36–50 (5 µg/ml), and after 3 days in culture proliferation was assessed. Supernatants were harvested and stored at -70°C until assayed.
Nitrite/nitrate assay
The production of NO, as measured by the formation of the stable decomposition product nitrite, was determined in cell-free supernatant and plasma (43, 44). Briefly 100 µl supernatant was mixed with 100 µl Griess reagent (1% sulfanilamide plus 0.1% N-1-naphtylethylenediamine in 2.5% H3PO4). For nitrite and nitrate studies, plasma aliquots were diluted 1/5 with phosphate buffer, and 0.6 ml was incubated at room temperature for 2 h with 0.05 ml nitrate reductase (1 U/ml; Sigma) and 0.05 ml NADPH (1.8 mM). Excess NADPH was oxidized by addition of 0.05 ml phenazine methosulfate (80 µM). The samples were deproteinized using 0.1 ml zinc acetate (0.5 M) and 0.1 ml NaOH (0.5 M), followed by centrifugation (1900 x g for 10 min). Nitrite levels were determined in the supernatant using Griess reagents. After 10 min at room temperature, the OD at 540 nm was measured. In parallel, a sodium nitrite/nitrate standard curve (1–200 µM) was generated. Samples were tested in triplicate, and results represent the mean ± SE of at least four to six mice per group.
Effect of macrophage-derived factors on lymphoid cell proliferation
Conditioned medium (CM) from encephalitogenic myelin basic protein-sensitized LNC cultures induces the expression of iNOS/NO in vitro (47). To examine whether macrophage-derived factors can influence lymphoid cell proliferation and to study the involvement of iNOS/NO, peritoneal macrophages from Wt and GR-deficient mice sensitized with MOG36–50 (day 10) were cultured for 24 h. The cells were pelleted, and the supernatants (referred to as CM) were either used directly or snap-frozen at -80°C (47). Supernatants from macrophage cultures established from healthy Wt and GR-deficient mice were used as controls.
The effects of CM on the proliferative capacity of Wt and GR-deficient mouse LNC or splenocyte cultures (taken on day 10) were determined by adding CM (diluted 5–100% in macrophage nutrient medium) to the cultures in the presence of MOG36–50 or LPS, and the proliferative response was measured. In preliminary dose-response studies, 75–100% CM gave maximal effects, and this concentration was selected for all experiments. No difference between fresh and snap-frozen supernatants was detected. The results for three different preparations of CM within each experimental group were equivalent.
To address the potential role of endogenous nitrogen species, macrophages from Wt and GR-deficient mice immunized with MOG were cultured with L-NIL (2 µg to 2 mg/ml), and the supernatants (CM-NIL) were applied to LNC of MOG-sensitized Wt and GR-deficient mice.
In an additional study the iNOS inhibitor, L-NIL (2 µg-2 mg/ml), or the NO donor, spermine nonoate, ([N-(4(-1–3-aminopropyl]-2-hydroxy-2-nitrosohydrazino)butyl-1,3-propanediamine, 1 mM; Sigma-Aldrich), was directly applied to lymphoid cultures from Wt and GR-deficient mice sensitized with MOG. In some cultures LPS-driven proliferation was assessed in the presence of the NO and carbon monoxide scavenger, ferrous hemoglobin (10-5–10-6M; Sigma-Aldrich). Nitrite levels were measured in the supernatants as described above.
Effects of NOS inhibitors during EAE induction
To examine the effect of NOS inhibition in vivo during the induction phase of EAE, Wt and GR-deficient mice were injected twice daily for 8 days i.p. with either pyrogen-free saline or the NOS inhibitor, L-NAME (125 mg/kg), dissolved in 1 ml pyrogen-free saline beginning on days 0–3 after immunization with MOG peptide in CFA or CFA alone (44). In a second study the specific iNOS inhibitor, L-NIL, was added to the drinking water (100 µg/ml) (19, 48) 0–1 days after immunization for 8 consecutive days. Solutions were prepared daily, fluid consumption in both L-NIL-treated and untreated mice was monitored due to changes in water consumption during the development of disease symptoms, and doses of L-NIL were adjusted accordingly (48). Alternatively, L-NIL (3 mg/kg) was administrated i.p. daily (48). Comparison between oral and i.p delivery of inhibitors yielded similar levels of inhibition of NO production. Mice were weighed and examined twice a day for signs of EAE, and blood from the tail vein was sampled for determination of plasma nitrite/nitrate levels. On day 10 after MOG sensitization, groups of saline- and L-NIL-injected mice (four mice per group) were sacrificed, and the proliferation of spleen and draining lymph node cells was measured.
Statistical analysis
Data were analyzed by two-way ANOVA, with group and time as independent variables, and by Student’s t test. Comparisons a posteriori between different experiments were made using the Newman-Keuls test (49).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In contrast to female Wt B6C3F1 (C57B
x C3H) mice, which developed relapsing remitting EAE (35 of 38), only
one of 35 GR-deficient mice developed clinical or histological signs of
disease (Fig. 1
A). Wt mice
exhibited typical signs of EAE, with an onset of 13.4 ± 1.7 days
and a mean maximal score of 3.2 ± 0.3 at 16.0 ± 0.8 days.
Following a period of remission, 18 of 35 mice exhibited relapses of
neurological deficits (mean day of onset, 32.3 ± 4.5; Fig. 1A). The GR-deficient mouse that exhibited neurological
signs showed tail weakness after 12 days (maximal score, 0.5) for 2
days and did not show signs of relapse until it was sacrificed on day
30 (Fig. 1A). From 12 days (12.2 ± 0.9 days) after
immunization, Wt, but not Tg, mice showed a progressive loss of body
weight. At the end of the experimental protocol the mean weight of Wt
mice was significantly reduced (20.5 ± 0.4 g;
p < 0.01) compared with that of controls (CFA and PBS
treated; 32.5 ± 1.2 g) or Tg mice (34.5 ± 2.9 g;
Fig. 1B). Histological examination of the CNS of Wt mice at
the peak of neurological disease (days 16–20) showed the presence of
typical signs of EAE, i.e., subpial inflammation in the cord with
marked lesions of demyelination (Fig. 2, A and B). In contrast, in GR-deficient mice,
neither inflammation nor demyelination was present at this time (Fig. 2, C and D).
|
|
GR-deficient mice exhibit an aberrant HPA axis response to stress,
demonstrated both in vivo and in vitro (31, 32, 33, 34, 35), and thus
one possible factor contributing to the EAE resistance could be the
endogenous levels of plasma cort. However, similar levels of cort were
present in the sham (PBS)-injected (i.e., preimmunization levels; Fig. 3A) Wt and GR-deficient mice.
Following immunization with CFA alone (data not shown) or
MOG36–50 in CFA (Fig. 3A) Wt mice
exhibited a robust plasma cort response that was sustained during the
course of EAE (p < 0.01 vs preimmunization
levels). Likewise, plasma cort levels in GR-deficient mice following
immunization with CFA alone (data not shown) or
MOG36–50 in CFA were significantly
(p < 0,01) increased above levels in
sham-immunized animals (Fig. 3A); however, these were
significantly lower than those in Wt mice during days 1–12 following
MOG immunization (p < 0.05). After this time
the levels in Tg mice increased to within the range observed in Wt
mice.
|
B). Together these data show that
both Wt and Tg mice have raised plasma levels of cort during EAE,
although the plasma cort levels in GR-deficient mice were in general
lower than that in Wt mice. Thus, the level of GCs after immunization
with MOG is not associated with EAE resistance. T cell proliferation is down-regulated in GR-deficient mice during immunization
To determine whether EAE resistance was related to modulation of
the immune response, the T cell proliferative responses to Con A and
MOG36–50 in the spleen and LN were examined
(Fig. 4, A and B).
Although the proliferative responses following PBS injection of
GR-deficient mice were higher (p < 0.05) than
those of Wt mice (Fig. 4, A and B,
preimmunization) following MOG36–50
immunization, the responses to both Con A and MOG were dramatically
reduced in GR-deficient mice compared with Wt mice
(p < 0.01), especially on days 7–20
postimmunization, when the proliferation index of both
splenocytes and LNC was 
3- to 7-fold lower in GR-deficient compared
with Wt mice (Fig. 4, A and B). These results
indicate a dramatic down-regulation of cell proliferative capacity.
Furthermore Wt mice responded to Con A and
MOG36–50 in a dose-dependent manner (Fig. 4C), particularly between 2.5 and 10.0 µg/ml, while
GR-deficient mice responded to the lower concentrations of 0.5 and 1.2
µg/ml Con A, and higher concentrations resulted in a sharp reduction
of proliferative capacity (Fig. 4C). This shift in the
proliferative dose-response curve of GR-deficient mice LNC suggests the
presence of powerful counterinhibitory mechanisms down-regulating T
cell responses in Tg mice early during EAE induction.
|
Macrophages of EAE-resistant animal strains can down-regulate
Ag-driven T cell proliferation (46, 50). To examine
whether the EAE resistance of GR-deficient Tg mice involves the
macrophage compartment, peritoneal macrophages or CM from macrophage
cultures were incubated with splenocytes or LNC in proliferation
assays. Addition of peritoneal macrophages or CM prepared from control
Wt or Tg mouse macrophages cultures did not influence the proliferative
effect of LNC or spleen cell cultures from Wt and GR-deficient mice
(data not shown). Similarly, macrophages or their conditioned media
from MOG-immunized Wt mice failed to modify the proliferative response
of MOG36–50 immunized mice (Fig. 5, A and B). In
direct contrast, addition of equal numbers of peritoneal macrophages
from GR-deficient mice to Wt LNC cultures inhibited the MOG-driven cell
proliferation in a dose-dependent manner (Fig. 5A).
Similarly, CM prepared from these Tg macrophage cultures established on
day 10 after MOG36–50 immunization (Tg-CM)
significantly inhibited (p < 0.01) the
proliferation of LNC from Wt mice in response to increasing
concentrations of MOG36–50 (Fig. 5B),
whereas Wt-CM was unable to modify the MOG proliferative response of Wt
LNC preparations (Fig. 5B). These results indicated that
factors produced by macrophages of GR-deficient Tg mice, but not Wt
mice, powerfully counter-regulate LNC proliferation.
|
Among the many inflammatory mediators inhibited by GCs, is
macrophage-derived NO (24, 25), a free radical synthesized
in a reaction catalyzed by iNOS. Since iNOS/NO is known to inhibit
proliferation of different cell types (20, 21, 46, 50, 51, 52), we investigated whether the GR deficiency of Tg mice
influences endogenous NO, by application of L-NAME and
L-NMMA or the specific iNOS inhibitor, L-NIL
(45, 48). It was observed that both Tg-CM (Fig. 5B) and Tg macrophages (Fig. 5C) lost the ability
to inhibit Ag-driven LNC proliferation when NOS inhibitors
(L-NAME, L-NMMA,
L-NIL), but not inactive enantiomers
(Ng-monomethyl-D-arginine,
not shown) were applied with the macrophage cultures.
Thus, in the mixed culture condition, NOS inhibition efficiently
counteracted the Tg macrophage-induced decrease in the Wt LNC
proliferative response to MOG36–50 (Fig. 5C). Both L-NAME and
L-NMMA dose-dependently reversed the reduction of
the Wt LNC proliferative response to MOG resulting from addition of Tg
macrophages, but L-NIL was 10- to 50-fold more
potent (Fig. 5C). Likewise, L-NIL (20
µM) sharply counteracted the Tg-CM-induced inhibition of the Wt LNC
proliferative response to increasing concentrations of
MOG36–50 (Fig. 5B). In contrast,
addition of NOS inhibitors to Wt macrophages failed to modify the
proliferative response of Wt cells in both CM (Fig. 5B) and
mixed culture conditions (Fig. 5C).
To further investigate this phenomenon, the ability of NOS inhibitors,
NO scavengers, or NO donors to directly affect MOG- or LPS-driven LNC
or spleen cell proliferation was studied. It was observed that the low
proliferation of mitogen-stimulated LNC from GR-deficient mice was
significantly increased by direct application of L-NIL (20
µM) to levels comparable to those measured in Wt mice, whereas the
iNOS inhibitor did not affect LNC proliferation of Wt-sensitized mice
cultures (Fig. 5D). Further evidence for the contribution of
NO in T cell down-regulation of GR-deficient mice was revealed when the
NO scavenger ferrous hemoglobin (10-6 M) was
added to Tg LNC or spleen cultures stimulated by LPS (50 ng/ml;
6500 ± 1050 cpm in the presence compared with 1600 ± 630
cpm in the absence of NO scavenger). In contrast, a sharp inhibition of
the Wt spleen cell proliferative response to LPS (50 ng/ml) was
measured in the presence of the NO donor spermine NONOate
(spermine NO complex), in which a complete suppression was observed at
1 mM (750 ± 95 cpm in the presence compared with 7500 ±
1200 cpm in the absence of the NO donor). Together, these findings
indicate that the GR deficiency of Tg mice up-regulates iNOS/NO
generation after MOG36–50 sensitization,
pointing to NO as the candidate immunosuppressor molecule.
Nitrite/nitrate levels are elevated in GR-deficient mice
As an indirect measure of NO formation, endogenous plasma
nitrite/nitrate production in vivo during EAE and by macrophages and
splenocytes in vitro was examined. In GR-deficient mice, circulating
nitrite and nitrate levels were elevated 3–20 days after immunization
(Fig. 6A), whereas in Wt mice,
levels were 3- to 4-fold lower before and during the acute disease
compared with days 30 and 40. Similarly, macrophage and splenocyte
cultures of GR-deficient mice established 3, 7, 10, and 20 days
following MOG36–50 immunization exhibited 3- to
5-fold higher LPS-induced NO production (Fig. 6B). In
peritoneal macrophage cultures, NO was maximally increased from day 7,
whereas in splenocytes, LPS-stimulated NO reached a maximal
concentration on days 10–20. In contrast, in Wt macrophage and
splenocyte cultures, nitrite levels were significantly lower, although
a progressive increase was observed (Fig. 6B).
|
Regulation of macrophage NO production by GC is reduced in GR-deficient mice
Next, we investigated the ability of GCs to regulate NO production in macrophage cultures from mice 10 days after MOG immunization compared with direct inhibition of iNOS by L-NIL
Basal nitrite production from peritoneal macrophages of GR-deficient
mice was significantly higher than that of Wt mice
(p < 0.01; Fig. 7C). Both basal and
LPS-induced production of nitrites were inhibited by
L-NAME, L-NMMA (data not
shown), and L-NIL (Fig. 6C) in a
dose-dependent manner. However, L-NIL was 10-
to 50-fold more potent, with an IC50 of 8 µM,
compared with L-NMMA (90 µM) or
L-NAME (700 µM), whereas the inactive
enantiomer of L-NAME,
Ng-monomethyl-D-arginine
was ineffective. These findings clearly supported the hypothesis that
iNOS activation was responsible for higher nitrite formation of Tg
macrophage cultures. As a comparison administration of cort to
LPS-stimulated macrophages induced a dose-dependent inhibition of NO
formation in macrophage cultures from Wt mice, with an
IC50 of 4.7 x 10-9 M
and complete prevention of NO formation at 10-7
M. This effect was mediated by GRs, since pretreatment of macrophage
cultures with the GR antagonist, RU486 (10-6 M),
completely counteracted cort-induced inhibition of nitrite production
(28.8 ± 3.5 µM after RU486 compared with 6.9 ± 0.7 µM
in the absence of RU486; p < 0.01). In macrophage
cultures from GR-deficient mice, cort-induced inhibition of NO
formation was reduced compared with that of Wt cultures
(IC50 = 6.78 x 10-8
M). Moreover, cort was unable to totally suppress NO formation even
under the highest concentrations
(10-7–10-6 M), whereas
in the presence of L-NIL, NO levels were
comparable to those in Wt cultures, again supporting iNOS activation in
macrophages cultures of GR-deficient mice. Thus, the high basal
production of nitrite levels in Tg macrophages was not suppressed by
exogenous application of cort, and macrophage sensitivity to the cort
inhibitory effect was reduced.
|
Specific iNOS inhibition reverts the resistance of GR-deficient mice into susceptibility
To examine the in vivo relevance of increased NO production in
GR-deficient mice, the NOS inhibitors, L-NAME and
L-NIL, were administered during the induction phase of EAE.
Wt and GR-deficient mice were injected i.p. with L-NAME for
8 consecutive days, whereas L-NIL was administered by the
oral route. L-NAME treatment alone of nonimmunized Wt and
GR-deficient mice failed to induce neuropathological effects (data not
shown). In Wt mice immunized with MOG, L-NAME treatment
resulted in a significant exacerbation of the clinical signs (mean EAE
score, 3.55 ± 0.29) compared with
saline-treated Wt mice (2.35 ± 0.35; p < 0.01;
Table I). Body weight was also significantly decreased in
L-NAME-treated mice (17.05 ± 1.5 g on
days 35–40) compared with untreated MOG-immunized Wt mice (22.70
± 1.20 g; p < 0.05; Table I). In addition, the
onset of EAE in L-NAME-treated Wt mice was
significantly earlier (7.4 ± 0.9 days) compared with that
(12.8 ± 1.5 days) in saline-treated Wt mice
(p < 0.01; Table I) and whereas 6 of 8 Wt
saline-injected mice exhibited clinical relapse, 8 of 10
L-NAME-treated mice exhibited relapses. In
GR-deficient mice, L-NAME treatment significantly
increased the number of mice with EAE (10 of 10 mice) compared with
untreated GR-deficient (1 of 10; Table I; p < 0.01)
and the clinical mean EAE score (3.5 ± 0.4) compared with
saline-treated Tg mice (1 of 10; EAE score, 1.5; p <
0.01). In addition, L-NAME treatment
significantly decreased body weight in MOG-immunized Tg (mean body
weight, 18.00 ± 2.0 g) compared with untreated Tg mice
(30.5 ± 2.50 g; p < 0.01; Table I). The
onset of disease in GR-deficient mice given
L-NAME was 9.5 ± 1.2 days, which was
significantly earlier than that in untreated mice (18.0 days). Although
the severity of disease declined (mean score, 1.5 ± 0.5), the
GR-deficient mice did not fully recover for the entire observation
period (35–40 days).
|
;
mean EAE score for L-NIL-treated WT mice, 3.45 ± 0.32
compared with 2.70 ± 0.24 in untreated mice; p <
0.01). The mean EAE score for L-NIL-treated Tg
mice was 3.50 ± 0.30 compared with that in untreated mice (in
eight of eight a clinical score of 0.0; p < 0.01;
Table IB). The onsets of disease in Wt and GR-deficient mice
given L-NIL were 8.0 ± 0.5 and 8.7 ±
1.0 days, respectively, significantly earlier than in untreated mice
(p < 0.01; Table I). Although Wt recovered
from clinical disease, GR-deficient mice did not fully recover for the
entire observation period (35–40 days). This effect was specific for
Tg mice, since Wt mice treated with L-NIL fully
recovered, and subsequently five of eight mice exhibited clinical
relapses. Finally, body weight was less severely affected by
L-NIL compared with L-NAME
in both Wt and GR-deficient mice (Table I). No significant differences were observed when L-NIL treatment was given i.p. compared with oral administration in either time of appearance or severity of disease.
L-NIL treatment suppressed the up-regulated plasma nitrite
and nitrate levels of MOG-sensitized GR-deficient mice and reduced the
already low plasma nitrite concentrations of Wt MOG-immunized mice
(Fig. 7A). A comparable inhibition of nitrite levels was
observed after L-NAME treatment (not shown).
Furthermore, L-NIL treatment significantly
increased both basal and Ag-driven proliferation of splenocytes and LNC
(Fig. 7B). Together these findings strongly support that the
GR deficiency of Tg mice is responsible for iNOS-derived NO
up-regulation and identifies GR-NO interaction as the key mechanism
responsible for the EAE-resistant phenotype.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The absence of a correlation between endogenous production of GCs and
the EAE resistance of GR-deficient mice observed in the present study
clearly indicates that in this sex and strain of mice, the plasma cort
response following MOG36–50 immunization is not
a predictor of disease susceptibility. The transgenic mouse used in the
present work has been characterized as a valuable model to study the
consequences of life-long central GR dysfunction for HPA axis
regulation and other GC-controlled brain functions (30, 34, 35). In previous reports GR-deficient mice displayed a reduced
GC negative feedback (33, 34), exaggerated
adrenocorticotropin hormone (ACTH) response to stress, and
exogenously administrated corticotropin-releasing hormone
(35), but maintained normal early morning levels of both
ACTH and cort, because of the hyposensitivity of the adrenal gland to
ACTH (34). Hence, these Tg mice show reduced GR capacity,
which is not compensated for by elevations in circulating GC levels
(35). In addition, plasma cort concentrations increase to
the same extent in Wt and GR-deficient mice after acute IL-1
challenge (58). In Wistar rats treated during neonatal
life with dexamethasone, the susceptibility to EAE in adult life is
increased in the face of a reduced cort response to endotoxin
challenge, an impaired macrophage capability to respond to LPS, and an
increased expression of proinflammatory cytokines in the spleen
(15). Early life exposure to Gram-negative bacterial
endotoxin results in a remarkable protection from adjuvant-induced
arthritis, while no difference in the cort response to immune challenge
is observed (16). In addition, splenocyte proliferative
response was sharply reduced in these animals (16). It is
interesting to note that animals exposed to Gram-negative bacterial
endotoxin early in life display decreased GR binding in hypothalamus,
frontal cortex, and hippocampus in adult life (12, 13),
but whether a reduction/dysfunction of GRs is also present in immune
tissues is unknown. Although a number of studies indicate that the
predisposition and/or severity of inflammatory and autoimmune diseases
is not always associated with HPA axis reactivity (16, 17, 59, 60), an altered dialogue between the HPA axis and the immune
system during development may program long-term effects in the
mechanisms regulating immune responsiveness to inflammation, thereby
contributing to modulate the severity or predisposition to inflammatory
disorders (15, 16, 17). The present work demonstrates that
activation of macrophage iNOS/NO as a result of GR deficiency in Tg
mice represents one crucial regulatory mechanism.
GR-deficient mice develop 4- to 6-fold higher plasma nitrite/nitrate
levels compared with Wt mice both before and during the acute EAE
phase. Conversely, plasma GC levels are 2- to 3-fold lower in Tg
compared with EAE-susceptible Wt mice. Since NO is involved in the
modulation of neuroendocrine functions (61), and during
inflammation NO can restrain the HPA axis (4), the
up-regulated nitrite levels of GR-deficient mice may at least in part
contribute to reduce the cort concentration in response to MOG
immunization. Interestingly, PVG rats, which have a very low
susceptibility to myelin basic protein-induced EAE
(50) develop higher serum levels of NO (50),
and spleen cell cultures stimulated in vitro with IFN-
produce
significantly higher levels of NO than the EAE-susceptible Lewis rat
(50). Consistent with this finding, NO has been shown to
play a critical role in the recovery of Lewis rats from EAE and the
maintenance of resistance to reinduction of disease (53).
Moreover, rodents immunized with CFA are resistant to subsequent
attempts to induce autoimmune diseases through mechanisms requiring a
functional iNOS gene (54). Rapid induction of Th1-type
cytokines and iNOS are also responsible for the protective effects of
high dose Ag therapy in EAE (62). The results of the
present study thus indicate that NO is involved in the control of
susceptibility to EAE in this sex and mouse strain, and that such
control is directly linked to GR.
GR-deficient mice have previously been shown to mount higher T cell
responses to T-dependent mitogens during postnatal maturation of thymus
and spleen (32). However, in this study MOG-immunized
adult GR-deficient mice showed a marked reduction of T cell responses
despite the higher basal proliferative capacity of GR-deficient
lymphocytes. In sharp contrast with Wt mice responses, application of
Con A or MOG in vitro induced a significant and long-lasting
down-regulation of cell proliferation, especially after the second
immunization challenge, suggesting that after a first priming signal
such an inhibitory mechanism(s) was more actively stimulated in
GR-deficient mice cultures. Although different events, including
alteration of cytokine production, changes in T cell sensitivity to
proapoptotic signals, and/or the production of suppressor molecules
(20, 21), may participate in the immunosuppression of Tg
mice, the final common pathway appears to be macrophage activation of
iNOS/NO, as demonstrated by studies ex vivo and in vitro, using NOS
inhibitors, NO scavengers, or NO donors. In addition, the ability of
L-NIL treatment in vivo to sharply increase MOG-reactive
LNC and spleen cell proliferation and revert the EAE resistance of
GR-deficient mice further supports the participation of iNOS/NO in
counteracting MOG reactive cell expansion during the early induction
phase of EAE (46, 55). Several endogenous and exogenous
molecules can induce (LPS, IFN-, TNF-, and IL-1) or repress
(TGF-, IL-4, IL-8, IL-10, corticosteroids, and estrogens) iNOS
activity. Thus, complex interplays between hormones and cytokines may
interact in vivo to temporally and spatially induce differential
expression of iNOS in organs and tissues of Tg mice, leading to
inhibition of autoreactive cell proliferation. That NO is involved in T
cell regulation is well documented, since both aggravation and
suppression of EAE have been attributed to NO (19, 20, 21 ; see
Ref. 55 for review). Mice lacking the ligand-binding chain
of the IFN- receptor (IFN--/-) develop
severe and usually fatal EAE as opposed to their Wt
IFN-+/+ counterparts, which are resistant to
EAE (46). Willenborg et al. (46) showed that
reduced NO production by peripheral macrophages correlated with
increased autoreactive T cell proliferation of
IFN--/- mice. In contrast, peritoneal
exudate from IFN-+/+-resistant mice produced
high levels of NO and down-regulated Ag-driven cell proliferation of
IFN -/- mice (46).
The down-regulatory effect of NO in primary EAE has been previously suggested to involve inhibition of T cell proliferation as well as inhibition of cell adhesion and migration (55). Such modulation may be dependent on the known functions of iNOS/NO, such as suppression of cytokine production, scavenging of superoxide, and/or apoptosis of macrophages or (encephalitogenic) T cells (20, 21). Inducible NOS-derived NO can modulate cytokine response of macrophages and T cells through activation/inactivation of ion channels, G proteins, protein tyrosine kinase, Janus kinases (Jak1, Jak2, Jak3, Tyk2), mitogen-activated protein kinases (i.e., extracellular signal-regulated kinase, p38 mitogen-activated protein kinase, c-Jun NH2-terminal kinases), caspases (i.e., caspase 1, caspase 3), metalloproteases, and phosphoproteases (63). Thus, complex mechanisms may underlay the NO-mediated inhibition of MOG-reactive cell proliferation observed in the present study.
To establish a direct link between the GR deficiency and the increased NO production in GR-deficient mice, we evaluated the capability of exogenously applied GC to inhibit macrophage production of nitrite corresponding to the phase of maximal inhibition of autoreactive cell proliferation in Tg spleen and LNC cultures. Hence, in GR-deficient mice macrophage appeared resistant to GC inhibition, since cort was unable to suppress both basal and LPS-stimulated NO production, whereas in Wt mice a similar dose of cort significantly suppressed nitrite production. The inverse correlation between up-regulation of NO and decreased GR transcription and binding capacity in immune tissues (32) suggests that increased NO production after Ag priming may contribute to further lower GR expression. Our previous studies indicated the presence and uniqueness of the antisense mRNA transgene in different tissue (cortex, hypothalamus, pituitary, heart, liver, kidney, adrenal) extracts of Tg mice, thus suggesting a less stringent neuronal expression directed by the 2.4-kb human neurofilament L gene promoter (30). Although not addressed in the present study, the decreased GR mRNA and binding capacity measured in thymus and spleen of Tg mice (31, 32) reflect transgene expression in immune tissues. The almost complete (70–90%) loss of GR mRNA transcription and binding capacity measured in lymphoid cells of GR-deficient compared with Wt mice (not shown) following Ag priming support this contention. Reaction of NO with sulfhydryl groups yields S-nitrosothiols (64), and nitrosylation of critical cysteine residues of the GR by NO has been demonstrated to cause a loss GR binding activity and GR signaling in L929 cells in vitro (64). Furthermore, incubation of GR with NO donors results in both time- and concentration-dependent decreases in ligand binding and GR signaling (64). It is therefore feasible that as well as the specific immunization paradigm (MOG and CFA) and the rodent strain, which are known to modulate NO production, GR down-regulation may also play a pivotal role in NO generation. Thus, the GR deficiency in the Tg mice may prime monocytes and macrophages to produce higher amounts of NO via iNOS induction directly or through increased production of Th1 cytokines (62). Our data argue that an excess production of NO can trigger a feedback loop involving loss of GR signaling, thus perpetrating or exacerbating NO output at a critical time during Ag priming, resulting in deletion and/or suppression of Ag-specific cells.
That iNOS is involved in EAE has been shown by Gold et al. (19) using the iNOS-specific inhibitor L-NIL, which induced a marked aggravation of neurological deficits in active myelin basic protein-induced EAE, whereas treatment of rats with adoptive EAE was protective. In another model of inflammation, i.e., streptococcal cell wall-induced arthritis in the rat, McCartney-Francis et al. (48) demonstrated that pharmacological inhibition of iNOS with L-NIL exacerbated streptococcal cell wall-induced arthritis, whereas the nonselective NOS inhibitor, L-NMMA, was disease-protective, suggesting different roles for constitutive and inducible NOS isoforms in the context of joint inflammation. In our study we demonstrate that L-NIL is more potent in inhibiting NO production by LPS-stimulated peritoneal macrophages in vitro, but is also equally as effective as a nonselective inhibitor of NOS in suppressing circulating nitrites in vivo during EAE. Regardless of the route of administration and the dose regimen used, both L-NAME and L-NIL treatment significantly accelerated the onset of and increased the severity of clinical EAE in GR-deficient and Wt mice. Interestingly, while Wt mice fully recovered from acute disease, the NOS inhibitors prevented full recovery in GR-deficient mice, suggesting that factors/mechanisms contributing to the recovery of Wt mice are not fully operative in mice with a dysfunctional GR.
In summary, this work identifies NO as a crucial mediator in EAE resistance of GR-deficient mice and highlights its prominent role as a key messenger of the HPA-immune axis. These results may explain the difference in the susceptibility to inflammatory diseases depending on prenatal experiences known to influence the expression and function of GRs and may thus have clinical implications for the development of autoimmune diseases in humans (65). Modulation of iNOS/NO by stress- or pharmacologically induced alterations at the GR level may also have potential therapeutic implications for such autoimmune diseases.
|
Footnotes |
|---|
2 Address correspondence and reprint requests to Dr. Bianca Marchetti, Neuropharmacology, OASI Institute for Research and Care on Mental Retardation and Brain Aging, Via Conte Ruggero 73, 94018 Troina (EN), Italy. E-mail address: bianca.marchetti{at}oasi.en.it
3 Abbreviations used in this paper: GC, glucocorticoid; CM, conditioned medium; cort, corticosterone; EAE, experimental autoimmune encephalomyelitis; GR, glucocorticoid receptor; HPA, hypothalamic-pituitary-adrenocortical; iNOS, inducible NO synthase; LNC, lymph node cell; L-NIL, L-N6-(1-iminoethyl)-lysine; L-NAME, N-nitro-L-arginine methylester; L-NMMA, Ng-monomethyl-L-argine; MOG, myelin oligodendrocyte glycoprotein; MS, multiple sclerosis; Tg, transgenic; Wt, wild type; ACTH, adrenocorticotropin hormone; Jak, Janus kinase.
Received for publication January 24, 2002. Accepted for publication March 28, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
B in mediation of immunosuppression by glucocorticoids. Science 270:283.B activity through induction of IB synthesis. Science 270:286.B. Mol. Pharmacol. 49:15.[Abstract]
B and inhibiting NF-B. Am. J. Physiol. 273:G1290.
, p < 0.01 compared with
untreated.