| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Max Planck Institute of Psychiatry (A.S., T.P., F.H., J.M.H.M.R.), Section of Neuropsychopharmacology, D-80804 Munich, Germany; Max Planck Institute of Neurobiology (A.S., C.L.), Department of Neuroimmunology, D-82152 Martinsried, Germany; University of Vienna (A.S., M.K.S., C.S., H.L.), Brain Research Institute, A-1090 Vienna, Austria; Department Of Neurology (M.K.S.), University of Graz, A-8010 Graz, Austria; and Department of Pharmacology (F.J.H.T.), Faculty of Medicine, Free University, 1081 BT Amsterdam, The Netherlands
Address all correspondence and requests for reprints to: Dr. J. M. H. M. Reul, Max Planck Institute of Psychiatry, Section of Neuropsychopharmacology, Kraepelinstrasse 2–16, D-80804 Munich, Germany. E-mail: reul{at}mpipsykl.mpg.de
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although the mechanistic basis of relapse/remission in MS is still poorly understood, several mechanisms that could be involved in modulating disease activity within the CNS have been identified in experimental autoimmune encephalomyelitis (EAE). EAE is an animal model of MS induced by either active immunization with CNS myelin autoantigens or, alternatively, the adoptive transfer of myelin antigen-specific T cell lines or clones (2). Regulation of disease activity in EAE has focused on immune mechanisms that influence the myelin-specific autoimmune response (3, 4, 5). However, these immunological mechanisms act in the context of a neuroendocrine response that is, itself, essential to resolve acute T cell-mediated inflammatory responses in the CNS (6, 7).
During the onset of clinical disease in EAE, the inflammatory process results in a marked increase in plasma corticosterone (CORT) via the effect of proinflammatory cytokines such as IL-1 and IL-6 on the hypothalamic-pituitary-adrenocortical (HPA) axis (8). The importance of this transient CORT response for the resolution of an acute episode of clinical disease was demonstrated in myelin basic protein (MBP)-induced EAE, an acute monophasic disease model in which recovery is spontaneous (6).
Here, we have investigated the importance of the glucocorticoid response in a novel model of chronic, relapsing/remitting EAE (CR-EAE) induced in female DA rats by immunization with the extracellular domain of the myelin oligodendrocyte glycoprotein (MOG). This disease model, in contrast to other EAE paradigms, reproduces all of the crucial immunopathological and clinical features of MS, including the formation of confluent plaques of demyelination in the brain, optic tract, and spinal cord (9, 10).
Demyelination in MOG-induced EAE is mediated by a MOG-specific autoantibody response that directs a combination of complement- and antibody-dependent cellular cytotoxicity-dependent effector mechanisms to selectively attack the myelin sheath (11, 12, 13, 14). This is in striking contrast to MBP-induced EAE, a purely T cell-mediated disease in which demyelination is minimal. In the current study, we developed a biphasic variant of MOG-induced CR-EAE, characterized by an initial acute inflammatory disease episode, a short clinical remission followed by relapse and a chronic progressive neurological deficit associated with extensive demyelination. We report that relapse and the onset of chronic progressive disease in this model of MS is associated with a failure of the disease process to sufficiently stimulate CORT production.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Induction of EAE and histopathological analysis
Animals were immunized, at the base of the tail, with 75 µl
MOG occulum, prepared by emulgating MOG (2 mg/ml) 1:1 with IFA,
resulting in a dose of 75 µg MOG/rat. Control animals were injected
with an emulsion of buffer in IFA. Clinical disease was scored on the
following scale: 0.5, partial loss of tail tone; 1.0, complete tail
atony; 2.0, hind limb weakness; 3.0, hind limb paralysis; 4,
moribund.
Histological evaluation was performed on paraformaldehyde-fixed paraffin-embedded sections of brains and spinal cords as previously described (10, 18, 19). Immunohistochemistry was performed on paraffin sections with monoclonal antibodies against the following targets: macrophages/activated microglia (ED1; Serotec, Oxford, UK), T-cells/polymorphonuclear cells (PMNs) (W3/13; Seralab, Sussex, UK), and an antiserum to IL-1ß (Serotec). Bound primary antibody was detected with a biotin-avidin technique. Control sections were incubated in the absence of primary Ab or with nonimmune rabbit serum. Using a morphometric lattice, numbers of immunolabeled cells were determined on three to five representative spinal cord cross-sections per animal. Cell density was expressed as cells/mm2. On adjacent serial sections, the demyelinated area was determined after Luxol fast blue staining. Using an optical grid, the cross-section area and the demyelinated areas were determined, and demyelination was expressed as a percent of cross-section area.
Plasma CORT and ACTH levels, stimulation of the HPA axis, and CORT
treatment
To determine plasma levels of CORT, great care was taken to keep
rats undisturbed the night before the experiment. The animals were
quickly anesthetized (<15 sec) with Halothane (Hoechst Marion Roussel, Inc., Frankfurt, Germany) immediately after removal
from their home cage, between 0700 and 0800 h. Trunk blood was
collected after decapitation in ice-chilled, EDTA-coated tubes
containing 140 µg aprotinin (Trasylol, Bayer Corp., Cologne, Germany). The whole procedure was
performed in less than 1 min. In some experiments, the thymus was
removed, cleaned, and weighed. Blood samples were centrifuged at 4 C
for 10 min, and plasma aliquots were stored at -80 C for analysis by
RIA (ICN Biomedicals, Inc., Costa Mesa, CA). The inter-
and intraassay coefficients of variance for CORT were 7 and 4%,
respectively, with a detection limit of 0.15 µg/100 ml. For ACTH, the
inter- and intraassay coefficient of variance were 7% and 5%,
respectively, with a detection limit of approximately 2 pg/ml.
As a disease-unrelated psychological HPA-axis stimulus, novelty stress was induced by placing animals individually in new cages for 30 min before collecting trunk blood. Other animals were injected ip with 2 µg/kg recombinant rat IL-1 (Batch 29109b, Dr. S. Poole, National Institute for Biological Standards and Control (NTBSC), South Mimms, UK) dissolved in 500 µl saline or saline only, and trunk blood was collected 90 min later. Some animals were sc implanted with pellets releasing CORT (200 mg, 21-d release; Innovative Research of America, Sarasota, FL) under Halothane anesthesia. To determine maintained levels of plasma CORT, in a separate group of pellet-implanted rats, trunk blood was collected 5 d after pellet implantation. In all experiments, stress-free sampling and sample processing were performed as described above.
3H-steroid binding assay
Hippocampal CORT-binding receptors were determined as described
previously (16, 20). Briefly, hippocampi were dissected
from animals, 24 h after adrenalectomy. They were homogenized (100
mg brain tissue/ml) in an ice-cold 5-mM Tris-HCl buffer (pH
7.4) containing 5% glycerol, 10 mM sodium molybdate, 1
mM EDTA, and 2 mM ß-mercaptoethanol,
using a glass homogenizer with a Teflon pestle milled at a clearance of
0.25 mm on the radius. Supernatant (cytosol) was prepared by high-speed
centrifugation (1 h at 2 C and 100,000 x g). Aliquots
of cytosol (100 µl) were incubated with
[3H]-steroids, at 0–4 C for 20–24 h, over a
concentration range of 0.1–10 nM (6–8
concentrations in duplicate; total vol of 150 µl). To measure MR,
aliquots were incubated with [3H]-aldosterone
(NEN Life Science Products, Cologne, Germany) in
the presence of a 100-fold excess of the specific GR ligand RU 28362
[11ß,17ß-dihydroxy-6-methyl-17
-(1-propionyl)
androsta-1,4,6-triene-3-one], and nonspecific binding was assessed in
the presence of a 1,000-fold excess of unlabeled CORT. Binding to the
GR was determined by incubation with
[3H]-dexamethasone (Amersham Pharmacia Biotech, Freiberg, Germany). The binding of
[3H]-dexamethasone to MR was evaluated by
adding an excess of RU 28362, and nonspecific binding was assessed
by adding a 1,000-fold excess of unlabeled dexamethasone.
Corticosterone binding globulin (CBG) binding was determined in 10-fold diluted plasma (from adrenally intact animals; see also below) by incubation with 35 nM [3H]-CORT (NEN Life Science Products) in the absence (for measurement of total binding) or presence (nonspecific binding) of a 1000-fold excess of unlabeled CORT.
After incubation for 20–24 h at 0–4 C, bound and free [3H]-steroid were separated by gel filtration on Sephadex LH-20 (Pharmacia Biotech, Uppsala, Sweden) columns, and bound radioactivity was measured in a liquid scintillation counter. The protein content was determined by the method of Lowry, with BSA as the standard. Binding data were expressed as femtomoles per milligram protein or, in case of CBG, as picomoles per milligram protein, and nonspecific binding was subtracted from total binding to yield specific binding. In this manner, the MR concentration could be directly measured. However, GR binding was estimated by subtraction of the specific binding of [3H]-dexamethasone + 100x RU 28362 from the specific binding of [3H]-dexamethasone. [3H]-dexamethasone + 100x RU 28362, rather than [3H]-aldosterone + 100x RU 28362, binding data were used to estimate the amount of the specific [3H]-dexamethasone binding to MR, because [3H]-dexamethasone + 100x RU 28362 binding to MRs has been found to be about 30% less than [3H]-aldosterone + 100x RU 28362 binding to this receptor type (20). The receptor binding capacity and binding affinity (dissociation constant) were derived from Scatchard analysis.
CBG binding levels were calculated, taking into account the presence of endogenous CORT by adjustment of the specific activity of [3H]-CORT [initial specific activity according to company (NEN Life Science Products) information: 70 Ci/mmol].
Tissue extracts
At given days after immunization, animals were perfused, through
the heart, with ice-chilled heparinized saline. Immediately thereafter,
the spinal cord was dissected out and, after weighing, the
anterior/posterior halves frozen on dry ice and stored at -80 C. The
whole procedure was completed in less than 10 min. Spinal cord
fragments (200–400 mg) were homogenized in 0.5 ml buffer (Iscove’s
medium containing 5% FCS, 100,000 international units/ml
aprotinin, 10 mM EDTA, 5 mM Benzamidin, 0.2
mM phenylmethylsulfonylfluoride). In supernatants obtained
after centrifugation, concentrations of cytokines were determined by
ELISA, and protein concentrations were measured by Bradford assay.
Spiking experiments with rat IL-1ß revealed extraction efficiencies
of approximately 80%. IL-1ß was assayed by ELISA, as previously
described (21), with a detection limit of 10 pg/ml
extract. Interferon (IFN)-
was analyzed using a commercial ELISA kit
(Biosource Technologies, Inc., Camarillo, CA) with
a detection limit of 13 pg/ml.
ELISA
MOG-specific antibodies were determined as previously described
(18). In brief, 96-well microtiter plates
(Costar, Cambridge, MA), coated with 5 µg/ml antigen (3
h, 37 C) in 50 mM carbonate/bicarbonate buffer, pH 9.6,
were incubated with serum samples diluted in PBS (pH 7.4) after
blocking with 1% BSA in PBS overnight at 4 C. Total anti-MOG levels
were determined directly using 100 µl peroxidase-conjugated rat IgG
and IgM-specific goat antibody [1:4000; Dianova,
Hamburg, Germany]. Isotype-specific anti-MOG antibody levels
were determined using 1:4000 dilutions of a panel of mouse mAbs
specific for rat IgM, IgG1, and Ig2a (Serotec), followed
by a mouse-specific peroxidase conjugate (1:8,000 in PBS; Dianova). To
determine the levels of IgE, 2% dried milk powder in PBS was used as a
blocking agent. Furthermore, the samples and a goat serum specific for
rat IgE (1:5000) (Dunn GmbH, Asbach, Germany) and a
horseradish peroxydase-conjugated donkey antigoat serum
[1:2000, Dianova (Germany)] were diluted in PBS containing 0.1% milk
powder. All plates were developed with O-phenylenediamine
dihydrochloride (Sigma, Deisenhofen, Germany), the
reaction was stopped with 3 M HCl, and optical
density was determined at 490 nm.
Data presentation and statistical analysis
Data are presented as group means ± SEM. Means
were compared by t test, one-way ANOVA followed by Duncan’s
post hoc test, or two-way ANOVA, as indicated.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
).
Clinical signs of disease were first observed 8–9 d post immunization
(d.p.i.), after which the neurological deficit progressed rapidly until
all animals exhibited complete hind-limb paralysis, 2–3 d later.
Histopathological analysis revealed that neurological symptoms were
accompanied by heavy infiltration of the CNS by inflammatory cells,
including T cells, PMNs, and macrophages, but only minor demyelination
(Table 1). The majority of animals (15/17
in a representative experiment) then entered a short period of
clear-cut, but often incomplete, remission of clinical disease that
lasted a further 2–3 d and was accompanied by a reduction of T cells
and PMNs in the CNS, whereas the number of macrophages remained
constant (Table 1). The remission was followed by a severe relapse that
started 16–17 d.p.i., characterized by the development of progressive
neurological deficit that proved fatal between 20–25 d.p.i. CNS
infiltration by T cells was comparable with the first episode of
disease, whereas macrophages increased 2-fold in number by 17
d.p.i., and PMNs increased more than 3-fold. In addition, demyelination
was found to increase 20-fold between 14 d.p.i. (remission) and
17 d.p.i. (relapse) (Table 1), suggesting that demyelination is a
dominant factor responsible for the neurological deficit in the second
phase of disease.
|
|
). During this phase, plasma
CORT levels rose 13-fold, compared with baseline values (baseline on
0 d.p.i.: 3.0 ± 1.6 µg/dl vs. 39.2 ± 5.5
µg/dl on 11 d.p.i., P < 0.05, Duncan’s
post hoc test; Fig. 1B). After 11 d.p.i., plasma CORT
dropped significantly, by approximately 50%, as the animals entered
clinical remission (i.e. 14 d.p.i.: 18.7 µg/dl
± 4.5 µg/dl, P < 0.05 vs. values at
11 d.p.i. by Duncan’s post hoc test). However, the
severe relapse, which occurred in all animals, was not accompanied by a
second rise in plasma CORT levels (Fig. 1B). Indeed, as disease
progressed, CORT levels decreased even further, from 18.8 ± 3.8
µg/dl, 17 d.p.i. (P < 0.05 vs.
baseline, Duncan’s post hoc test) to 14.9 ± 3.7
µg/dl, 20 d.p.i., when levels were no longer significantly
different from baseline values (Fig. 1B).
ACTH followed a pattern similar to that of CORT initially (Fig. 1D),
with a significant surge in plasma levels during the first phase of
MOG-EAE (11 d.p.i.: 134.1 ± 34 pg/ml vs. 40.5 ±
12.5 pg/ml on 0 d.p.i., one-way ANOVA, F (4, 37) =
25.8, P < 0.0001). Yet, whereas CORT levels remained
elevated on 14 and 17 d.p.i., ACTH levels dropped back to baseline
levels more rapidly, with no significant elevation at any time-point
other than 11 d.p.i. (6 d.p.i.: 56.7 ± 6.3 pg/ml, 14
d.p.i.: 61.9 ± 8.4 pg/ml, 17 d.p.i.: 61.8 ± 11.5
pg/ml).
Similarly, CBG was elevated significantly only during the first phase
of MOG-EAE [Fig. 1C (11 d.p.i.: 14.8 ± 1.8 pmol/mg
vs. 4.6 ± 0.4 pmol/mg on d 0, one-way ANOVA, F (4, 37) = 5.8, P = 0.001)] and dropped back to
baseline levels thereafter (d 6 p.i.: 8.9 ± 2.4 pmol/mg,
14 d.p.i.: 8.9 ± 0.8 pmol/mg, 17 d.p.i.: 7.7 ±
1.0 pmol/mg).
The continuous biological activity of circulating CORT is underlined by
the significant and progressive thymic involution observed over the
entire course of CR-EAE (Fig. 1E). Corresponding to the first clinical
phase, thymic wet weight on 11 d.p.i. had decreased by almost two
thirds, compared with baseline values (11 d.p.i.: 115.7 ± 26.1 mg
vs. 275.8 ± 12.6 mg on d 0, one-way ANOVA, F (5, 37) = 5.8, P = 0.001), and continued to drop, to
less than 20% of its original weight by 20 d.p.i. (6 d.p.i.:
226.3 ± 6.6 mg, 14 d.p.i.: 89.5 ± 23.5 mg, 17
d.p.i.: 59.6 ± 8.6 mg, 20 d.p.i.: 57.4 ± 12.3 mg).
Hippocampal corticosteroid receptor binding is altered during
CR-EAE
Hippocampal MRs and GRs are known to be critically involved in HPA
axis regulation (20, 23, 24, 25). Therefore, alterations in
receptor density, over the course of disease, may contribute to the
observed changes in HPA axis activity. We measured binding to
hippocampal MRs and GRs, and we found that, interestingly, the density
of both receptor types exhibited changes that were reciprocal to plasma
CORT levels (Table 2). The changes
in MR binding were particularly pronounced (one-way ANOVA: F (4, 20) = 201, P < 0.0001) and, corresponding to
maximal HPA axis activation between 11–14 d.p.i. MR levels, decreased
by approximately 50%, from 58.5 ± 1.8 fmol/mg (6 d.p.i.) to
25.6 ± 1.6 fmol/mg (14 d.p.i.). Levels of GR fell less
dramatically [one-way ANOVA: F (4, 20) = 98.2, P
< 0.0001], reaching 94% of control levels (137 ± 4.7 fmol/mg)
on 11 d.p.i., and 75% on 14 d.p.i. In addition, over the
course of CR-EAE, the ligand-binding affinity of MR, as derived from
Scatchard analysis, was reduced in parallel with the receptor density
on 11 and 14 d.p.i. Thus, one might expect that the observed
changes in receptor binding properties would contribute to relapse
induction. However, at the onset of relapse, by 17 d.p.i., both
receptor capacity and receptor affinity had returned to (or even
exceeded) the levels seen preimmunization (Table 2).
|
, A and B). Novelty stress
induced significant changes in plasma CORT, both in healthy animals and
animals suffering from EAE (two-way ANOVA: effect of EAE: F (1, 31) = 14.33, P = 0.001; effect of novelty-stress:
F (1, 31) = 22.6, P 
0.0005; two way
interaction: F (1, 31) = 0.24, P = 0.627). In
unstressed animals with EAE, on 17 d.p.i., CORT levels were
elevated as expected (22.8 ± 5.7 µg/dl, see also Fig. 1B); and
exposure to novelty stress triggered a significant further increase in
plasma CORT, to 45.6 ± 6.7 µg/dl (P = 0.02,
t test; Fig. 2, A and B). This clearly shows that the HPA
axis does not succumb to exhaustion during the course of disease and
remains responsive to disease-unrelated stress, even as the animals
enter relapse.
|
). However, this increase was only half of
that observed in healthy animals injected with the same dose of IL-1ß
(Fig. 2C) [83.2 ± 7.1 µg/dl (IL-1ß injected rats)
vs. 15.1 ± 3.4 µg/dl (saline-injected rats), two-way
ANOVA, effect of EAE: F (1, 22) = 11.4, P = 0.003;
effect of IL-1ß: F (1, 22) = 57.4, P < 0.0001;
two-way interaction: F (1, 22) = 4.9, P = 0.037].
These data show that, during the second disease phase (17 d.p.i.), the
HPA system is normoresponsive to novelty stress but hyporesponsive to
an inflammatory stimulus such as that provided by IL-1ß, an effect
that may be, in part, responsible for the dissociation of disease
activity and plasma CORT levels during relapse.
Supplementation with exogenous CORT prolongs remission
The observation that the HPA axis becomes hyporesponsive to
exogenous inflammatory stimulation (17 d.p.i.) suggests that
stimulation of CORT production by the inflammatory response was simply
insufficient to block disease progression. This hypothesis was tested
by supplementing animals with established disease with exogenous CORT
using CORT-releasing pellets (200 mg, 21-d release). Animals suffering
from EAE were implanted sc, 11 d.p.i., corresponding to the time
of maximal endogenous CORT production (Fig. 3A). In a parallel experiment, pellet
implantation maintained a plasma concentration of CORT (51.6 ±
4.3 µg/dl), which was similar to the levels of endogenous CORT
attained in EAE-diseased rats at 11 d.p.i. (i.e.
39.2 ± 5.5 µg/dl; Fig. 3B). Recovery from the first phase of
disease was not affected by pellet implantation, but the remission
phase was prolonged for more than 2 wk, compared with 2 d in
sham-treated animals (Fig. 3A). This demonstrates that CORT can
suppress disease activity for a prolonged period providing plasma
levels are sufficiently high. However, all animals eventually relapsed
(at the latest, when the implants were exhausted; approximately 20
d after implantation), indicating that CORT suppresses disease
progression without eliminating the underlying pathogenic autoimmune
response.
|
were far lower than those seen during the initial episode of disease.
The in situ cytokine response was determined by assaying
IL-1ß and IFN- levels in spinal cord extracts, by ELISA, and found
to be markedly lower at 17 d.p.i. vs. 11 d.p.i.
(Fig. 4), though changes in IFN- did
not reach statistical significance because of high variability between
the animals [one-way ANOVA, F (2, 16) = 1.9, P =
0.16] (Fig. 4A). In contrast, the concentrations of IL-1ß in the
spinal cord showed significant changes over the course of disease
(one-way ANOVA, F (2, 35) = 7.1, P = 0.0025) and
were low in naïve animals (5.6 ± 0.9 pg/100 mg) (Fig. 4B). At 11 d.p.i., they had increased approximately 80-fold and
reached a value of 459.3 ± 101.2 pg/100 mg (P <
0.05 vs. baseline, Duncan’s post hoc test) but
by 17 d.p.i. had fallen significantly, by approximately 50%
(233.0 ± 53.8 pg/100 mg; P < 0.05 vs.
d 11, Duncan’s post hoc test). This result was unexpected,
because macrophages (the major source of IL-1ß) were more abundant in
the lesions during the relapse, compared with the first disease episode
(Table 1). We thus investigated IL-1ß expression in situ,
by immunocytochemistry, and found that, despite the high numbers of
macrophages in the lesions, a significantly lower proportion of these
cells expressed IL-1ß, in comparison with those in the acute attack
of disease (Table 1, Fig. 5).
Thus, the decreased IL-1ß levels within the tissue might best be
explained by an altered activational state of inflammatory cells within
the respective lesions.
|
|
Late CR-EAE is associated with high titers of demyelinating
anti-MOG antibodies
In contrast to CNS inflammation that peaks early during CR-EAE and
then declines, demyelination increased 20-fold between first and second
attack (Table 1). Measurement of serum anti-MOG antibodies, which, in
the rat, mediate primary demyelination, revealed a 3-fold increase in
titers in this period (Fig. 6). Total
anti-MOG IgG/IgM (Fig. 6A) titers were 1:6,300 on 11 d.p.i., rose
to 1:11,200 until 14 d.p.i., and peaked at 1:19,500 by 17
d.p.i. (d 20: 1:14,800, data not shown). This rise in serum titers
affected both Th1-associated IgG2b (Fig. 6C) and Th2-associated IgE
(Fig. 6D) to a similar degree.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It is well known that chronic activation induces major changes in the regulation and responsiveness of the HPA axis (32, 33, 34). However, during CR-EAE, the HPA axis does not develop a state of general unresponsiveness, but rather a selective desensitization to inflammatory mediators. The HPA response to a disease-unrelated stressor was normal at relapse, whereas the response to an exogenous proinflammatory cytokine, IL-1ß, was markedly reduced, relative to healthy controls. This may be another example of the desensitization to a homotypic stressor, because also repeated injection of lipopolysaccharide (LPS) causes desensitization of the HPA axis response (35, 36). Although this phenomenon has never been shown before for the condition EAE, a parallel might exist between our observation of the effect of IL-1ß in late-stage EAE and repeated LPS treatments. Nevertheless, the mechanistic basis of this effect of disease on the HPA axis response remains to be clarified. However, it could not be attributed to changes in the affinity and binding capacity of either hippocampal MRs or GRs, as these were similar to control values at the onset of relapse. Still, it should be noted that major changes in both MR and GR occur during the first episode of clinical disease that may, by reduced feedback inhibition, participate in activation of the HPA axis at this time (23). The marked decrease in the ligand-binding affinity of MR may indeed be the result of CNS inflammation, given that similar changes were observed after LPS or IL-1 administration (37). However, although MR and GR parameters that we measured normalize by the time the animals relapse, we cannot rule out the possibility that the initial disease episode induces a persisting alteration in the associated signaling pathways, a feature of the disease process that we are currently investigating.
The low levels of endogenous CORT produced during relapse are clearly
insufficient to suppress disease activity in the CNS; but can this be
simply attributed to desensitization of the HPA axis or are there other
factors involved? Stimulation of the HPA axis during EAE is driven by
proinflammatory cytokines, such as IL-1ß, produced during the course
of the local immune response (8). However, during relapse,
levels of IL-1ß in spinal cord tissue extracts, as well as the number
of IL-1ß reactive cells in spinal cord sections, were substantially
lower than in the initial phase of disease, despite the fact that
inflammation, in terms of the numbers of cells infiltrating the CNS,
was greatest during relapse. This observation was apparently
attributable to differences in macrophage activation rather than a
selective decrease in the number of macrophages, the most important
sources of IL-1ß. Moreover, a similar trend was observed for IFN-
levels in CNS tissue extracts assayed at the same time-points. Although
the data for IFN- was not statistically significant, it suggests
that disease activity during relapse continues and that the
neurological deficit increases in severity despite a generalized
reduction in the production of proinflammatory cytokines in the CNS.
Analysis of anti-MOG antibody isotypes suggests that this is not caused
by a significant Th1-Th2 shift over the course of disease. Clearly,
many proinflammatory and regulatory cytokines apart from IL-1ß and
IFN- are involved in the pathogenesis of EAE, and a full
characterization of these cytokines in MOG-EAE and their influence on
the HPA axis is in progress.
The observation that severe clinical disease develops during relapse despite a relative reduction in the production of proinflammatory cytokines can be explained by the significant rise in demyelination after remission, which will not only induce clinical deficit in its own right but also render the CNS more susceptible to inflammatory mediators (38). It is well established that demyelination is antibody-, rather than T cell-mediated in rat models of EAE and correlates with the levels of anti-MOG antibodies (11). Furthermore, antibody-dependent demyelination triggers severe clinical disease in the context of a subclinical inflammatory response in the CNS (11, 14, 18). These findings are reproduced in the present study, where we found a rapid increase both in anti-MOG antibody titers and effector cells of antibody-mediated demyelination, such as CNS-macrophages, between remission and relapse in CR-EAE.
In summary, this study indicates that relapse and/or conversion to a progressive clinical disease in MOG-induced EAE is attributable to a combination of multiple factors that prevent successful control of the disease process: 1) reduced HPA axis sensitivity to inflammation; 2) reduced HPA axis drive by CNS inflammation; and 3) predominance of antibody-mediated demyelination over inflammation during relapse, which results in an increased vulnerability of the damaged CNS (38). The significance of these findings is underlined by the recent demonstration of disease mechanisms resembling MOG-EAE in MS patients (39), indicating that a similar scenario of immune-endocrine interactions may also contribute to the chronicity of the disease process in MS.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: CBG, Corticosterone binding globulin; CNS, central nervous system; CORT, corticosterone; CR-EAE, chronic relapsing experimental autoimmune encephalomyelitis; d.p.i., days post immunization; HPA, hypothalamic-pituitary-adrenocortical; IFA, incomplete Freund’s adjuvant; IFN, interferon; LPS, lipopolysaccharide; MBP, myelin basic protein; MOG, myelin oligodendrocyte glycoprotein; MS, multiple sclerosis; PMN, polymorphonuclear cell.
Received August 3, 2000.
Accepted for publication April 3, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  M. W. Meagher, R. R. Johnson, E. G. Vichaya, E. E. Young, S. Lunt, and C. J. Welsh Social Conflict Exacerbates an Animal Model of Multiple Sclerosis Trauma Violence Abuse, July 1, 2007; 8(3): 314 - 330. [Abstract] [PDF]
|
|
|
|
 |
|
 A. Vroon, A. Kavelaars, V. Limmroth, M. S. Lombardi, M. U. Goebel, A.-M. Van Dam, M. G. Caron, M. Schedlowski, and C. J. Heijnen G Protein-Coupled Receptor Kinase 2 in Multiple Sclerosis and Experimental Autoimmune Encephalomyelitis J. Immunol., April 1, 2005; 174(7): 4400 - 4406. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Bourquin, A. Schubart, S. Tobollik, I. Mather, S. Ogg, R. Liblau, and C. Linington Selective Unresponsiveness to Conformational B Cell Epitopes of the Myelin Oligodendrocyte Glycoprotein in H-2b Mice J. Immunol., July 1, 2003; 171(1): 455 - 461. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. K. Droste, A. Gesing, S. Ulbricht, M. B. Muller, A. C. E. Linthorst, and J. M. H. M. Reul Effects of Long-Term Voluntary Exercise on the Mouse Hypothalamic-Pituitary-Adrenocortical Axis Endocrinology, July 1, 2003; 144(7): 3012 - 3023. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
) (200 mg/21 d release) or placebo pellet (
)
11 d.p.i. (arrow) and monitored for clinical
disease. B, In a separate experiment, plasma CORT levels were
determined in healthy animals (n = 5) 5 d after pellet
implantation (P = 0.001, t test) and
were found to be comparable with disease-induced CORT levels on day
11 p.i. (P = 0,168, t test).
), 14 (•) or 17 (
) d.p.i. were analyzed for the presence of
total IgG + IgM anti-MOG antibodies, by ELISA, and titers were
determined at an OD490 = 1. In samples from individual
animals, total IgG + IgM (B) was compared with Th1-associated IgG2b (C)
and Th2-associated IgE (D) at a serum dilution of 1:4000.