| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Divisions of Medical Pharmacology (A.M.K., O.C.M., E.R.d.K.) and Pharmacology (I.C.J.v.d.S., E.C.M.d.L., A.G.d.B.), Leiden/Amsterdam Center for Drug Research, Leiden University, Leiden, 2300 RA, The Netherlands; and Institute for Neurobiology (P.J.L.), University of Amsterdam, 1090 GB, Amsterdam, The Netherlands
Address all correspondence and requests for reprints to: A. M. Karssen, Division of Medical Pharmacology, Leiden/Amsterdam Center for Drug Research, P.O. Box 9503, Leiden University, Leiden, 2300 RA, The Netherlands. E-mail: Karssen{at}LACDR.LeidenUniv.nl
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
[3H]Cortisol poorly penetrated the brain of adrenalectomized wild-type mice, but the uptake was 3.5-fold enhanced after disruption of Pgp expression in mdr 1a-/- mice. In sharp contrast, treatment with [3H]corticosterone revealed high labeling of brain tissue without difference between both genotypes.
Interestingly, human MDR1 Pgp also differentially transported cortisol and corticosterone. LLC-PK1 monolayers stably transfected with MDR1 complementary DNA showed polar transport of [3H]cortisol that could be blocked by a specific Pgp blocker, whereas [3H]corticosterone transport did not differ between transfected and host cells.
Determination of the concentration of both steroids in extracts of human postmortem brain tissue using liquid chromatography mass spectrometry revealed that the ratio of corticosterone over cortisol in the brain was significantly increased relative to plasma.
In conclusion, the data demonstrate that in both mouse and human brain the penetration of cortisol is less than that of corticosterone. This finding suggests a more prominent role for corticosterone in control of human brain function than hitherto recognized.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
These observations raised the possibility that the blood brain barrier (BBB) limits the access of dexamethasone to the brain (3, 8, 9). Recently, it was indeed demonstrated that the penetration of dexamethasone into the brain is hampered because the multidrug resistance 1a (mdr1a) P-glycoprotein (Pgp) excludes this exogenous compound from mouse brain (10, 11). The drug-transporting Pgp is expressed at the luminal membranes of endothelial cells of the BBB (12, 13). This transmembrane protein is encoded by the mdr1a gene in rodents and by the highly homologous MDR1 gene in humans (14, 15).
Thus, Pgp may explain why moderate amounts of dexamethasone primarily
act at the anterior pituitary level to suppress stress-induced ACTH
release (3). In contrast, in rodents corticosterone
primarily acts on centrally regulated functions underlying behavioral
adaptation and the activity of the hypothalamic-pituitary-adrenal axis
(16, 17, 18). In many other species, cortisol is the principal
endogenous glucocorticoid; e.g. in human blood, cortisol
circulates in 10- to 20-fold higher levels than corticosterone
(19, 20, 21, 22). As a naturally occurring glucocorticoid,
cortisol is commonly accepted to exert similar actions in human brain
as corticosterone does in rat and mouse brain. However, although it has
a high affinity for MR, a tracer dose of cortisol has been reported to
be poorly retained in cell nuclei of rat hippocampi (7).
This may not be surprising, because rat and mouse lack the
17
hydroxylase needed for synthesis of cortisol, which therefore
makes this steroid exogenous in these species and hence, as Pgp is
known to transport many unrelated but all exogenous compounds
(15), a potential target for Pgp-mediated export from the
brain.
In the present study, we have first tested the hypothesis that mdr1a Pgp at the mouse BBB limits in vivo brain penetration of cortisol. For this purpose we have used adrenalectomized mdr1a null and wild-type mice injected with a tracer dose of [3H]cortisol or [3H]corticosterone, which freely crosses the BBB. In addition, we have investigated whether a species difference exists between the multidrug resistance Pgp of mouse and man, which would allow free entrance of cortisol into the human brain. To explore this possibility, we have examined the corticosteroid transport capabilities of monolayers of human MDR1-transfected porcine LLC-PK1 cells compared with nontransfected LLC-PK1 cells. To examine the in vivo effect of MDR1 Pgp we have extracted both corticosteroids from human plasma as well as from postmortem human brain material to simultaneously determine cortisol and corticosterone concentrations using liquid chromatography-mass spectrometry (LC-MS).
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
After transport, the mice were housed individually at our laboratory at ambient temperature and at a 12-h light, 12-h dark schedule (lights on at 0700 h, lights out at 1900 h) with free access to food and water. To remove the source of endogenous corticosterone, mice were bilaterally adrenalectomized under gas anesthesia (isoflurane) by a dorsal approach. After adrenalectomy (ADX) the animals had free access to 0.9% saline. At the time of the experiment the animals weighed 27 ± 2.7 g (mean ± SD).
Two days after ADX, the animals were sc injected with tritiated steroids (dissolved in 2% ethanol/0.9% saline) for in vivo autoradiography. Wild-type (n = 4) and mutant mice (n = 6) were injected with 13 µCi per 10 g (1, 2, 6, 7)-[3H]cortisol (Amersham Pharmacia Biotech, Little Chalfont, UK; specific activity 63 Ci/mmol). As a control for nonspecific retention, one mouse of each genotype was pretreated with a 100-fold excess of unlabeled cortisol. In a separate but similar experiment, mice (n = 7–8) were treated with 2.5 µCi per 10 g (2, 4, 6, 7)-[3H]corticosterone (Amersham Pharmacia Biotech; specific activity 70 Ci/mmol). One hour after injection the animals were decapitated. Trunk blood was collected in EDTA-coated tubes and centrifuged for determination of radioactivity and of remaining corticosterone in the plasma using a 125I-labeled corticosterone RIA kit (ICN Biomedicals, Inc., Costa Mesa, CA). Liver, testis, intestine, and cerebellum were dissected and frozen on dry ice. The pituitary was dissected and mounted on top of the brain (without cerebellum), which were then frozen together in isopentane precooled on dry ice/ethanol. All tissues were stored at -80 C until further use.
All organ tissues studied, except for the brain, were homogenized using Soluene-350 (Packard Bioscience, Groningen, The Netherlands). Hionic-Fluor (Packard Instruments Co., Meriden, CT) was added to tissue homogenates and plasma and radioactivity was determined in a Tricarb ß-counter (Packard Instruments Co.). Twelve-micrometer coronal sections of brain were cut on a cryostat and thaw-mounted on poly-L-lysine (Sigma, St. Louis, MO) coated microscopic slides. The slides were put in an x-ray exposure holder (Amersham Pharmacia Biotech) under Ultrofilm (Leica Corp., Heerbrugg, Switzerland) and apposed for 8 weeks. Optical density of radiolabeled steroid retained in pituitary and different brain areas was quantified after subtraction of film background using a computerized Olympus Corp. (Paes, The Netherlands) image analysis system equipped with a Cue CCD camera. From each brain, 3–5 sections were measured by outlining the different brain regions.
Transepithelial transport and inhibition studies
To examine the interactions of cortisol and corticosterone with
the human Pgp we used monolayers of the porcine kidney epithelial
cell-line LLC-PK1, and LLC-PK1 cells stably transfected with
complementary DNA of the human MDR1 gene (LLC-PK1:MDR1).
Cells obtained from the American Type Culture Collection
(Manassas, VA) were kindly provided by the Dutch Cancer
Institute (Amsterdam, The Netherlands) (11). Human
Pgp has been shown before to be specifically expressed on the apical
surface of LLC:PK1:MDR1 cells in these monolayers (23).
Therefore, Pgp substrates entering these cells from the basal side will
be translocated to the apical compartment, whereas those entering the
apical membrane will be pumped back into the medium, thus resulting in
polarized transport of substrates. This system models the way Pgp is
likely to function in the BBB in excluding drugs from the brain.
Cells were cultured at 37 C in the presence of 5% CO2 in complete medium, which consisted of DMEM (BioWhittaker Europe, Verviers, Belgium) supplied with 25 mM HEPES and 4.5 g/liter glucose and supplemented with 100.000 U/liter penicillin, 100 mg/liter streptomycin, 2 mM L-glutamine, and 10% (vol/vol) FCS. The LLC-PK1 and LLC-PK1:MDR1 cell lines were subcultured by trypsinization every 3–4 days, and medium was replaced twice a week.
During the experiments complete medium was used. The LLC-PK1 and LLC-PK1:MDR1 cells were seeded on microporous polycarbonate membrane filters (0.4-µM pore size, 12-mm diameter, Transwell; Costar, Cambridge, MA) at a density of 120 x 103 cells/cm2. The cells were grown for 5–6 days in complete medium with one medium replacement at day 3. Two hours before the start of the experiment, the medium was replaced with 800 µl fresh medium at both the apical and basal side of the monolayer. In the inhibition experiments, 1 h later, the potent and selective Pgp blocker LY 335979 (1 µM; kindly provided by Eli Lilly & Co.) or water was added at the basal side. To measure the transepithelial transport from the apical to the basal side or from the basal to the apical side 8 µl of a 100x stock of tritiated steroid ([3H]cortisol, [3H]corticosterone, or (1,2(n))-[3H]cortisone; Amersham Pharmacia Biotech; specific activity 50 Ci/mmol) in ethanol was added in triplicate at the apical or basal side, respectively, at the start of the experiment (0 h). The starting concentrations for each experiment are mentioned in the legends of the appropriate figures. In the dose-response experiment, different concentrations of unlabeled cortisol were used, supplemented with [3H]cortisol. Over the 4 h of study, 75-µl aliquots were taken once every hour from both compartments. Eight microliter samples of the 100x stock, and samples from the compartment opposite that to which activity was added, were counted in a Tricarb ß-counter after adding 3 ml Emulsifier Safe (Packard Bioscience). Basal-to-apical and apical-to-basal transport is presented as percentage of total activity added at the beginning of the experiment. Transepithelial electrical resistance was measured before and after the experiments to check the integrity of the monolayers.
Corticosteroid determination in postmortem human brain
Human brain material was collected through the rapid autopsy
program of The Netherlands Brain Bank (NBB) (Amsterdam, The
Netherlands; coordinator: Dr. R. Ravid). The NBB abides to all local
ethical legislation. All tissue has been obtained with informed consent
of the donor or next of kin to perform brain autopsy and the subsequent
use of brain tissue for scientific purposes; consent is requested in
advance together with the permission to use the medical records.
Patient tissue was carefully selected; none of the subjects was
reported to suffer at the moment of death or to have suffered before
from a known neurological or psychiatric disease, or from conditions
that might have affected BBB integrity, like transient ischemic
attacks, (suspected) prolonged arterial blood pressure changes,
prolonged fever, or the presence of multiple brain infarcts. Moreover,
none of the subjects had been treated with synthetic steroids or
antidepressants at time of death or at any time during life. From every
subject, a standard set of brain areas has been carefully investigated
(24) by neuropathologists Prof. Dr. D. Troost (Academic
Medical Center, Amsterdam, The Netherlands), Prof. Dr. F. C. Stamand,
and Dr. W. Kamphorst (both from Free University, Amsterdam, The
Netherlands). The final diagnosis was established by relating this
neuropathological examination to the outcome of the clinical diagnosis.
Following this careful examination, all present subjects were confirmed
to be true controls because the tissue was free of any such changes.
Postmortem delay was kept as short as possible and was on average
6.75 h. Further clinicopathological details are presented in
Table 1
. All 11 brain tissue samples used
in this study were dissected from superior parietal cortex of male
control subjects (mean age 65 ± 5.1), rapidly frozen in liquid
nitrogen, and then stored at -80 C until use.
|
Samples were prepared for assay by dichloromethane (DCM)/ethanol extraction. The brain samples (weighing about 350 mg) were homogenized in 2 ml 0.1 M perchloric acid with a Potter-Elvehjem tissue homogenizer (Eli Lilly & Co.) (10 times up and down, 1,000 rpm). To check for differences in recovery, 100 ng dexamethasone was added to each sample. The homogenates were transferred with a 4-ml wash of DCM to screw-capped glass tubes. After adding an extra 4 ml DCM, the tubes were shaken on a horizontal reciprocating shaker for 30 min and subsequently centrifuged at 1,000 x g at 4 C for 10 min. The DCM layer was transferred to a clean coned tube, rinsed with 1 ml water, and centrifuged at 700 x g for 10 min. Then, the DCM-layer was transferred to a long tube and evaporated to dryness in a SpeedVac. To maximize the amount transferred, the extracts were redissolved in 750 µl ethanol and, after transferring to an Eppendorf, evaporated again. The final extracts were resuspended in 100 µl 25% methanol and centrifuged at 13,000 rpm for 5 min. To avoid possible dissimilarities between different extraction methods, the 250-µl plasma samples were extracted in the same way.
LC-MS was the method of choice to measure the levels of cortisol and corticosterone in the supernatants of the extracts as it allowed the simultaneous measurements of both hormones in small samples with dexamethasone as internal standard. The assays were performed on a Triple Stage Quadrupole mass spectrometer (Finnigan MAT TSQ-700, San Jose, CA) with a custom-made atmospheric pressure chemical ionization interface. A modification of the method of Van der Hoeven et al. (25) was used. The analysis was performed in negative ionization mode using selective ion monitoring of [M+CH3COO]- of cortisol, corticosterone, and dexamethasone, alternately scanning mass/charge (m/z) 421, 406, and 452. The ion-source temperature and the nebulization heater were kept at 200 and 400 C, respectively. The voltages on the corona needle and on the electron-multiplier were set at -3200 and -1800 V, respectively. Each experiment, a new calibration series was made in 25% methanol with eight concentrations ranging from 5–500 ng/ml of both cortisol and corticosterone. Dexamethasone (1 µg/ml) was used as an internal standard. An alkyl-diol-sianol C18 column was used to separate the steroids. After injection of 20 µl of the calibration or extraction samples, the column was washed with acetonitrile-water (40/60%, vol/vol) containing 1 g/liter acetic acid at a flow rate of 0.5 ml/min. The detection limit of this assay was 5 ng/ml. Corticosteroid concentrations were calculated from a standard plot of area under the curve vs. concentration. Presented data are corrected for recovery of dexamethasone, which was in the order of 20–40%.
Statistical analysis
Human and mouse data were evaluated by Student’s t
test. The results of the monolayer experiments were analyzed by
repeated measures ANOVA. Significance was taken at P less
than 0.05.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). In contrast, the amount of
[3H]corticosterone in cerebellum did not differ
between the two genotypes. For both corticosteroids the presence or
absence of the mdr1a gene did not affect their concentration
in plasma, liver, testis, and intestine (Fig. 1 and
Table 2). The autoradiograms (Figs. 2 and 3)
extend these results to the regional distribution of the
[3H]steroids in the brain. The
mdr1a+/+ animals showed hardly any labeling of
brain tissue after administration of
[3H]cortisol (Fig. 2A). Labeling in brain
sections was restricted to the cerebral ventricles. However, the
mdr1a-/- mutants showed increased labeling of
whole brain (Fig. 2B). In particular, radioactivity was retained in
hippocampal cell fields and to a lesser extent the amygdala. In
contrast, after treatment with
[3H]corticosterone the null and wild-type ADX
mice did not differ in their strong labeling of hippocampal neurons or
of any other part of the brain (Fig. 2, C and D). In both mutants and
controls, [3H]cortisol labeling of the
pituitary, which lies outside the BBB, was not affected by the absence
of the mdr1a Pgp. Pretreatment with excess unlabeled cognate steroid to
block specific labeling to receptors resulted in loss of labeling of
hippocampal neuronal fields and amygdala, but not of the rest of the
brain (data not shown). Moreover, hippocampal optical density showed an
inverse correlation with residual levels of endogenous corticosterone,
illustrating that the signal represents specific receptor-bound steroid
(data not shown). The cortex and the pituitary lack this correlation,
but cortex labeling showed a clear effect of disruption of the
mdr1a gene (Fig. 3). These data evidently demonstrate that
the presence of mdr1a Pgp in the BBB hampers the access of cortisol to
the mouse brain, but does not have any effect on the access of the
endogenous glucocorticoid corticosterone.
|
|
|
|
), although some polar transport was
observed in both cell lines in all our experiments (Figs. 4B and 5B).
Nonetheless, this demonstrates the absence of human MDR1 Pgp-mediated
transport of corticosterone. However, cortisol was transported in a
polarized fashion in the MDR1-transfected monolayers, but not in the
host cells (Fig. 4A). Polarized transport in MDR1 monolayers of
cortisol was abolished in presence of LY335979, a potent and selective
Pgp blocker (26, 27), resulting in similar fractions
transported as in untransfected cells (Fig. 5A). This confirms that cortisol
transport is mediated by human Pgp. LY335979 did not change the
fraction of corticosterone translocated through the membrane (Fig. 5B).
We also examined transepithelial transport of different concentrations
of cortisol, ranging from 5 to 625 nM, but did
not demonstrate any saturation at higher dose. At all concentrations
tested, about 3.5 times more [3H]cortisol had
been transported from basal to apical sides than from apical to basal
sides after 4 h (data not shown). Interestingly, these data show
that human MDR1 Pgp is also able to transport cortisol, whereas
corticosterone passage remains unchanged.
|
|
). To subsequently
exclude the possibility that we have measured the metabolites we
further examined the transport of [3H]cortisol
or [3H]corticosterone in presence of
carbenoxolone (10-6 and
10-5
M). The addition of this 11ß-HSD inhibitor did
not change the transport properties of LLC-PK1 or LLC-PK1:MDR cells in
any way (data not shown).
|
Corticosteroid levels in human brain
To test whether MDR1 Pgp in human BBB increases the ratio of
corticosterone over cortisol in brain, we determined the concentrations
of both glucocorticoids in human brain samples (Table 3). Thus, we were able to establish the brain corticosterone to
cortisol ratio of 11 subjects, which was 0.31 ± 0.04 (mean
± SEM) (Fig. 7). In
contrast, we measured a corticosterone to cortisol ratio in plasma
samples of age-matched males of 0.06 ± 0.01 (Fig. 7). Statistical
analysis showed that the difference between the brain and plasma ratios
was significant (t(1, 20) = 6.444, P < 0.01).
Thus, corticosterone appears to penetrate more easily in the human
brain than cortisol, resulting in a higher ratio of corticosterone
over cortisol present in brain compared with plasma.
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Previous studies had already established in vivo the low cell nuclear retention of cortisol in rat brain (7). In rats, the first pass uptake in brain after a carotid injection of [3H]cortisol appeared to be negligible in contrast to uptake of [3H]corticosterone, whereas uptake of the labeled steroids in liver after portal injection was not different (29). A tracer dose of corticosterone is known to label only the high affinity hippocampal MR, leaving the lower affinity GR undetectable. The uptake of this steroid is not affected by disruption of the mdr1a gene (Ref. 10 and this study). Cortisol also binds with a rather high affinity to MR (16). In fact, our autoradiography study revealed a pattern of cortisol labeling in the mdr1a-/- mouse hippocampus reminiscent of that of corticosterone. In the present study, the effect of mdr1a ablation on specific binding of cortisol to the low capacity MR in the hippocampus is less pronounced than its effect on uptake in whole brain. This is probably due to the lower affinity of cortisol for MR in rodent brain (30, 31). Anyhow, our data convincingly demonstrate that the mouse mdr1a Pgp hampers the brain uptake of cortisol but not of corticosterone.
As a model for Pgp function in human BBB, we have used monolayers of pig kidney epithelial LLC-PK1 cells stably transfected with the MDR1 gene to measure transport of steroids by human Pgp. Such monolayers of epithelial cells are a plausible model for Pgp-mediated transport at the BBB, given the apical localization of Pgp forming a barrier between the two compartments. A confounding factor is that LLC-PK1 host cells contain low levels of porcine Pgp (23, 32, 33). Thus, in theory, porcine Pgp might be responsible for polar transport of corticosterone seen in both cell lines, although any effect after application of the potent and selective Pgp-blocker was absent. LLC-PK1 cells also have endogenous 11ß-hydroxysteroid dehydrogenase type 2 activity, able to inactivate cortisol and corticosterone (28). Because we used radiolabeled glucocorticoids it is possible that we have actually measured transport of labeled metabolites rather than the parent hormone. However, the transport of [3H]cortisol and [3H]corticosterone did not change in the presence of the 11ß-HSD inhibitor carbenoxolone, indicating that 11ß-HSD activity did not interfere. Therefore, we conclude that our monolayers are a suitable model of Pgp function in human BBB.
Our data corroborate several studies on transport of cortisol and corticosterone by Pgp (23, 34, 35, 36, 37, 38, 39). Differential transport of these two steroids by murine Pgp has been observed in several drug-resistant cell lines, using steroid induced apoptosis (35, 40) or steroid accumulation (38, 39) as read-outs. The murine mdr1b Pgp has some capacity to transport corticosterone (38), but this second murine multidrug resistance Pgp is not expressed at the BBB. In view of the lack of corticosterone transport that we have observed in cells stably transfected with the human MDR1 gene, the corticosteroid transport capabilities of the human MDR1 Pgp apparently correspond to that of murine mdr1a rather than to that of mdr1b Pgp. Using comparable monolayers to those in this study, Ueda et al. (23) have already demonstrated that cortisol is transported by the human MDR1 Pgp, but corticosterone was not included in their assay. In human colon carcinoma cells the amount of accumulated [3H]cortisol is lower than of corticosterone (39), whereas both steroids equally increase accumulation of the Pgp substrate [3H]vinblastine, exemplifying a difference between actual transport by and binding of steroids to the pump.
The difference in interaction of Pgp with cortisol and corticosterone is remarkable considering their large similarity in molecular structure. Pgp is an efflux transporter with a surprisingly broad substrate spectrum (41), but corticosterone only differs from cortisol in the lack of the 17-hydroxyl group. However, there are indications that both the 17hydroxyl and the 11-hydroxyl group determine the ability of steroids to be transported by Pgp (40). Pgp transports steroids having both these hydroxyl-groups whereas steroids lacking one of these groups are probably minimally if at all transported. A caveat is that these indications are based on the extent of glucocorticoid resistance, which also depends on GR affinity. It is difficult to assess the influence of the 11-hydroxyl group because steroids lacking this group do also have a low affinity for GR. Therefore, cortisone could not be identified as a substrate of Pgp in the previous study; however, using LLC-PK1:MDR1 monolayers, we demonstrated that cortisone is also transported by Pgp.
Our study with the mdr1a null mice is the first to directly show the involvement of Pgp in excluding a naturally occurring glucocorticoid from the brain. Previous studies have demonstrated that access of the synthetic glucocorticoid dexamethasone to the brain was also enhanced in the mdr1a-/- mouse (10, 11). In vitro studies have confirmed that dexamethasone is a Pgp substrate (23, 40, 42). In fact, in our stably MDR1-transfected LLC-PK1 monolayers, dexamethasone behaved very similar to cortisol (43). Thus, human MDR1 Pgp, like mouse mdr1a Pgp, transports both cortisol and dexamethasone, but not corticosterone.
Our in vitro results using monolayers of stably MDR1-transfected LLC-PK1 cells show that the endogenous presence in a species of a naturally occurring glucocorticoid is not a prerequisite to exclude transport by Pgp. We have clearly demonstrated that human MDR1 Pgp is able to discriminate between cortisol and corticosterone. Both glucocorticoids are present in human plasma, although cortisol circulates in about 10–20 times higher levels than corticosterone (19, 20, 21, 22, 44). The data strongly suggest that corticosterone rather than cortisol can freely gain access to the human brain. Alternatively, BBB passage of cortisone and subsequent conversion of cortisone to cortisol by 11ß-HSD type 1 present in brain (45), might regenerate cortisol in brain. The fact that MDR1 Pgp also transports cortisone, argues against the possibility that cortisol would be able to circumvent Pgp in the BBB through this route, because less cortisone would also be available for 11ß-HSD type 1 conversion in brain. Therefore, the limited access of cortisol and cortisone is likely to result in overall lower brain levels of glucocorticoids and in an increase of corticosterone relative to cortisol in human brain compared with plasma.
An in vivo cell nuclear retention study in ADX rhesus monkeys, which have cortisol as their main glucocorticoid, showed a similar regional pattern for both corticosteroids, but the amount of cortisol radioactivity was lower than that of corticosterone (46). This observation substantiates that even in an animal that normally produces cortisol, this glucocorticoid less efficiently penetrates into the brain than corticosterone. We demonstrated a shift in the corticosterone to cortisol ratio in favor of corticosterone in human autopsy brain samples compared with plasma samples. These results support data reported by Brooksbank et al. (47), who also demonstrated that corticosterone is accumulated in the brain to a substantially greater extent than cortisol. They found a ratio of corticosterone to cortisol of about 0.4. Earlier, Fazekas and Fazekas (48) also determined corticosteroid levels in human brain using paper chromatography and similarly reported high levels of corticosterone relative to cortisol.
The privileged uptake of corticosterone in the brain is also expected to promote its receptor occupancy relative to cortisol. There are indications that corticosterone might have a higher affinity for the MR than cortisol. At least this is the case for the rat MR (30, 31), but data presented by Arriza et al. (49) also suggests that corticosterone is the more potent competitor at human MR. Furthermore, transactivation of human MR in response to cortisol and corticosterone indicates that corticosterone is more effective than cortisol (50). Thus, besides the hampered uptake in human brain, cortisol might also less effectively mediate the human brain MR response. Should it indeed be confirmed that levels of GR are relatively low in the human hippocampus, as was recently claimed for the rhesus monkey (51), glucocorticoid-mediated effects on hippocampal functioning might then mainly reflect corticosterone acting through MR rather than cortisol. At least, our data suggest that the human glucocorticoid feedback system might be more complex than the rodent system in view of the potential different roles for cortisol and corticosterone.
The influence of cortisol on brain functioning and its role as main corticosteroid in glucocorticoid feedback to the human brain is commonly accepted. However, in contrast to rodents where corticosterone readily enters the brain, the main glucocorticoid in human appears to be partially excluded from the brain. It would be interesting to know how much either corticosterone and cortisol contributes to stabilization of neuronal excitability (52), maintenance of neuronal integrity (53), suppression of hypothalamic-pituitaryadrenal activity (18), and facilitation of behavioral adaptation (54). The preferential uptake of corticosterone in human brain may further be used as a lead toward the development of novel selective steroids for treatment of stress-related brain disorders.
In conclusion, we have demonstrated the involvement of Pgp in hampering the access of the naturally occurring glucocorticoid cortisol rather than corticosterone to both mouse and human brain. Therefore, the data suggest that corticosterone may play a more prominent role in the modulation of human brain function than hitherto recognized.
|
Acknowledgments |
|---|
Received November 13, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  B. L. Mason, C. M. Pariante, and S. A. Thomas A Revised Role for P-Glycoprotein in the Brain Distribution of Dexamethasone, Cortisol, and Corticosterone in Wild-Type and ABCB1A/B-Deficient Mice Endocrinology, October 1, 2008; 149(10): 5244 - 5253. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. R. de Kloet and R. A. Sarabdjitsingh Everything Has Rhythm: Focus on Glucocorticoid Pulsatility Endocrinology, July 1, 2008; 149(7): 3241 - 3243. [Full Text] [PDF]
|
|
|
|
|
|
S. K. Droste, L. de Groote, H. C. Atkinson, S. L. Lightman, J. M. H. M. Reul, and A. C. E. Linthorst Corticosterone Levels in the Brain Show a Distinct Ultradian Rhythm but a Delayed Response to Forced Swim Stress Endocrinology, July 1, 2008; 149(7): 3244 - 3253. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. L. Schmidt and K. K. Soma Cortisol and corticosterone in the songbird immune and nervous systems: local vs. systemic levels during development Am J Physiol Regulatory Integrative Comp Physiol, July 1, 2008; 295(1): R103 - R110. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Ye, C. J Kenyon, S. M MacKenzie, K. Nichol, J. R Seckl, R. Fraser, J. M C Connell, and E. Davies Effects of ACTH, dexamethasone, and adrenalectomy on 11{beta}-hydroxylase (CYP11B1) and aldosterone synthase (CYP11B2) gene expression in the rat central nervous system J. Endocrinol., February 1, 2008; 196(2): 305 - 311. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Sanderson, A. Khan, and S. Thomas Distribution of Suramin, an Antitrypanosomal Drug, across the Blood-Brain and Blood-Cerebrospinal Fluid Interfaces in Wild-Type and P-Glycoprotein Transporter-Deficient Mice Antimicrob. Agents Chemother., September 1, 2007; 51(9): 3136 - 3146. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. S. McEwen Physiology and Neurobiology of Stress and Adaptation: Central Role of the Brain Physiol Rev, July 1, 2007; 87(3): 873 - 904. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. J. Mark and B. J. Waddell P-Glycoprotein Restricts Access of Cortisol and Dexamethasone to the Glucocorticoid Receptor in Placental BeWo Cells Endocrinology, November 1, 2006; 147(11): 5147 - 5152. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 V. Arya, V. G. Demarco, M. Issar, and G. Hochhaus CONTRARY TO ADULT, NEONATAL RATS SHOW PRONOUNCED BRAIN UPTAKE OF CORTICOSTEROIDS Drug Metab. Dispos., June 1, 2006; 34(6): 939 - 942. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. N. Samtani, N. A. Pyszczynski, D. C. DuBois, R. R. Almon, and W. J. Jusko Modeling Glucocorticoid-Mediated Fetal Lung Maturation: I. Temporal Patterns of Corticosteroids in Rat Pregnancy J. Pharmacol. Exp. Ther., April 1, 2006; 317(1): 117 - 126. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. B. Goralski, P. D. Acott, A. D. Fraser, D. Worth, and C. J. Sinal BRAIN CYCLOSPORIN A LEVELS ARE DETERMINED BY ONTOGENIC REGULATION OF MDR1A EXPRESSION Drug Metab. Dispos., February 1, 2006; 34(2): 288 - 295. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Karssen, O. C. Meijer, A. Berry, R. Sanjuan Pinol, and E. R. de Kloet Low Doses of Dexamethasone Can Produce a Hypocorticosteroid State in the Brain Endocrinology, December 1, 2005; 146(12): 5587 - 5595. [Abstract] [Full Text] [PDF]
|
, 
) and from apical-to-basal (
, •) compartment was measured
in wild-type LLC-PK1 (broken line) or MDR1-transfected
LLC-PK1 (solid line) monolayers. Presented is the
fraction of the dose of radioactivity, which is 9 nM for
both steroids, added to the respective compartment. Each
point represents the mean of three monolayers ±
SEM. Repeated measures ANOVA showed a significant
interaction of time* cell type* transport for cortisol
(P = 0.00) but not for corticosterone.
) and from
apical-to-basal (
) compartment was measured in MDR1-transfected
LLC-PK1 monolayers. Presented is the fraction of the dose of
radioactivity, which is 8 nM for cortisol and 28
nM for corticosterone, added to the respective compartment.
Each point represents the mean of three monolayers ±
SEM. Repeated measures ANOVA showed a significant
interaction of time* cell type* transport for cortisol
(P = 0.00), but not for corticosterone.
