Glucocorticoid Insensitivity at the Hypoxic Blood–Brain Barrier Can Be Reversed by Inhibition of the Proteasome
Background and Purpose—Glucocorticoids potently stabilize the blood–brain barrier and ameliorate tissue edema in certain neoplastic and inflammatory disorders of the central nervous system, but they are largely ineffective in patients with acute ischemic stroke. The reasons for this discrepancy are unresolved.
Methods—To address the molecular basis for the paradox unresponsiveness of the blood–brain barrier during hypoxia, we used murine brain microvascular endothelial cells exposed to O2/glucose deprivation as an in vitro model. In an in vivo approach, mice were subjected to transient middle cerebral artery occlusion to induce brain infarctions. Blood–brain barrier damage and edema formation were chosen as surrogate markers of glucocorticoid sensitivity in the presence or absence of proteasome inhibitors.
Results—O2/glucose deprivation reduced the expression of tight junction proteins and transendothelial resistance in murine brain microvascular endothelial cells in vitro. Dexamethasone treatment failed to reverse these effects during hypoxia. Proteasome-dependent degradation of the glucocorticoid receptor impaired glucocorticoid receptor transactivation thereby preventing physiological glucocorticoid activity. Inhibition of the proteasome, however, fully restored the blood–brain barrier stabilizing properties of glucocorticoid during O2/glucose deprivation. Importantly, mice treated with the proteasome inhibitor Bortezomib in combination with steroids several hours after stroke developed significantly less brain edema and functional deficits, whereas respective monotherapies were ineffective.
Conclusions—We for the first time show that inhibition of the proteasome can overcome glucocorticoid resistance at the hypoxic blood–brain barrier. Hence, combined treatment strategies may help to combat stroke-induced brain edema formation in the future and prevent secondary clinical deterioration.
Disruption of the blood–brain barrier (BBB) and successive edema formation are pathological hallmarks of various disorders of the central nervous system and can dramatically deteriorate neurological symptoms especially in patients with stroke. Glucocorticoids (GCs) are frequently applied to fight BBB leakage in different central nervous system disorders, but GC efficacy is highly variable. Although GCs diminish edema formation in neuroinflammatory diseases such as acute multiple sclerosis lesions, and in certain brain tumors, this substance class is ineffective or even harmful in acute ischemic stroke.1,–,5 This is unfortunate because excessive edema formation is a frequent cause of secondary infarct growth and successive death in patients with stroke.6 The mechanisms underlying this discrepant GC efficacy are largely unknown and their revelation could help to develop novel antiedematous strategies in many neurological diseases.
The biological effects of GC are mediated by the glucocorticoid receptor (GR), which is a member of the nuclear receptor superfamily.7 After ligand, that is, steroid engagement, GR binds to specific DNA sequences (glucocorticoid-response element) in the 5′-flanking region of target genes thereby activating gene transcription.7,8 The GR is expressed in a wide range of tissues, including the vascular endothelium, and GR expression levels correlate with its transcriptional activity. An important level of regulation of GC responses is the post-translational modification and degradation of GR by the ubiquitin–proteasome system.9
Proteasomal degradation of nuclear receptors on the one hand is a physiological process necessary to terminate transcriptional activity (eg, hormone response) after ligand binding. On the other hand, it can restrict transcriptional signaling by steroids under certain pathophysiological conditions thereby compromising steroid function.9 Hypoxia, for example, has been shown to induce proteasomal degradation of the estrogen receptor α in breast cancer cells resulting in reduced estrogen sensitivity10,11 and several studies have confirmed that lack of oxygen is a critical mediator of nuclear receptor disintegration in various tissues (reviewed in Meller12).
We demonstrate here that O2/glucose deprivation (OGD) fosters degradation of the GR in an in vitro model of the BBB in a proteasome-dependent manner leading to impaired GC sensitivity. Inhibition of the proteasome restored the responsivity of the BBB to GC during hypoxia. Importantly, combined inhibition of the proteasome together with GC treatment also significantly attenuated edema formation and neurological deficits after transient middle cerebral artery occlusion (tMCAO) in mice, which is an established in vivo model of hypoxic BBB damage.
Materials and Methods
Detailed materials and methods are available in the Supplement (available at http://stroke.ahajournals.org).
Animal experiments were approved by the Regierung von Unterfranken, Germany, and conducted according to the recently published recommendations for research in mechanism-driven basic stroke studies.13 tMCAO using an intraluminal filament was performed as described previously14 (see Supplemental Methods). Middle cerebral artery occlusion time was 60 minutes in all experiments.
Two hours after reperfusion, mice were treated with intravenous injections of dexamethasone (Sigma-Aldrich, Taufkirchen, Germany; 10 mg/kg), methylprednisolone (Sigma-Aldrich; 60 mg/kg), Bortezomib (Janssen-Cilag GmbH, Neuss, Germany; 0.2 mg/kg), or the combination of Bortezomib plus dexamethasone or Bortezomib plus methylprednisolone (n=10/group). Control mice (n=10) received 0.9% NaCl as a carrier solution.
Direct, that is, without correction for brain edema, and indirect, that is, corrected for brain edema, infarct volumes were calculated from 3 coronal 2,3,5-triphenyltetrazolium chloride (TTC)-stained brain sections14,15 according to the following equations: in which the term (V1−VC) represents the volume difference between the ischemic hemisphere and the control hemisphere and (V1−VC)/VC expresses this difference as a percentage of the control hemisphere.
Brain edema volumes were then calculated by subtracting indirect from direct infarct volumes.
The extent of BBB damage after tMCAO was assessed by measuring the volume of Evan Blue leaked into the brain parenchyma as described16 (see Supplemental Methods).
The Bederson score17 was used to quantify stroke-related functional deficits in mice.
Isolation and Culture of Murine Cerebral Endothelial Cells
The immortalized mouse brain capillary endothelial cell line cEND was generated as described.18
Cell Cultures, In Vitro Stroke Conditions, and Cell Viability Studies
Transfection and Luciferase Assay
Essentially, transfection and luciferase assays were carried out as described.18
Transfection of cEND Cells With K426A-GR
Transient transfection experiments using the Effectene reagent (Qiagen) were performed using 2 μg of ubiquitination-defective K426A-GR expression vector.9
Electrophoresis and Immunoblotting, Quantitative Real-Time Reverse Transcription–Polymerase Chain Reaction, and Transendothelial Electric Resistance Measurements
Electrophoresis and immunoblotting, quantitative real-time reverse transcription–polymerase chain reaction (RT-PCR), and transendothelial electric resistance measurements were performed following standard procedures (see Supplemental Methods).
Immunohistochemistry on cryopreserved mouse brain slices was performed as previously described.20 A GR-specific primary antibody (sc-1004; Santa Cruz, Santa Cruz) was applied at a dilution of 1:200. To identify the cellular origin, we performed double staining of GR with the endothelial marker platelet-endothelial cell adhesion molecule-1 (CD31; 1:200; Abcam, Cambridge, UK). Counterstaining with 4′,6-diamidino-2-phenylindole was used to identify cell nuclei. Secondary antibodies were Alexa Fluor 488-coupled goat antimouse (1:100; BD Biosciences, Heidelberg, Germany) and Cy3-coupled goat antirabbit (1:300; Dianova, Hamburg, Germany). Negative controls included omission of the primary or secondary antibody and revealed no detectable signal (not shown). Stainings (n=3/group) were examined in a blinded fashion by microscopy (Axiophot2; Zeiss, Oberkochen, Germany) with a charge-coupled device camera (Visitron Systems, Tuchheim, Germany).
Results are presented as mean±SD. For statistical analysis Bonferroni-corrected 1-way analysis of variance or, in the case of 2 independent variables, Bonferroni-corrected 2-way analysis of variance was used (PrismGraph 4.0 software package; La Jolla, CA). Probability values <0.05 were considered statistically significant.
Establishment of an In Vitro Model of Hypoxic BBB Damage
To investigate the molecular basis for GC insensitivity of the BBB under conditions of hypoxia, we first established an in vitro OGD model using cEND cells. OGD increased lactate dehydrogenase release, a measure of plasma membrane damage and cytotoxicity,21 from cEND cells in a time-dependent manner (Figure 1A): 2 hours of OGD moderately elevated lactate dehydrogenase release, that is, relative cytotoxicity, to 10.7%±2.1% compared with control conditions (P>0.05). After 4 hours, relative cytotoxicity had reached 15.5%±1.1% (P>0.05). Prolonged OGD over 8 hours resulted in significant cEND damage (21.4%±2.6%, P<0.02), which further increased until 16 hours (66.1%±2.0%, P<0.02; Figure 1A).
We next assessed the activity of the GR in cEND cells in different settings over time. For this purpose, transactivation of the established GC-responsive test promoter mouse mammary tumor virus22 was determined after OGD or control conditions in the presence or absence of dexamethasone (Figure 1B). Significant mouse mammary tumor virus promoter–construct transactivation, that is, GC responsivity, was observed in dexamethasone-treated cEND cells compared with cells not receiving dexamethasone under physiological conditions after 4 hours (4.2±0.1-fold induction, P<0.05), 8 hours (4.7±0.5-fold induction, P<0.05), and 16 hours (4.3±0.2-fold induction, P<0.05), respectively (Figure 1B). In contrast, OGD completely abolished mouse mammary tumor virus promoter–construct transactivation indicating loss of dexamethasone-mediated GR activity already after 1 hour of OGD (Figure 1B).
Taken together, these findings point out that shortage of O2 and glucose damages cEND cells and reverses their natural GC responsivity. Hence, the cEND system is suitable to emulate the in vivo conditions at the hypoxic BBB present, for example, during ischemic stroke. Given the results from this preliminary test series, 4 hours of OGD followed by a reoxygenation period for 24 hours in the presence or absence of 100 nmol/L dexamethasone was chosen as a standard condition for all following in vitro experiments.
To secure that the general ability to induce a transcriptional response is preserved in cEND cells under OGD conditions, we determined the induction of the hypoxia-inducible factor-dependent gene vascular endothelial growth factor23 by quantitative RT-PCR (Figure 1C). After 4 hours or 6 hours of hypoxia, vascular endothelial growth factor expression was induced 1.5±0.1-fold or 1.5±0.3-fold (P<0.05), respectively. This finding excludes that OGD-induced cell stress leads to a general loss of the transcriptional response in our cEND system.
GC Barrier Stabilizing Effects Are Lost in Brain cEND Under Hypoxic Conditions in Vitro
As a next step, functional consequences of the reduced GC sensitivity in ischemic cEND cells on surrogate markers of barrier maintenance were analyzed. OGD markedly compromised barrier tightness in cEND monolayers as reflected by a significant decrease in transendothelial electric resistance (TER) compared with monolayers maintained under normoxic conditions (195±47 Ωcm2 versus 371±59 Ωcm2, P<0.05; Figure 2A). Treatment of normoxic cEND cells with dexamethasone significantly increased TER to 671±43 Ωcm2 (P<0.05), whereas dexamethasone failed to restore BBB integrity in hypoxic cEND cells (167±39 Ωcm2 versus 195±47 Ωcm2, P>0.05; Figure 2A).
The compactness of the brain endothelial barrier depends on the expression of tight junction proteins sealing the paracellular space.24 We determined gene expression levels of the tight junction proteins occludin, claudin-3, claudin-5, and claudin-12 as well as the adherens junction protein VE-cadherin in the cEND system in the presence or absence of O2/glucose or dexamethasone using real-time RT-PCR (Figure 2B). In line with previous reports,18 dexamethasone treatment during normoxia induced the expression of the validated GC target gene occludin8,24 compared with untreated control cells (1.7±0.1-fold induction, P<0.02). Moreover, dexamethasone significantly raised claudin-3 (2.9±0.1-fold induction, P<0.001) and claudin-12 (3.3±0.3-fold induction, P<0.05) mRNA levels under normoxia but had no effect on the expression of claudin-5 and VE-cadherin (P>0.05), thus excluding claudin-5 and VE-cadherin as relevant GC target genes at the murine BBB (Figure 2B). OGD alone had no significant effect on the expression of tight junction or adherens junction protein genes in cEND cells but completely abolished the gene inducing properties of dexamethasone (Figure 2B).
At the protein level, administration of dexamethasone under physiological conditions increased the immunoreactivity for occludin (120%±7%, P<0.05) and claudin-5 (110%±7%, P>0.05) but did not alter claudin-3 and claudin-12 expression compared with untreated cEND cells as revealed by Western blot analysis (Figure 2C). OGD decreased the protein amount of occludin (82%±3%, P<0.05) and claudin-5 (61%±4%, P<0.001) in comparison to normoxic cells, whereas it had no effect on claudin-3 and claudin-12 protein contents (Figure 2C). Application of dexamethasone under conditions of OGD could not recover the hypoxia-induced loss of occludin and claudin-5 and likewise did not influence the expression levels of the other TJ proteins investigated (Figure 2C).
The activity of GC is directly linked to the amount of GR protein available.25 We therefore determined the expression of the GR in our in vitro system. As expected for a GC responsive cell line,25 GR protein was readily detectable in cEND cells under normoxic conditions. In response to dexamethasone binding, GR protein levels decreased to 39%±12% (P<0.02) of the levels in untreated cells thereby confirming previous reports (Figure 3A).26 OGD significantly reduced cellular GR protein contents (28%±9%, P<0.002) and the application of dexamethasone under these conditions could not further influence GR levels (Figure 3A). To investigate whether ischemia-induced GR degradation is a specific feature of the cerebral endothelium or a more general phenomenon also occurring in other epithelial cell lines, the renal tubular epithelial cell line Hek293, which has been described to maintain GC sensitivity after OGD,27 was used. OGD or administration of dexamethasone did not alter GR protein levels in Hek293 cells (Figure 3A). In line with these results, mouse mammary tumor virus promoter–construct transactivation, that is, GR activity, was preserved in hypoxic Hek293 cells supplemented with dexamethasone (Figure 3B).
To rule out GR gene expression effects (Figure 3A), quantification of GR transcripts under conditions of OGD or normoxia as well as in the presence and absence of dexamethasone was conducted by real-time RT-PCR (Figure 3C). Neither condition had a significant effect indicating that active GR protein degradation rather than altered GR gene expression is responsible for the low GR protein contents observed in cEND cells on hypoxia (Figures 3A, C).
Proteasomal Degradation of the GR Underlies the Reduced GC Sensitivity of Brain Endothelial Cells During OGD In Vitro
We26 and others9 have previously shown that proteasome-mediated GR degradation restricts GC-induced transcriptional signaling. To address the question whether enhanced proteasomal GR degradation is also responsible for GC insensitivity of endothelial brain barriers during hypoxia, we took advantage of GR-K426A-transfected cEND cells resistant to proteasomal degradation.9 OGD significantly reduced the TER in K426A monolayers (239±58 Ωcm2 versus 371±59 Ωcm2, P<0.05; Figure 2A). In contrast to untransfected cEND cells, however, the sensitivity to dexamethasone on hypoxia was preserved in GR-K426A-transfected cEND cells (167±39 Ωcm2 versus 619±67 Ωcm2, P<0.05; Figure 2A). Our findings on maintained barrier function in steroid-treated proteasome-resistant GR-K426A cells were consistent with the results from real-time RT-PCR and Western blot analyses of TJ proteins. Gene expression levels of occludin (1.8±0.2-fold induction, P<0.01), claudin-3 (4.1±1.1-fold induction, P>0.05), and claudin-12 (2.4±0.3-fold induction, P>0.05) were upregulated in GR-K426A-transfected cEND cells exposed to OGD in combination with dexamethasone to a magnitude comparable to that in untransfected cEND cells receiving dexamethasone during normoxia (Figure 2B). Accordingly, a significantly higher level of occludin immunoreactivity (122%±5%, P<0.05) was detected in hypoxic GR-K426A-transfected cells treated with dexamethasone in comparison to normoxic or deprived untransfected cEND cells (Figure 2C). In contrast, protein levels of claudin-3 and claudin-12 remained unaffected in GR-K426A-cEND cells by either exposure to OGD or dexamethasone as did the non-GC targets claudin-5 and VE-cadherin (Figure 2B–C) underlining the divergent effects of GC action on different barrier constituting proteins at the brain microvascular endothelium. Consistent with these findings, the GR was still responsive to dexamethasone after OGD in proteasome-resistant GR-K426A cells as indicated by significant mouse mammary tumor virus promoter–construct transactivation, which was similar to that observed under physiological conditions (25.9±3.3-fold induction versus 25.3±3.2-fold; P<0.05; Figure 2D).
To assure at the pharmacological level that proteasome-mediated degradation of the GR is indeed responsible for the GC insensitivity of brain microvascular endothelial cells, the proteasome inhibitor Bortezomib was added to OGD-treated cEND cells (Figure 3A).
Bortezomib (87%±7%, P<0.001) significantly mitigated the degradation of GR protein present after hypoxia (28.0%±12%, P<0.002). Dual application of dexamethasone plus Bortezomib was likewise able to attenuate hypoxia-induced reduction of GR protein (45%±7%, P<0.002) although to a lesser extent than Bortezomib monotherapy (P<0.001). This finding goes congruent with the observation of reduced GR protein expression in dexamethasone treated normoxic cEND cells (Figure 3A) and confirms previous reports on the downregulation of the GR by GC as a general mechanism of GC action7,9,28and specifically in mouse microvascular endothelial cells.26 Inhibition of the proteasome together with GC treatment seems to exert differential effects on GR levels in the presence or absence of oxygen and glucose in the cEND system. The molecular basis for this difference is unknown at present and should be addressed in future studies.
Taken together, our findings in the cEND in vitro system suggest that hypoxia and reduced energy supply impair the GC sensitivity of the brain endothelium through proteasome-induced GR degradation. Accordingly, inhibition of proteasomal pathways restored GC responsivity in this model.
Inhibition of the Proteasome Restores the BBB Stabilizing Effects of GC in Acute Ischemic Stroke In Vivo
To analyze whether interruption of proteasomal pathways is also able to reverse steroid resistance at the BBB under hypoxic conditions in vivo, the tMCAO model of acute ischemic stroke in mice was used. In this model, excessive BBB damage and successive edema formation are well-established events.16
At 6 hours and 24 hours after tMCAO, GR expression was markedly reduced in ischemic brains from vehicle-treated mice compared with Bortezomib-treated animals as revealed by Western blot analysis and immunohistochemistry (Figure 4A–B) thereby confirming the results from our in vitro experiments. Moreover, double labeling for GR and CD31 (an endothelial marker) could identify cerebral blood vessels as a major source of GR protein in the murine brain (Figure 4B).
Monotherapy with either the proteasome inhibitor Bortezomib or the GC dexamethasone or methylprednisolone, respectively, 2 hours after stroke onset did not reduce edema-corrected infarct volumes (ie, indirect infarct volumes) in comparison to untreated controls as determined by planimetry from TTC-stained brain sections (P>0.05; Figure 5A, left panel). In contrast, combined application of Bortezomib plus dexamethasone or Bortezomib plus methylprednisolone significantly attenuated corrected stroke size on Day 1 after tMCAO (77.1±19.1 mm3 versus 43.2±22.3 mm3 or 47.1±24.9 mm3, respectively; P<0.05; Figure 5A, left panel). Importantly, the reduction of infarct size was also functionally relevant because the Bederson score assessing global neurological function was significantly better in mice receiving combination therapy (3.2±0.6 versus 2.1±0.4 or 2.1±1.1, respectively, P<0.05; Figure 5A, right panel).
Next we sought to elucidate whether smaller infarct volumes and less severe functional deficits after dual application of Bortezomib and GC are indeed a consequence of preserved BBB integrity and reduced edema formation. Again, application of dexamethasone, methylprednisolone, or Bortezomib alone had no influence on BBB disruption 24 hours after tMCAO as assessed by extravasation of the vascular tracer Evan's Blue (P>0.05; Figure 5B, left panel). In line with this finding, brain edema volumes in these animals as calculated from direct and indirect infarct volumes did not differ from untreated controls (Figure 5B, right panel; P>0.05). Simultaneous treatment with Bortezomib and dexamethasone or methylprednisolone, however, stabilized the BBB (Evan's Blue extravasation: 56.6±10.0 mm3 versus 18.8±13.2 mm3 or 14.2±11.6 mm3, respectively, P<0.001; Figure 5B, left panel) and significantly attenuated the parenchymal brain water content (edema volume: 1.52±0.6 mm3 versus 0.37± 0.08 mm3 or 0.49±0.16 mm3, respectively, P<0.001; Figure 5B, right panel).
We here for the first time elucidate the molecular basis for the well-established but yet unexplained insensitivity of the BBB to GC under hypoxia. OGD reduced GR expression and activity in an in vitro model of the BBB (cEND cells) thereby abolishing the physiological barrier stabilizing effects of dexamethasone. Proteasome resistance or inhibition prevented GR degradation and preserved dexamethasone function in hypoxic cEND cells. Importantly, mice subjected to ischemic stroke and treated with the proteasome inhibitor Bortezomib in combination with GC developed significantly less brain edema and functional deficits. This indicates that inhibition of the proteasome is also effective in vivo and can overcome GC resistance at the hypoxic BBB.
Brain capillary endothelial cells are the major structural constituents of the BBB.28 Among the TJ proteins expressed at the BBB, occludin and claudin-5 have been shown to be critical for BBB maintenance.29 Degradation of these proteins is involved in BBB breakdown during multiple sclerosis, meningitis, or brain neoplasm.24 Disintegration of the BBB is also a pathological hallmark in acute ischemic stroke and is visible already several hours after stroke onset in rodents and humans.30,31 The impact of focal cerebral ischemia on the regulation of TJ proteins has only recently been addressed. Expression and subcellular localization of claudin-5 were altered under hypoxic conditions in mouse brain bEND.3 cells.32 This was accompanied by a decrease in TER. Kago and coworkers33 investigated the expression of occludin and zonula occludens-1 in isolated brain capillaries ex vivo and found downregulation after microsphere-induced cerebral embolism. Interestingly, degradation of occludin and claudin-5 in brain endothelial cells has also been reported in mice with brain edema due to acute liver failure.34 We confirm and further extend these findings by demonstrating that OGD selectively reduces the amount of the TJ proteins occludin and claudin-5 in murine brain-derived cEND cells leading to increased transendothelial permeability. In contrast, expression levels of other important TJ or adherens junction proteins such as claudin-3, claudin-12, or VE-cadherin were not affected underlining the divergent GC effects on different barrier constituting proteins at the brain microvascular endothelium.
The GR is expressed in a wide range of tissues, including the vascular endothelium, and is essential for mediating the physiological GC effects. GR expression levels closely correlate with transcriptional GR activity.25 Inhibition of GR binding to target gene promoters by antagonists or GR mutation has been shown to abolish transactivation.35,36 We found that OGD induced significant downregulation of the GR in our cEND in vitro system as did tMCAO in murine brains in vivo. Lack of functional GR prevented transcriptional modulation by GC thereby reversing genomic GC effects such as induction of TJ protein target genes leading to elevation of TER. Degradation of cerebral GR on hypoxia has also been described in vivo in neonatal rats subjected to cerebral ischemia37 and is further corroborated by our findings in the tMCAO model in adult mice. Of note, ischemia-induced downregulation of the GR by proteasomal degradation appears to be a specific feature of the cerebral endothelium because OGD in a parallel experiment did not alter GR expression or activity in hypoxic renal tubular epithelial cells (Hek293), which is in line with previous reports in other cell types.27 Notably, GR expression and GC responsivity are even heterogenous between different vascular beds as demonstrated by our previous studies using endothelial cells of neural and nonneural origin.26
We now establish the endothelial GR as a substrate of the proteasome after hypoxia. The proteasome inhibitor Bortezomib prevented destruction of GR on OGD and GR transactivation after GC binding was preserved in K426A-transfected cells resistant against proteasomal GR degradation. As a consequence, the capacity of GC to stabilize barrier functions reflected, for example, by inducing TJ proteins or increasing TER was restored. Most importantly, inhibition of the proteasome was also able to overcome GC resistance at the hypoxic BBB in vivo because mice subjected to ischemic stroke and treated with Bortezomib in combination with GC developed significantly less brain edema. Several studies have already confirmed that hypoxia is a critical mediator of protein ubiquitinylation and proteasome activity, suggesting detrimental effects of the ubiquitin–proteasome system during ischemia.12 The molecular targets of ubiquitinylation, however, are largely unknown. Only recently, hypoxia-induced poly-ubiquitination and subsequent proteasomal degradation have been ascribed to another steroid hormone receptor, namely the estrogen receptor α; this study demonstrated reduced expression of estrogen receptor α in hypoxic breast cancer cells leading to impaired cellular estrogen sensitivity.10,11
In contrast to our present results, other studies have reported beneficial effects of certain proteasome inhibitors when given as monotherapy during brain ischemia/reperfusion injury.38,39 The exact reasons for this discrepancy are unclear at present. Differences in stroke models (eg, embolic stroke versus filament occlusion of the middle cerebral artery) and animal species (rat versus mouse) as well as divergent proteasome inhibitor formulations might play a role here.
Taken together, we here identify proteasomal GR degradation and inactivation as major reasons for the paradox insensitivity of the BBB to GC under conditions of hypoxia. Blockade of the proteasome can overcome this GC insensitivity in vitro and in vivo and may therefore provide the basis for targeting stroke-induced edema formation by GC in the future.
Sources of Funding
This work was supported by grants from the Deutsche Forschungsgemeinschaft (SFB688, A05 to C.F., A13 to C.K., and B01 to G.S.) and from the European Union (HEALTH-F2-2009-241778) to C.F. and C.K. This work was also supported by grants from the Bundesministerium für Bildung und Forschung (BMBF01 EO1004), to C.F.
The K426A glucocorticoid receptor construct was a generous gift from Dr John Cidlowski, Laboratory of Signal Transduction, North Carolina, USA. We thank Katharina Mattenheimer, Melanie Glaser, and Bianca Schneiker for excellent technical assistance. We are thankful to Dr P.M. Kane, Department of Biochemistry and Molecular Biology, SUNY Upstate Medical University of Syracuse, Syracuse, NY, for critical reading of the manuscript.
The online-only Data Supplement is available at http://stroke.ahajournals.org/cgi/content/full/STROKEAHA.110.592238/DC1.
- Received June 7, 2010.
- Revision received October 27, 2010.
- Accepted November 3, 2010.
- © 2011 American Heart Association, Inc.
- Norris JW
- Norris JW,
- Hachinski VC
- Poungvarin N
- Bardutzky J,
- Schwab S
- Wallace AD,
- Cidlowski JA
- Meller R
- Kleinschnitz C,
- Schwab N,
- Kraft P,
- Hagedorn I,
- Dreykluft A,
- Schwarz T,
- Austinat M,
- Nieswandt B,
- Wiendl H,
- Stoll G
- Kleinschnitz C,
- Stoll G,
- Bendszus M,
- Schuh K,
- Pauer HU,
- Burfeind P,
- Renne C,
- Gailani D,
- Nieswandt B,
- Renne T
- Austinat M,
- Braeuninger S,
- Pesquero JB,
- Brede M,
- Bader M,
- Stoll G,
- Renne T,
- Kleinschnitz C
- Bederson JB,
- Pitts LH,
- Tsuji M,
- Nishimura MC,
- Davis RL,
- Bartkowski H
- Narasimhan P,
- Liu J,
- Song YS,
- Massengale JL,
- Chan PH
- Leonard MO,
- Godson C,
- Brady HR,
- Taylor CT
- Garside H,
- Waters C,
- Berry A,
- Rice L,
- Ardley HC,
- White A,
- Robinson PA,
- Ray D
- Zhang L,
- Zhang ZG,
- Buller B,
- Jiang J,
- Jiang Y,
- Zhao D,
- Liu X,
- Morris D,
- Chopp M