GW9662

PPARγ activation suppresses the expression of MMP9 by downregulating NF-κB post intracerebral hemorrhage

Abstract

PeroXisome proliferator-activated receptor-gamma (PPARγ) is critical in protecting against inflammatory and oXidative stresses post brain injury. We have previously reported that rosiglitazone, an agonist of PPARγ, was effective to prevent microglia from apoptosis and ameliorate neuronal injuries post intracerebral hemorrhage (ICH) with suppression of matriX metalloproteinase-9 (MMP9) expression. However, molecular mechanisms linking how PPARγ decreases MMP9 remain unknown. Here, we hypothesize that PPARγ downregulates MMP9 expression post hemorrhage by inhibiting nuclear factor kappa B (NF-κB), an upstream regulator of MMPs gene and also key transcription factor involved in the control of immune and neuroinflammatory responses. We found both in vivo and in vitro that PPARγ was significantly downregulated post ICH with prominent increases of NF-κB and MMP9. Activation of PPARγ using rosiglitazone decreased the expression of both NF-κB and MMP9, while reversed effects were observed when administrating the PPARγ antagonist GW9662. Besides, inhibiting NF-κB by JSH-23 also suppressed the expression of MMP9, with only limited effect on PPARγ. Further studies revealed prominent colocalizations of NF-κB with PPARγ and MMP9, respectively. Finally, direct interactions of NF-κB with PPARγ and MMP9 gene were also confirmed, respectively, by protein and chromatin immunoprecipitations. These results suggested a role of NF-κB in mediating the reduction of MMP9 by PPARγ, potentially providing a new therapeutic target for brain hemorrhage.

1. Introduction

Intracerebral hemorrhage (ICH) is a common cause of morbidity and disability, but to date specific therapy is still lacking. Common factors leading to ICH include hypertension, amyloid angiopathy, tumors, hemorrhagic transformation of an ischemic stroke, vasculitis and vascular malformations [1]. Inflammation is often present concurrently post ICH, which might serve as a neuroprotective role at the initial stage but subsequently aggravate brain injury when the inflammatory response is out of control.

MatriX metalloproteinase-9 (MMP9), one of downstream effectors of PPARγ, shows upregulated expression post ICH [7,8]. Clinical evidences have suggested independently association between MMP9 gene variation and symptomatic intracerebral hemorrhage [9]. Besides, MMP9
inhibitors were also proved effective for the neuroprotection in ICH [10]. However, molecular mechanisms linking how PPARγ regulatesligand-activated transcription factor which is well-known to be involved in the lipid metabolism and cell proliferation. Recent studies have suggested roles of PPARγ in anti-inflammatory process and clearance of hematoma in central nervous system disorders [2,3]. We and others have previously reported that rosiglitazone, a specific agonist of PPARγ,(NF-κB), which was reported to be suppressed by PPARγ [12], is an important regulator of MMPs gene expression [11]. Here, we hypothe- sized and demonstrated that PPARγ suppresses the expression of MMP9 post ICH by directly acting on NF-κB.

2. Materials and methods
2.1. Animals

SD rats were obtained from the Animal experiment center of Guizhou Medical University. All rats were housed under standard conditions with 12 h light/dark cycle, food and water ad libitum. Adult male rats of 8~12 weeks of age, weighing about 180~220 g were used for experiments. All animal experiments were approved by the Animal Care and Use Com- mittee of Guizhou Medical University.

2.2. Cell culture

BV2 cell was obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). All cells were cultured in the Dulbecco’s modified Eagle’s medium (Gibco) supplemented with 10 % FBS (Gibco) under a condition of 37 ◦C, 5% CO2.

2.3. Cerebral hemorrhage model

Rats were anesthetized using 10 % Chloral hydrate (350 mg/kg), and fiXed in a stereotaxic instrument (RWD, Shenzhen, China). The scalp was sterilized using iodophors and 75 % ethanol and incised along the skull midline. The surface of skull was cleaned by 3% H2O2, then a hole was drilled at anterior-posterior 0.6 mm, lateral-medial 3 mm (from the Bregma). About 80 μL autologous blood was collected from the tail ar- tery, and a total of 50 μL blood was injected into the cerebral caudate nucleus (-6.0 mm from the skull) using an automatic microinjection system (World Precision Instruments, USA) at a rate of 10 μL/min. The needle syringe was left in place for 10 min before withdrawal. The skin was sutured and sterilized with iodophors. The neurological symptoms of rats were evaluated by the Longa scales [13] 30 min following ana- lepsia, In brief, rat behaviors were scored by: 0 – no neurological find- ings, 1 – failure to extend left forepaw fully, 2 – circling to the left, 3 – falling to the left, 4 – no spontaneous walking and being depressed. Rats showing prominent hematoma in the brain, and with the Longa score 2 were considered hemorrhage, the else were excluded from analysis.

2.4. Immunostaining

For immunohistochemistry, rats were killed with overdose of 10 % Chloral hydrate and intracardially perfused with saline followed by 4 % paraformaldehyde (PFA). Rat brains were removed, post-fiXed in 4 % PFA and then cryoprotected in 30 % sucrose. Brain sections of 30 μm
were sliced using a cryostat microtome (Leica). brain sections were pre- treated with 3 % H2O2 at 37℃ for 30 min. For immunofluorescence,
cultured cells were harvested on ice and fiXed using 4 % PFA.

Free-floating sections or fiXed cells were washed in PBS, blocked in 5% bull serum albumin and 0.3 % Triton x-100 for 2 h, and then incu-
bated with primary antibodies at 4 ◦C for 24 h: PPARγ (1:400, Abcam), NF-κB (1:400, Abcam) and MMP9 (1:500, Abcam), washed in PBS and
then incubated with secondary antibodies (1:400, Invitrogen). For immunohistochemistry staining, immunoreactions were developed using a DAB-staining kit (ZSGB-Bio), dehydrated through graded ethanol series, and sealed with neutral balsam. For immunofluores- cence, cells were co-stained with DAPI (1:2000, Beyotime) at room temperature for 2 h. Images were taken by a microscope (Olympus).

2.5. Western blotting

Rats brains were removed and tissues were isolated on ice. Cultured cells were harvested 1 h after thrombin treatment. Brain or cell samples were homogenized using RIPA lysis buffer and then centrifuged at 12,000 rpm for 30 min. The supernatant was collected and miXed with loading buffer. Proteins were separated in SDS-PAGE gels, transferred onto nitrocellulose membranes. The membranes were then blocked with 5% BSA for 2 h, and incubated with primaries and secondary antibodies: PPARγ (1:2000, Abcam), NF-κB (1:2000, Abcam) and MMP9 (1:2000,
Abcam) and β-actin (1:1000, Abcam). Blots were detected using a luminol reagent (Beyotime).

2.6. Real-time quantitative reverse transcription PCR (RT-qPCR)

Cell were collected, and total RNA was extracted using a TRIzol re- agent (Life Technologies, Carlsbad, CA). Samples were assayed using the Biosystems QuantStudio 3 Systems (Thermo Fisher). The system including 3 mM MgCl2, 0.5 μM primers each, 2 μL SYBR Green PCR master miXes and 2 μL cDNA. Primers were as follows: PPARγ-forward 5′- AGGGCGATCTTGACAGGAAA -3′, PPARγ-reverse5′- CGAAACTGG- CACCCTTGAAA -3′; NF-kB-forward5′- GCCAAAGAAGGACACGAC -3′, NF-kB-reverse5′- ATCACCCTCCAGAAGCAG -3′; MMP9-forward5′- AAAGGCCATTCGAACACCAC -3′, MMP9-reverse5′- GGATGA- CAATGTCCGCTTCG -3′; GAPDH-forward5′- ATGGGTGTGAACCAC- GAGA -3′, GAPDH-reverse5′- CAGGGATGATGTTCTGGGCA -3′. Data
was normalized by the expression level of house-keeping gene GAPDH.

2.7. Co-immunoprecipitation

BV2 cells were homogenized on ice in the RIPA buffer (Beyotime) at 4 ◦C and then centrifuged at 12,000 g for 15 min. A total of 200 μg total proteins were incubated with NF-kB antibody for 8 h followed by in- cubation with protein G agarose (Beyotime) for 2 h at 4 ◦C on rotation.The agarose beads were then washed and proteins were resuspended in 40 μL of SDS-containing loading buffer, denatured at 95 ◦C for 10 min, and finally analyzed by Western blotting.

2.8. Chromatin immunoprecipitation

Cells were incubated with 1 % formaldehyde for 10 min, washed and then scraped into cold PBS with protease inhibitors. After centrifugation, the pellet was resuspended in a buffer containing 20 mM HEPES, 25 % glycerol, 420 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA and protease inhibitors cocktail, incubated on ice for 20 min, and then centrifuged. The pellet was resuspended and sonicated for 5 s, and Triton buffer was added. An aliquot was reserved as the input, and the remainder was divided into immunoprecipitation with IgG or NF-kB (Abcam) antibody followed by incubation with protein G beads. Samples were washed in Triton buffer, SDS buffer was added and then samples were subjected to vortex and incubated at 65 ◦C overnight to reverse cross-linking. DNA was isolated using phenol/chloroform extraction and re-suspended in distilled H2O. Primers used for PCR were: MMP9-forward GACGGTCTCCTCCATCAATA, MMP9-reverse CTTTAGTCCAGCTT- CATCCC. PCR products were finally analyzed in 2% agarose gel electrophoresis

2.9. Statistical analysis

Data were presented as means ± SEM, and analyzed or plotted using GraphPad Prism5. One-way ANOVA with Turkey’s post hoc tests were used, with p < 0.05 considered as statistically significant. 3. Results We firstly examined the expression of PPARγ, NF-κB and MMP9 in a rat model of cerebral hemorrhage. In consistent with previous findings, PPARγ was significantly downregulated post hemorrhage with a prom- inent increase of MMP9 measured by both immunohistochemistry (Fig. 1A and B) and Western blotting (Fig. 1C and D). Besides, we found here that the expression of NF-κB was also upregulated in the brain of rats with cerebral hemorrhage. To test whether the increases of NF-κB and MMP9 were downstream effects of PPARγ reduction, we treated hemorrhage rats with rosiglitazone, an agonist of PPARγ. Rosiglitazone significantly reversed the decrease of PPARγ as well as the increase of NF-κB and MMP9 (Fig. 1). By contrast, PPARγ antagonist GW9662 produced opposite ef- fects on NF-κB and MMP9 (Fig. 1). These results suggested that PPARγ was an upstream regulator of NF-κB and MMP9. Fig. 1. Opposite changes in the expression of PPARγ with NF-κB and MMP9 in the rat model of brain hemorrhage. Immunohistochemical staining (A and B) and Western blotting (C and D) results showing that PPARγ increased in hemorrhage with the reduction of NF-κB and MMP9, while activation or suppressing of PPARγ using rosiglitazone and GW9662, respectively, exerted opposite effects on those proteins. Inhibiting NF-κB by JSH-23 downregulated NF-κB and MMP9 but showed limited effect on PPARγ. One-way ANOVA with Turkey’s post hoc tests, * p < 0.05, compared with the sham group; # p < 0.05, compared with the vehicle group; n = 5 rats in each group. Fig. 2. Direct interaction of NF-κB with PPARγ and MMP9 in the BV2 cell model of hemorrhage. (A and B) EXpression changes of PPARγ, NF-κB and MMP9 proteins. One-way ANOVA with Turkey’s post hoc tests, * p < 0.05, compared with the normal control (NC) group; # p < 0.05, compared with the vehicle group; n = 3 cell wells in each group. (C) mRNA changes of PPARγ, NF-κB and MMP9 proteins detected by RT-qPCR. Data were normalized to the level of GAPDH gene. One-way ANOVA with Turkey’s post hoc tests, * p < 0.05, compared with the NC group; # p < 0.05, compared with the vehicle group; n = 3 cell wells in each group. (D) Prominent co-localization of NF-κB with PPARγ and MMP9, respectively in BV2 cells treated with thrombin. (E) The direct interaction between NF-κB and PPARγ was detected in co-immunoprecipitation. (F) Immunoprecipitation of MMP9 gene fragments by NF-κB antibody in BV2-thrombin models. To further confirm the regulation of NF-κB and MMP9 by PPARγ in cerebral hemorrhage, rats were administrated with JSH-23, a specific inhibitor of NF-κB. We found that JSH-23 produced similar effects on NF-κB and MMP9 as rosiglitazone post hemorrhage, but showed limited effect on PPARγ (Fig. 1). Taken together, these data indicated that the reduction of PPARγ in cerebral hemorrhage resulted in the upregulation of MMP9 through disinhibiting NF-κB. We next examined the interaction of NF-κB with PPARγ and MMP9, respectively, in BV2 cells treated with thrombin, a well-accepted in vitro model mimicking the microglia activation and neuroinflammation in brain hemorrhage. In consistent with results from in vivo studies, the expression of PPARγ reduced in the thrombin-treated cells, with prominent increases of NF-κB and MMP9, while rosiglitazone significantly reversed those changes (Fig. 2A–C). In addition, GW9662 showed opposite effects to rosiglitazone, and JSH-23 prevented the increase of NF-κB and MMP9 post hemorrhage but showed limited effects on PPARγ (Fig. 2A–C). Importantly, we observed prominent co-localizations of NF-κB with PPARγ and MMP9, respectively, in BV2 cells (Fig. 2D). Furthermore, the direct interaction between NF-κB and PPARγ was evidenced by co- immunoprecipitation experiments. The NF-κB-PPARγ binding was weaken in hemorrhage, which was oppositely regulated by rosiglitazone and GW9662 (Fig. 2E). Moreover, we also examined the regulation of MMP9 gene by NF-κB. NF-κB was predicted to bind the sequence 5′- CAGTTTCCCC-3′ of the MMP9 gene promoter, and this direct binding was evidenced by the chromatin immunoprecipitation (CHIP) of MMP9 gene using NF-κB antibody (Fig. 2F). Similarly, the regulation of MMP9 gene expression by NF-κB changed with the administration of rosiglitazone and GW9662. 4. Discussion Taken together, we found in the present study that the decrease of PPARγ post ICH was negatively associated with NF-κB and MMP9, and the link of PPARγ with MMP9 by NF-κB was evidenced by specific drugs and immunoprecipitation assays. In general, our findings were in line with previous hypothesis that the suppression of MMP9 gene expression by PPARγ was resulted from the inhibition of NF-κB [11,14]. In response to the enhanced inflammatory responses post ICH [1,15], we observed here that PPARγ was significantly decreased in both in vivo and in vitro models of ICH. Indeed, PPARγ has been reported to promote the clearance of hematoma in a rat model of ICH, possibly through modulating the haptoglobin-hemoglobin-CD163 signaling pathway, and PPARγ agonist was effective in reducing the hematoma volume, brain edema and hemoglobin after ICH [2]. Besides, PPARγ has been also reported to be involved in the activation of microglia and production of TNF-α in a progressive Parkinson’s disease model [6]. Not only that, the neuro- protective role of Maraviroc, a CC-chemokine receptor 5 antagonist, against the trauma and hemorrhage-induced hepatic injury through is also dependent on PPARγ cascades [16]. We found here that activation of PPARγ using rosiglitazone also increased its expression level, while the PPARγ antagonist GW9662 significantly decreased the expression of PPARγ, these results were in consistent with our and others previously findings that rosiglitazone was effective in inhibiting microglia activation and ameliorating neuroinflammation [4,5].

MMP9 has long been reported to be involved in the pathogenesis of ICH. Numerous studies have revealed the increases of matriX metal- loproteinases, specifically the MMP9 levels post stroke, with disruptions of the blood brain barrier, increased risk of hemorrhagic complications and worsened outcome [17]. Clinical evidences suggested that, adjust- ing for major clinical determinants, only the MMP9 gene variation proved independently associated with death and symptomatic intrace- rebral hemorrhage [9]. However, another case-control study in Chinese Han population reported only nonsignificant association between the MMP9 gene variation and ICH susceptibility [18]. The discrepancy might be due to the different subpopulation surveyed. Despite the dual role and temporal profile of MMP9, its inhibitors still represent poten- tially effective treatments for the neuroprotection in ICH [8]. For example, edaravone, a free radical scavenger was evidenced to have dose-dependently suppressed the activity and mRNA expression of MMP and ameliorated ICH [19]. In the present study, we found that the expression of MMP9 is upregulated both in vivo and in vitro ICH models, and shows a negative correlation with the PPARγ either under the regulation with PPARγ agonist or antagonist. In fact, the inhibition of MMP9 by PPARγ activators has also been revealed by others in human bronchial epithelial cells [14]. We have further found that the sup- pression of MM9 by PPARγ was dependent on the inhibition of NF-κB. Indeed, PPARγ has been widely demonstrated to be important in the modulation of NF-κB transcriptional activity, and both the two proteins regulate inflammatory and hypoXia responses [20,21]. Not only in the ICH, the suppression of NF-κB by the PPAR agonist troglitazone was also involved in such as cancer cell growth and anti-virus infection [12,22]. Furthermore, NF-κB play a role in the regulation of MMP9 gene expression. It was reported that the matriX metalloproteinase-9 expression induced by relaxin, a hormone secreted in the ovary by the corpus luteum, is associated with activation of the NF-κB pathway [11,23]. Here, we demonstrated the direct interactions between PPARγ and NF-κB, NF-κB and MMP9, respectively, and reported their associations in ICH.

In conclusion, we found in the present study a role of NF-κB in mediating the reduction of MMP9 by PPARγ post ICH. Our data revealed a new mechanism underlying the regulation of MMP9 expression, and suggested a new therapeutic target for the brain hemorrhage.