Therapeutic targeting of Krüppel-like factor 4 abrogates microglial activation
© Kaushik et al; licensee BioMed Central Ltd. 2012
Received: 15 December 2011
Accepted: 19 March 2012
Published: 19 March 2012
Skip to main content
© Kaushik et al; licensee BioMed Central Ltd. 2012
Received: 15 December 2011
Accepted: 19 March 2012
Published: 19 March 2012
Neuroinflammation occurs as a result of microglial activation in response to invading micro-organisms or other inflammatory stimuli within the central nervous system. According to our earlier findings, Krüppel-like factor 4 (Klf4), a zinc finger transcription factor, is involved in microglial activation and subsequent release of proinflammatory cytokines, tumor necrosis factor alpha, macrophage chemoattractant protein-1 and interleukin-6 as well as proinflammatory enzymes, inducible nitric oxide synthase and cyclooxygenase-2 in lipopolysaccharide-treated microglial cells. Our current study focuses on finding the molecular mechanism of the anti-inflammatory activities of honokiol in lipopolysaccharide-treated microglia with emphasis on the regulation of Klf4.
For in vitro studies, mouse microglial BV-2 cell lines as well as primary microglia were treated with 500 ng/mL lipopolysaccharide as well as 1 μM and 10 μM of honokiol. We cloned full-length Klf4 cDNA in pcDNA3.1 expression vector and transfected BV-2 cells with this construct using lipofectamine for overexpression studies. For in vivo studies, brain tissues were isolated from BALB/c mice treated with 5 mg/kg body weight of lipopolysaccharide either with or without 2.5 or 5 mg/kg body weight of honokiol. Expression of Klf4, cyclooxygenase-2, inducible nitric oxide synthase and phospho-nuclear factor-kappa B was measured using immunoblotting. We also measured the levels of cytokines, reactive oxygen species and nitric oxide in different conditions.
Our findings suggest that honokiol can substantially downregulate the production of proinflammatory cytokines and inflammatory enzymes in lipopolysaccharide-stimulated microglia. In addition, honokiol downregulates lipopolysaccharide-induced upregulation of both Klf4 and phospho-nuclear factor-kappa B in these cells. We also found that overexpression of Klf4 in BV-2 cells suppresses the anti-inflammatory action of honokiol.
Honokiol potentially reduces inflammation in activated microglia in a Klf4-dependent manner.
Microglia, the resident macrophages of the central nervous system (CNS), are amongst the most important cells of innate immunity within this organization and become activated in response to pathological stimuli, such as endotoxins and other insults to the CNS [1, 2]. Activated microglia respond by secreting proinflammatory cyto-chemokines such as macrophage chemoattractant protein-1 (MCP-1), TNF-α and IL-6 as well as proinflammatory enzymes such as inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (Cox-2) [3–6]. These exaggerated responses by microglia can result in neuronal damage  that can be detrimental to normal CNS functions . Microglial inflammation is also observed in several neurodegenerative diseases including Alzheimer's disease , Parkinson's disease and multiple sclerosis [9, 10]. Understanding the molecular mechanisms of microglial activation and targeting the key players of inflammation within microglia are essential in designing therapeutic strategies to deal with inflammatory conditions of the CNS.
Phospho-nuclear factor-κB (pNF-κB) is one of the most studied transcription factors for inflammatory pathways and is attributed with the upregulation of these proinflammatory cytokines [5, 6]. In addition to pNF-κB, zinc finger transcription factor Krüppel-like factor 4 (Klf4) has been reported to play a key role in endothelial- and macrophage-mediated inflammation [11–16] alongside its role in cellular growth and differentiation in the epithelial lining of the gut and skin [17, 18]. More recently, our laboratory has shown that Klf4 is important in mediating the upregulation of these proinflammatory cytokines in microglial cells . We observed a significant increase in the expression of total as well as nuclear Klf4 in microglial cells upon lipopolysaccharide (LPS) treatment in a time-dependent manner. Upon LPS treatment to BV-2 cells, Klf4 also interacted with the iNOS promoter, which was confirmed by electrophoretic mobility shift assay studies. During this study, we also observed that siRNA-mediated knockdown of Klf4 resulted in a significant reduction in the levels of these cytokines and proinflammatory enzymes. In addition, we found that Klf4 potentially interacted with pNF-κB in response to LPS, suggesting that the two transcription factors may be equally important in regulating the downstream proinflammatory markers of microglial activation. It is quite likely that the proinflammatory actions of Klf4 are associated with its interaction with pNF-κB. Since most studies have focused on the role played by pNF-κB in mediating inflammation, our main focus was to evaluate the inflammatory potential of Klf4. Our previous studies confirmed that Klf4 is a key transcription factor in modulating inflammation. We therefore wanted to investigate whether Klf4 can be a potential target for anti-inflammatory agents such as honokiol (HNK).
HNK is a biphenolic anti-inflammatory agent isolated from the barks of stems and roots of various Magnolia species and used as a traditional herbal medicine . HNK possesses neuroprotective , anti-cancerous [22, 23] and anti-inflammatory properties [24, 25]. There is plenty of literature about the anti-inflammatory roles of HNK in dendritic cells , macrophages  and several other cellular systems [28–32]. It is reported that the anti-inflammatory actions of HNK in LPS-mediated inflammation act via p38, extracellular signal-regulated kinases 1/2 (ERK1/2) and c-Jun N-terminal kinases 1/2 (JNK1/2) pathways . Recently, a study focusing on the use of HNK on gliosarcoma revealed that it can easily cross the blood-brain barrier and blood-cerebrospinal fluid barrier , which makes it a drug of choice for CNS disorders. Moreover, HNK does not antagonize the binding of LPS to its receptors and does not alter the surface expression of CD14 or toll-like receptor 4 . In addition, HNK also inhibits cytotoxicity induced by LPS in RAW264.7 macrophage cells . Whether HNK treatment modulates Klf4 expression upon LPS treatment and decreases microglial activation stimulation has never been studied. Therefore, this study focuses on understanding the role of HNK in modulating the expression of Klf4 as well as LPS-induced neuroinflammation. Our current study shows that HNK downregulates the expression of Klf4 and pNF-κB and suppresses the overall expression of iNOS and Cox-2 as well as MCP-1, IL-6 and TNF-α in LPS-treated microglia as well as in mouse brain. Our study highlights Klf4 as a potential target for therapeutic activities of HNK.
Mouse microglial cell line BV-2 was originally obtained from Dr Steve Levison, University of Medicine and Dentistry, New Jersey, USA and maintained in the laboratory as previously described . Briefly, BV-2 cells were grown at 37°C in DMEM supplemented with 5% sodium bicarbonate (NaHCO3), 10% FBS, penicillin at 100 units/mL and streptomycin at 100 μg/mL. All reagents related to cell culture were obtained from Sigma Aldrich, St. Louis, MO, USA unless otherwise stated. Primary mixed glial cultures were prepared from postnatal day 0-2 (P0-P2) mouse brains as described previously . The cells were plated into 75-cm2 tissue culture flasks at a density of 2 × 105 viable cells/cm2 in minimum essential medium supplemented with 10% FBS, penicillin at 100 units/mL, streptomycin at 100 μg/mL, 0.6% glucose and 2 mM glutamine. Media were changed every two to three days after plating. On day 12, when the mixed glial culture was confluent, the flasks were shaken on an Excella E25 orbital shaker (New Brunswick Scientific, Edison, NJ, USA) at 250 rpm for 60 to 75 minutes to dislodge microglial cells which were then allowed to adhere in bacteriological petri dishes for 60 to 90 minutes. The adherent cells were scraped, centrifuged and plated in chamber slides at a density of 8 × 104 viable cells/cm2, and incubated at 37°C until the treatment.
For in vitro experiments, treatment involved a pre-incubation with HNK (dissolved in 1X PBS and dimethyl sulfoxide (DMSO) in a 1:1 ratio; Sigma Aldrich) for 3 hours in serum-free media and then co-treatment with LPS (Sigma Aldrich) for an additional 6 hours for reactive oxygen species (ROS) and nitric oxide (NO) generation for a luciferase assay, or for 12 hours for immunoblotting, immunocytochemistry and cytokine bead array (CBA) analysis, under serum-free conditions. The control cells received only serum-free culture media and a mixture of DMSO and 1X PBS in a 1:1 ratio at the same volume as HNK.
For in vivo experiments, three groups of 6- to 8-week-old BALB/c mice were injected intraperitoneally with 2.5 or 5 mg/kg body weight of HNK dissolved in 1X PBS and DMSO in a 1:1 ratio over 12 hours while the control group received the sham mixture (1X PBS and DMSO in a 1:1 ratio). After 12 hours of HNK administration, mice were injected intraperitoneally with 5 mg/kg body weight of Salmonella enterica LPS dissolved in 1X PBS for an additional 24 hours. Control animals received the same volume of 1X PBS alone. All the animals were killed after 24 hours of treatment for tissue or protein. All experiments were performed according to the protocol approved by the Institutional Animal Ethics Committee.
We examined the effects of LPS and HNK + LPS on microglial cells by determining the levels of ROS as well as NO. Intracellular ROS generation in control and treated cells was assessed using the cell permeable, non-polar hydrogen peroxide-sensitive dye 5-(and-6)-chlromethyl-2', 7'- dichlorodihydrofluorescein diacetate (Sigma Aldrich) and the mean fluorescent intensities (MFIs) were measured on the FL-1 channel on a fluorescence-activated cell sorting (FACS) Calibur flow cytometer (Becton Dickinson, Franklin Lakes, NJ, USA) as described previously . For NO production, nitrite generation by BV-2 cells was used as an indicator of NO release and measured by the Griess reaction as described previously . The readings were taken at 540 nm using a microplate spectrophotometer (Bio-Rad, Gladesville, NSW, Australia) and the concentration was calculated from a sodium nitrite standard reference curve.
For in vitro studies, untreated control and treated BV-2 cells were lysed in buffer containing 1% Triton-X-100, 10 mM Tris(hydroxymethyl)aminomethane-Cl (pH 8.0), 150 mM sodium chloride, 0.5% octylphenoxypolyethoxyethanol (Nonidet P-40), 1 mM ethylenediaminetetraacetic acid, 0.2% ethylene glycol tetraacetic acid, 0.2% sodium orthovanadate and protease inhibitor cocktail (Sigma Aldrich).
For in vivo studies, the whole brain tissues, barring the olfactory lobes and the cerebellum, from three animals after 24 hours of LPS administration and untreated mock controls were dissected and homogenized in 1X PBS with protease inhibitor. Protein concentrations were determined using the bicinchoninic assay method and 30 μg of each protein sample was electrophoresed on a 7.5% to 10% SDS-PAGE and transferred onto a nitrocellulose membrane. After blocking in 5% skimmed milk in 1X PBS-Tween-20 (1X PBST) for 4 hours at room temperature, the blots were incubated in 1X PBST overnight at 4°C with gentle agitation with either rabbit anti-Klf4 (1:500; Chemicon International, Temicula, CA, USA), anti-iNOS (1:1,000; Millipore, Billerica, MA, USA), Cox-2 (1:1,000; Millipore), pNF-κB (1: 1,000; Ser-536; Cell Signaling Technology, Danvers, MA, USA) and anti-β-actin (1:10,000; Sigma Aldrich). After five washes in 1X PBST, blots were then incubated with goat anti-rabbit horseradish peroxidase (1:5,000; Vector Laboratories, Burlingame, CA, USA) for 1.5 hours, with agitation. The blots were developed by using chemiluminescence reagent (Millipore) in ChemiGenius Bioimaging System (Syngene, Cambridge, UK). The images were captured and analyzed using the GeneSnap and GeneTools software, respectively, from Syngene. The protein levels were normalized to β-actin levels.
Luciferase reporter gene constructs pGL2-iNOS (kindly provided by Dr Mark A. Feinberg and Dr Mukesh K. Jain, Harvard Medical School, Boston, MA, USA; the construct was originally made by Dr Mark A. Perella, also Harvard Medical School) [12, 34] and pCOX301 (a kind gift from Dr Manikuntala Kundu, Bose Institute, Kolkata, India)  were used, containing mouse iNOS promoter region (-1485 to +31 bp) and Cox-2 promoter region (-891 to +9 bp) cloned upstream of the luciferase reporter gene in the pGL2-promoter and pGL2-basic vector, respectively (Promega, Madison, WI, USA). The luciferase assay was carried out using the luciferase assay kit (Promega) according to the manufacturer's protocol and as described previously . The reading was taken using a Sirius single tube luminometer (Berthold Detection Systems GmBH, Bad Wildbad, Germany). The luciferase units were measured as relative luciferase units and these values were normalized to the amount of protein present in the sample.
Oligonucleotide primers were designed to amplify the coding sequence of full-length mouse Klf4 cDNA yielding a 1452 bp (1.4 kb) product. The oligonucleotide primers were as follows: forward primer: 5'-ATGAGGCAGCCACCTGGCGA-3'; reverse primer: 5'- TTAAAAGTGCCTCTTCATGTGTAA-3'. The amplified product was then electrophoresed onto 0.8% agarose and the 1452 bp fragment was gel extracted using a Gel Extraction Kit (Qiagen, Hilden, Germany). The fragment was then inserted into the pcDNA3.1/V5 His TopoTA expression vector (Invitrogen, Carlsbad, CA, USA) to generate a pcDNA3.1-Klf4 (K-10) construct which was sequenced commercially (ILS Bioservices, Manesar, India).
For immunocytochemical studies, mouse primary microglial cells were fixed in 4% paraformaldehyde for 20 minutes at 25°C post-treatment and then incubated in blocking solution for 1 hour at 25°C. The cells were then incubated with rabbit anti-Klf4 (1:250; Millipore) overnight at 4°C. Anti-rabbit Alexa fluor 594 secondary antibody (1:1,000; Vector Laboratories) was used as the secondary antibody and mounting was carried out with mounting medium containing 4,6-diamidino-2-phenylindole (DAPI; Vector Laboratories). Images were captured using a Zeiss Apotome microscope (20× magnification; Zeiss, Hamburg, Germany). For immunohistochemistry, LPS-treated and age-matched control BALB/c mice in sets of three were perfused transcardially with 1X PBS and isolated whole brains were fixed with 4% paraformaldehyde in 1X PBS for 24 hours at 4°C. These were then kept in 30% sucrose for about 24 to 48 hours at 4°C or until the tissue was immersed at the bottom of the tube. The brains were then sectioned on a cryostat and stained for a marker of activated microglia, Iba1, as described previously . Briefly, microglia were labeled with rabbit anti-Iba1 antibody (1:250; Wako, Osaka, Japan) by incubating the sections overnight at 4°C. After five washes with 1X PBS, the sections were incubated with horse anti-mouse fluorescein isothiocyanate (1:250; Vector Laboratories) for 1.5 hours. The slides were then mounted with mounting medium containing DAPI and images were captured using a Zeiss Apotome microscope (40× magnification; Zeiss).
The CBA kit (BD Biosciences, Franklin Lakes, NJ, USA) was used for the quantitative measurement of cytokine levels in control, LPS and LPS + HNK conditions. Using 50 μL of mouse inflammation standard and sample dilutions, the assay was performed according to the manufacturer's instructions and analyzed on the FACSCalibur flow cytometer (Becton Dickinson). Analysis was performed using CBA software that allows the calculation of cytokine concentrations in unknown lysates as described previously [19, 36]. The levels of cytokines were measured in picograms per milliliter.
Data are represented as the mean ± standard error of the mean (SEM) from at least three independent experiments. The data generated were analyzed statistically by paired two-tailed Student's t-test. P < 0.01 and < 0.05 were considered significant.
We also estimated nitrite production in order to measure NO release by LPS-stimulated BV-2 cells upon HNK treatment. Similar to our observations with ROS production, LPS increased generation of nitrite by more than 2.5-fold in comparison to the untreated control condition (2E), which was reduced by HNK treatment. Although the reduction was not significant with 1 μM HNK dose, we observed a significant decrease in NO levels with 10 μM HNK dose when compared with the LPS alone-treated condition (2E). These findings confirm that HNK reduces the oxidative stress generated in BV-2 microglial cells upon LPS stimulation. Reduction in NO production also supports our observation of a decrease in iNOS levels upon HNK treatment to LPS-stimulated cells.
Inflammation of the CNS is marked by the excessive production of numerous proinflammatory cyto-chemokines including TNF-α, MCP-1 and IL-6 in response to LPS [3, 4]. Previous reports from our laboratory have suggested that Klf4 is involved in mediating microglial activation and the regulation of proinflammatory mediators in response to LPS . This study revealed HNK for the first time as an important factor regulating inflammation in microglial cells in addition to pNF-κB, another key player in inflammatory pathways. The focus of our study was to determine whether HNK, a biphenolic compound that can easily cross the blood-brain barrier  and also has a neuroprotective role in the brain , can reduce microglial activation and whether these activities are mediated by regulating Klf4. For our experiments, we induced inflammation in BV-2 cells and BALB/c mice using LPS and then treated them using HNK, whose anti-inflammatory role in the CNS has not previously been studied. Our studies show that HNK significantly reduces microglial activation and secretion of proinflammatory cyto-chemokines in response to LPS; that HNK downregulates both Klf4 and pNF-κB in vitro as well as in vivo; and that overexpression of Klf4 suppresses the anti-inflammatory properties of HNK.
HNK also suppresses neuroinflammation stimulated by LPS in the mouse brain and induces the activation of microglia within the brain cortex of LPS-stimulated mice. Microglia in HNK + LPS mice resemble the resting state morphology, clearly showing that HNK does reduce neuroinflammation. It would be of interest to understand the role played by HNK in affecting other cell types of the brain as well as its underlying molecular mechanisms as other cell types, including astrocytes, also have been reported to have neuromodulating properties.
Oxidative stress is an important feature of inflammation and the role of HNK for its anti-oxidant properties have been studied in detail in the peripheral system and there is plenty of literature on its mode of action. Studies on collagen-induced arthritis in mice have shown that HNK suppresses production of TNF-α and IL-1β along with a reduction in oxidative damage . Studies on ischemic injury in rats have shown that HNK can act as a potent ROS inhibitor by potentially modulating the enzymes in ROS production, such as myeloperoxidase, Nicotinamide adenine dinucleotide phosphate oxidase and Glutathione peroxidase in neutrophils [50, 51]. In cerebellar granule neuronal cells, HNK was found to reverse hydrogen peroxide-mediated neurotoxicity . Similarly, HNK administration resulted in a significant suppression of oxidative damage in mouse brain treated with N-methyl-d-aspartic acid . In consensus with these studies, we report that HNK can significantly reduce the production of ROS in LPS-stimulated microglial cells. However, the detailed mechanisms of ROS inhibition by HNK in activated microglia is not clearly understood. It would be interesting to determine the mechanisms by which HNK mediates these anti-oxidant activities.
Our studies further identify a crucial role for Klf4 in mediating inflammation in BV-2 cells. We demonstrate that overexpression of Klf4 increases the production of iNOS and Cox-2 in these cells. In addition to these observations, we also report that Klf4 overexpression results in a decrease of anti-inflammatory activity of HNK. In K-10 + LPS conditions, HNK was not able to decrease the production of iNOS and Cox-2 significantly with respect to the K-10 + LPS alone condition. In addition, the levels of iNOS as well as Cox-2 remained significantly higher in the HNK treatment group with respect to the LPS alone condition. It is likely that overexpressed Klf4 remains bound to the respective promoter elements in spite of HNK treatment to these cells. There is, however, some reduction observed when HNK treatment was carried out in these Klf4 overexpressing cells which may be due to HNK's action on the basal level of cellular Klf4. These findings strongly suggest that Klf4 is one of the key players of inflammation and therefore one of the key targets of therapeutic intervention.
We hereby propose that Klf4 may be an important target for drugs that are known to traditionally block pNF-κB pathways. It would be interesting to understand the molecular mechanisms of HNK-mediated Klf4 expression. Furthermore, Klf4 as a potential target for other anti-inflammatory agents needs to be further evaluated.
cytokine bead array
central nervous system
Dulbecco's modified Eagle's medium
fluorescence-activated cell sorting
fetal bovine serum
inducible nitric oxide synthase
Krüppel-like factor 4
macrophage chemoattractant protein-1
mean fluorescent intensity
phospho-nuclear factor kappa B
reactive oxygen species
standard error of the mean
small interfering RNA
tumor necrosis factor alpha
This work was supported by a core grant from the Department of Biotechnology to the National Brain Research Centre and Life Science Research Board, Defence Research & Developmental Organization (DLS/81/48222/LSRB-213/EPB2010), Government of India and also supported by a National Bioscience Award for Career Development from the Department of Biotechnology to AB. DKK is the recipient of a Senior Research Fellowship from the Indian Council of Medical Research, Government of India. The authors thank Mr Manish Dogra for excellent technical assistance.