Cardoso - Jcr-2009

ARTICLE IN PRESS Journal of Controlled Release xxx (2009) xxx–xxx

Contents lists available at ScienceDirect

Journal of Controlled Release j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / j c o n r e l

Tf-lipoplex-mediated c-Jun silencing improves neuronal survival following excitotoxic damage in vivo A.L.C. Cardoso a,b, P. Costa a,b, L.P. de Almeida a,c, S. Simões a,c, N. Plesnila d, C. Culmsee e,1, E. Wagner e, M.C. Pedroso de Lima a,b,⁎ a

Center for Neuroscience and Cell Biology, University of Coimbra, 3004-517 Coimbra, Portugal Department of Biochemistry, Faculty of Science and Technology, University of Coimbra, Apartado 3126, 3001-401 Coimbra, Portugal Laboratory of Pharmaceutical Technology, Faculty of Pharmacy, University of Coimbra, 3000-295 Coimbra, Portugal d Department of Neurosurgery & Institute for Surgical Research, University of Munich Medical Center – Großhadern, Ludwig-Maximilians University, Germany e Department of Pharmacy, Ludwig-Maximilians University, Munich, Germany b c

a r t i c l e

i n f o

Article history: Received 24 July 2009 Accepted 3 November 2009 Available online xxxx Keywords: Tf-lipoplexes c-Jun Neuroprotection Excitotoxicity Kainate siRNA

a b s t r a c t Excitotoxicity is one of the main features responsible for neuronal cell death after acute brain injury and in several neurodegenerative disorders, for which only few therapeutic options are currently available. In this work, RNA interference was employed to identify and validate a potential target for successful treatment of excitotoxic brain injury, the transcription factor c-Jun. The nuclear translocation of c-Jun and its upregulation are early events following glutamate-induced excitotoxic damage in primary neuronal cultures. We present evidence for the efficient knockdown of this transcription factor using a non-viral vector consisting of cationic liposomes associated to transferrin (Tf-lipoplexes). Tf-lipoplexes were able to deliver anti-c-Jun siRNAs to neuronal cells in culture, resulting in efficient silencing of c-Jun mRNA and protein and in a significant decrease of cell death following glutamate-induced damage or oxygen–glucose deprivation. This formulation also leads to a significant c-Jun knockdown in the mouse hippocampus in vivo, resulting in the attenuation of both neuronal death and inflammation following kainic acid-mediated lesion of this region. Furthermore, a strong reduction of seizure activity and cytokine production was observed in animals treated with anti-c-Jun siRNAs. These findings demonstrate the efficient delivery of therapeutic siRNAs to the brain by Tf-lipoplexes and validate c-Jun as a promising therapeutic target in neurodegenerative disorders involving excitotoxic lesions. © 2009 Elsevier B.V. All rights reserved.

1. Introduction Hypoxic/ischemic events contribute to neuronal degeneration in many acute central nervous system disorders, including stroke, traumatic brain injury and epilepsy and may also play a role in chronic diseases, such as amyotrophic lateral sclerosis and Alzheimer's disease [1]. The impact of all these disorders in developed countries is considerable, with stroke, for example, being one of the leading causes of death and disability worldwide. The fundamental process responsible for triggering neuronal cell death after an ischemic event is

⁎ Corresponding author. Department of Biochemistry, Faculty of Sciences and Technology, University of Coimbra, Apartado 3126, 3001-401 Coimbra, Portugal. Tel.: +351 239 820 190; fax: +351 239 853 607. E-mail addresses: [email protected] (C. Culmsee), [email protected] (M.C. Pedroso de Lima). 1 Current address: Clinical Pharmacy – Pharmacology and Toxicology, Faculty of Pharmacy, Philipps-University of Marburg, Karl v. Frisch Strasse 1, 35043 Marburg, Germany. Tel.: + 49 6421 25780; fax: +49 6421 25720.

known as excitotoxicity [2] and refers to the excessive stimulation of excitatory amino acid receptors, with consequent increase in intracellular Ca2+ levels and activation of pro-death signalling pathways. Calcium-dependent activation of proteases, lipases and nucleases leads to cytoskeleton breakdown, oxidative stress and nuclear DNA degradation, resulting in impaired neuronal function and, ultimately, neuronal death [3]. Cell death mechanisms involve a series of proteins, ranging from ion-channels and membrane receptors to mitochondrial proteins and transcription factors. Revealing the role of these proteins under physiological conditions and after excitotoxic damage is crucial for the design of new and improved strategies aiming at neuroprotection and neuronal recovery after brain injury. Stress-activated pathways, such as the mitogen-activated protein kinase (MAPK) cascade, were found to be strongly activated following cerebral ischemia [1,4], brain trauma and seizure [5] and, depending on the duration of the hypoxic/ischemic insult and extent of energy depletion, the activation of these processes may be neuroprotective or detrimental to the cells. For example, cerebral ischemia enhanced the activity of several members of the MAPK family, extracellular signalregulated kinases (ERK1 and ERK2), p38 and c-Jun N-terminal kinases

0168-3659/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jconrel.2009.11.004

Please cite this article as: A.L.C. Cardoso, et al., Tf-lipoplex-mediated c-Jun silencing improves neuronal survival following excitotoxic damage in vivo, J. Control. Release (2009), doi:10.1016/j.jconrel.2009.11.004

GENE DELIVERY

COREL-05279; No of Pages 12

GENE DELIVERY

ARTICLE IN PRESS 2

A.L.C. Cardoso et al. / Journal of Controlled Release xxx (2009) xxx–xxx

(JNK1, JNK2 and JNK3) were shown to be active after ischemia [6]. In contrast to ERK1 and ERK2, which mediate neuroprotection when activated by neurotrophins [7], p38 and JNKs are thought to be mostly involved in inflammatory cytokine production and apoptosis [8,9]. Several studies have shown that JNK inhibition or knockout in the brain not only improves the functional outcome following kainic acidinduced excitotoxic lesion [10,11], but also reduces lesion size after cerebral ischemia [12]. The downstream events of JNK activation involve c-Jun phosphorylation and formation of the activated protein1 (AP-1) transcription factor, which leads to the expression of immediate early genes (IEGs) and production of several proteins important for processes of neuronal injury and inflammation, such as FasL, Bim, COX-2 and TNF-α [13,14]. Most of these IEGs were found to be upregulated following hypoxic/ischemic injury, which suggests a major role for c-Jun and AP-1 in the activation of some of the most important executing pathways of delayed neuronal cell death. The aim of the current study was to follow up our previous findings on the neuroprotective role of c-Jun silencing [15] in vitro, and investigate the potential of siRNA-mediated downregulation of c-Jun as a therapeutic approach in an excitotoxic lesion model in vivo, using a lipid-based siRNA delivery system [16].

for in vivo application, the dried lipid film was hydrated in 1.6 ml of 5% HBG buffer (Hepes–glucose buffer: 5% glucose, 20 mM Hepes, pH 7.4) and sonicated for 5 min. The resulting liposomes were then extruded 21 times through two stacked polycarbonate membranes (50 nm pore diameter) and diluted in HBG buffer to a final DOTAP concentration of 22.5 mM. The liposomes were stored at 4 °C until use. For the in vitro studies, Tf-lipoplexes were prepared by preincubating a given volume of the liposome suspension with ironsaturated human transferrin (32 µg/µg of siRNA) for 15 min, followed by addition of the necessary volume of siRNA stock solution to achieve a final siRNA concentration of 50 or 100 nM in each well and a 2/1 lipid/siRNA charge ratio. The mixture was further incubated for 30 min at room temperature before delivery to the cortical neurons in culture. Alternatively, for in vivo administration, Tf-lipoplexes prepared at a 6/1 lipid/siRNA (+/−) charge ratio were obtained by mixing 0.8 µl/animal of the liposome stock solution (22.5 mM DOTAP) with 0.5 µl/animal of human Tf solution (192 mg/ml in HBG), followed by 15 min incubation prior to the addition of siRNAs (2 µg/ animal). The resulting mixture was further incubated for 30 min. All formulations were used immediately after being prepared. 2.4. Primary neuronal cultures

2. Materials and methods 2.1. Materials The cationic lipid 1,2 dioleoyl-3(trimethylammonium)propane (DOTAP) and cholesterol (Chol) were purchased from Avanti Polar Lipids (Alabaster, AL, USA). Iron-saturated human transferrin (Tf) was obtained from Sigma (Sigma, St.Louis, MO, USA). The anti-c-Jun siRNA (5′-AGTCATGAACCACGTTAAC-3′) was obtained from Ambion (Ambion, Austin, Texas, USA). The control non-silencing siRNA as well as all the QRT-PCR reagents was obtained from Qiagen (Qiagen, Hilden, Germany). The c-Jun antibody was purchased from Cell Signalling (Cell Signalling, Danvers, USA), the α-tubulin antibody was obtained from Sigma (Sigma, Saint Louis, USA), the CD11b antibody was purchased from Serotec (Serotec, Oxford, United Kingdom) and the GFAP and NeuN antibodies were purchased from Millipore (Millipore, Billerica, USA). All other chemicals were obtained from Sigma unless stated otherwise. 2.2. Animals All efforts were made to minimize the number of animals and suffering according to the guidelines of the German animal protection law and derived guidelines on the ethical use of animals and the relevant international laws and policies (Directive 86/609/EEC and Guide for the Care and use of Laboratory Animals, US National Research Council, 1996). C57BL/6 mice were obtained from Charles River, Sulzfeld, Germany. All animals were kept under controlled light and environmental conditions (12 h dark/light cycle, 23 ± 1 °C, 55 ± 5% relative humidity), having free access to food and water. 2.3. Liposome and lipoplex preparation Cationic liposomes composed of DOTAP:cholesterol (1:1 molar ratio) were prepared as previously described by Campbell [17] for in vitro application. Briefly, a mixture of 1 ml of DOTAP and 1.5 ml of cholesterol in chloroform (from stock solutions of 25 mg/ml DOTAP and 37.8 mg/ml cholesterol), was dried under nitrogen in order to obtain a thin lipid film. The film was dissolved in 100 µl of ultrapure ethanol and the resulting ethanol solution was injected into 900 µl of HGB buffer (Hepes–glucose buffer: 5% glucose, 20 mM Hepes, pH 7.4) maintained under continuous vortex, employing a 250 µl Hamilton syringe. The resulting MLV (multilamellar vesicles) were sonicated briefly to obtain SUV (small unilamellar vesicles) and diluted in HBS to a final lipid concentration of 1.43 mM DOTAP (1 mg/ml). Alternatively,

Primary mouse embryonic cortical neurons were obtained from C57/BL6 mice, at day 16 of gestation, as described previously [18]. After dissociation and centrifugation of the dissected cortices, the tissue was ressuspended in Neurobasal medium (Invitrogen, San Diego, CA, USA), enriched with 2% (v/v) B27 supplement (Invitrogen), 2 mM glutamine and 100 U/ml penicillin/streptomycin (Invitrogen). For survival experiments, QRT-PCR and Western blot analysis, cells were plated at a density of 0.5 × 106 cells/well onto 12-well plates previously coated with poly-L-lysine. For fluorescence microscopy studies, cells were plated at a density of 0.12× 106 cells/well onto 12-well plates containing glass coverslips previously coated with poly-L-lysine. Characterization of the embryonic neuronal cultures confirmed the presence of 95% neurons in these cultures, as determined by GFAP and NeuN-immunostaining. Primary cultures were kept at 37 °C in a humidified atmosphere containing 5% CO2. All experimental treatments were performed in 13–14 day old cultures. 2.5. In vitro Tf-lipoplex-mediated siRNA delivery and induction of neuronal cell death After 13 days in culture, 50 µl of Tf-lipoplexes containing anti-c-Jun or non-silencing (Mut) siRNAs was added to the cells to a final siRNA concentration of 50 nM per well. After a 4 h incubation period (in 5% CO2, at 37 °C), the Neurobasal medium was replaced with a fresh medium and the cells were further incubated for different periods of time (24 h for QRT-PCR analysis and 48 h for Western blot or cell viability analysis). In order to induce oxygen–glucose deprivation (OGD), glucose-free EBSS medium (6.8 g/l NaCl, 0.4 g/l KCl, 0.264 g/l CaCl2∙2H2O, 0.2 g/l MgCl2∙2H2O, 2.2 g/l NaHCO3, 0.14 g/l NaH2PO4∙2H2O, pH 7.2) supplemented with gentamicin (5 mg/l) was purged with 95% N2 /5% CO2 for 30 min, resulting in an oxygen content of 2–3%. After 14 days in culture, neurons were washed 3 times with this medium and incubated for 4 h in an oxygen-free 95% N2/5% CO2 atmosphere (OGD). Control cultures were incubated in EBSS with 10 mM glucose. In order to induce glutamate excitotoxicity, neurons (14 days in culture) were exposed to 125 µM glutamate in EBSS with 10 mM glucose, for 20 min. Following the indicated incubation periods (4 h for OGD and 20 min for glutamate excitotoxicity), the medium was replaced by standard Neurobasal medium. Eighteen hours after the onset of OGD or glutamate challenge, cells were harvested for cell viability analysis or protein and mRNA extraction. In parallel experiments, neurons were washed two times in phosphate-buffered saline (PBS) and fixed in PBS containing 4% paraformaldehyde for immunocytochemistry studies.

Please cite this article as: A.L.C. Cardoso, et al., Tf-lipoplex-mediated c-Jun silencing improves neuronal survival following excitotoxic damage in vivo, J. Control. Release (2009), doi:10.1016/j.jconrel.2009.11.004

A.L.C. Cardoso et al. / Journal of Controlled Release xxx (2009) xxx–xxx

3

2.6. Evaluation of cell viability

2.9. Histological evaluation and immunostaining

Cell viability of primary neuronal cultures was determined, following Tf-lipoplex delivery, by 3-(4,5 dimethylthiazol-2-y1)-2,5diphenyltetrazolium bromide (MTT) reduction. Forty-eight hours after transfection and 18 h after glutamate or OGD exposure, cells were washed with fresh culture medium and incubated with MTT (0.25 mg/ml) for 2 h at 37 °C. The reaction was terminated by adding dimethylsulfoxide solution and the absorbance was determined at 590 nm and 630 nm in an ELISA microplate reader (Spectra FluorPlus; Tecan, Durham, NC, USA). Cell viability was calculated as percentage of control cells (non-treated cells) using the formula: (A590-A630) of treated cells × 100/(A590-A630) of control cells.

For histological evaluation using cresyl violet staining, brain sections containing the hippocampus region were mounted and dried in gelatine-coated glass slides, and each slide was stained for 5 min in 0.5% cresyl violet solution in acetate buffer, rinsed twice in water, briefly dehydrated in ethanol, cleared in xylene solution and mounted with Entellan (Merck, Darmstadt, Germany). The striatal slices were examined under a Zeiss Axiovert microscope (Zeiss, Thornwood, New York, USA) equipped with 5× and 20× objectives. In order to quantify neuronal loss, neuron counts were made in the CA3 region of every six section (180 µm separation distance), using an unbiased counting frame. Only neurons with a visible nucleus and in which the entire outline of the cell was apparent were considered intact. Intensively stained, condensed and fragmented nuclei were considered damaged and counted as pycnotic nuclei. For each section, the number of intact and pycnotic neurons was counted in the CA3 region of both hemispheres. Cell counts were averaged and mean numbers were used for statistical analysis. Results were expressed as the percentage of intact neurons of the total CA3 neurons per hemisphere. Immunocytochemistry and immunohistochemistry were performed in cultured cells or in 30 μm brain slices, respectively, according to established protocols. Briefly, the cells or tissue were permeabilized for 30 min in PBS/0.5% Triton X-100, and non-specific binding was blocked with PBS/4% goat serum (Invitrogen, Karlsruhe, Germany) for an additional 30 min period. The cells or tissue were incubated overnight at 4 °C with the primary antibodies anti-c-Jun (dilution 1:100), anti-NeuN (specific neuronal marker, dilution 1:500), anti-GFAP (specific marker of astrocytes, dilution 1:1000), anti-CD11b (specific marker of microglia, dilution 1:500) or anti-αtubulin (dilution 1:1000), diluted in PBS containing 0.25% goat serum, followed by two washing steps in PBS. Alexa Fluor® 488 – or Alexa Fluor® 594-conjugated secondary antibodies (Molecular Probes, Leiden, The Netherlands) in PBS/0.25% goat serum were applied at a 1:500 dilution for 2 h at room temperature. After further rinsing twice in PBS, the coverslips were mounted in glass slides using the Prolong® Anti-fade kit (Molecular Probes) and the brain slices were mounted on gelatine-coated glass slides using the FluorSave™ reagent (Calbiochem, Darmstadt, Germany). All coverslips or glass slides sections were observed under a Zeiss Axiovert epifluorescence microscope, equipped with the 20× or 40× objectives and the rhodamine, FITC and DAPI filters. In order to evaluate c-Jun translocation to the nucleus following KA-induced excitotoxicity, cell counts were performed following exposure of neuronal primary cultures to KA. The number of blue nuclei and red nuclei was counted in at least six separate coverslips per condition and the results were expressed as the percentage of red nuclei (cells presenting nuclear c-Jun) of the total cell number per field. Loss of the neuronal marker NeuN was also evaluated following KAinduced lesion in vivo, by immunohistochemistry. Neuron counts were made in the CA3 region of at least four sections per animal. For each section, the number of red cells (cells positive for NeuN) was counted in the CA3 region of both hemispheres and results were expressed as a percentage of NeuN positive neurons in control animals (sham-operated animals).

2.7. Stereotactic injection of Tf-lipoplexes and kainic acid For hippocampal injections of Tf-lipoplexes, C57/BL6 mice were anaesthetized with 10 µl/g of avertin (1.3% tribromoethanol and 0.8% amylalcohol in MiliQ water) and placed in a stereotactic apparatus. A midline incision was made, the soft tissues were reflected and a burr hole was made in the skull with the aid of a surgery micro-drill, at a point −2.00 mm (posterior) from bregma and 2.25 mm lateral from the midline, according to Paxinos and Franklin [19]. A volume of 2 µl of Tflipoplexes in HGB, containing 2 μg of anti-c-Jun or non-silencing (Mut) siRNAs was injected at a rate of 0.2 µl/30 s in the right hemisphere (ipsilateral hemisphere) of each animal, via a stainless steel needle connected to a Hamilton syringe (Hamilton Bonaduz, Bonaduz, GR, Switzerland), 2.00 mm ventral from dura. For kainic acid (KA) administration, each mouse received a single injection of 3 µl PBS containing 1 nmol of KA in the lateral ventricle of the left brain hemisphere, at a point −1.00 mm posterior from bregma, 1.75 mm lateral from the midline and 1.75 mm ventral from dura. Control animals received a single injection of PBS (sham operated). Five minutes after the injections were completed, the needle was withdrawn slowly and the skin was sutured. No symptoms of toxicity or loss of basic activity were observed in any of the mice following Tf-lipoplex injection or vehicle application. All animals were sacrificed 1, 3 or 5 days following Tf-lipoplex injection and 1 or 3 days following KA administration. For histological evaluation of cell viability and immunohistochemistry analysis, the animals were transcardially perfused with 20 ml of an ice-cold 0.9% NaCl solution, followed by further perfusion with 20 ml of ice-cold 4% paraformaldehyde in 0.9% NaCl solution. The brains were removed and postfixed (12 h) in the same fixative solution, followed by 2–3 days in a cryoprotective solution containing 25% sucrose. After this period, the brains were rapidly frozen in dry ice, dipped in OCT embedding medium (Sakura Finetek, Mijdrecht, The Netherlands), and 30 µm sections were cut at −20 °C in a cryostat (Leica CM 3050 S, Leica, Wetzlar, Germany) 1000 µm anterior and 1000 µm posterior from the injection site and placed in PBS. For protein and mRNA extraction, the brains were removed and placed on an acrylic matrix. A 2 mm coronal section containing the injection site was cut with a stainless steel razor and the hippocampal region from both hemispheres of each animal was dissected and placed in the appropriate ice-cold lysis buffer.

2.8. Monitoring of seizure activity

2.10. Extraction of RNA and cDNA synthesis

Following KA administration, the mice were monitored continuously for 4 h for the onset and extent of seizure activity. Seizures were rated according to a previously defined scale [20]; stage 1: immobility, stage 2: forelimb and/or tail extension, stage 3: repetitive movements, head bobbing, stage 4: rearing and falling, stage 5: continuous rearing and falling, blackouts and stage 6: severe tonic-clonic seizures and death. In order to be included in this study, mice needed to demonstrate at least stage 3 seizures.

Total RNA was extracted from 1 × 106 neuronal cells or from hippocampal samples using the RNeasy Mini Kit (Qiagen, Hilden, Germany), according to the manufacturer's recommendations for cultured cells or brain tissue, respectively. Briefly, after cell lysis, the total RNA was adsorbed to a silica matrix, washed with the recommended buffers and eluted with 40 µl of RNase-free water by centrifugation. After RNA quantification, cDNA conversion was performed using the Superscript III First Strand Synthesis Kit (Invitrogen,

Please cite this article as: A.L.C. Cardoso, et al., Tf-lipoplex-mediated c-Jun silencing improves neuronal survival following excitotoxic damage in vivo, J. Control. Release (2009), doi:10.1016/j.jconrel.2009.11.004

GENE DELIVERY

ARTICLE IN PRESS

GENE DELIVERY

ARTICLE IN PRESS 4

A.L.C. Cardoso et al. / Journal of Controlled Release xxx (2009) xxx–xxx

Karlsruhe, Germany), according to the manufacturer's instructions. For each sample, cDNA was produced from 0.5 µg of total RNA in a iQ 5 thermocycler (Bio-Rad, Munich, Germany), by applying the following protocol: 10 min at 25 °C, 30 min at 55 °C and 5 min at 85 °C. After cDNA synthesis, a 30 min incubation period with RNase H at 37 °C was performed in order to remove any remaining RNA contamination. Finally, the cDNA was diluted 1:3 with RNase-free water prior to quantification by QRT-PCR.

2.11. Quantitative real time polymerase chain reaction (QRT-PCR) Quantitative PCR was performed as described previously [16] in an iQ5 thermocycler (Bio-Rad) using 96-well microtitre plates and the iQ SYBR Green Supermix Kit (Bio-Rad). The primers for the target genes (c-Jun, IL-1β, IL-6 and TNF-α) and the two tested housekeeping genes (GAPDH and HPRT) were pre-designed by Qiagen (QuantiTect Primer, Qiagen). A master mix was prepared for each primer set, containing a fixed volume of SYBR Green Supermix and the appropriate amount of each primer to yield a final concentration of 150 nM. For each reaction, 20 µl of master mix were added to 5 µl of template cDNA. All reactions were performed in duplicate (two cDNA reactions per RNA sample) at a final volume of 25 µl per well, using the iQ5 Optical System Software (Bio-Rad). The reaction conditions consisted of enzyme activation and well-factor determination at 95 °C for 1 min and 30 s followed by 40 cycles at 95 °C for 10 s (denaturation), 30 s at 55 °C (annealing) and 30 s at 72 °C (elongation). The melting curve protocol started immediately after amplification and consisted of 1 min heating at 55 °C followed by eighty 10 s steps, with 0.5 °C increases in temperature at each step. Threshold values for threshold cycle determination (Ct) were generated automatically by the iQ5 Optical System Software. The percentage of c-Jun knockdown or interleukin decrease was determined following the guidelines for relative mRNA quantification in the presence of target and reference genes with different amplification efficiencies. The amplification efficiency for each target or housekeeping gene was determined according to the formula: E = 10(− 1/S) − 1, where S is the slope of the standard curve obtained for each gene.

2.13. Statistical analysis All data are presented as mean± standard deviation (SD). In vitro data result from three independent experiments, each performed at least in triplicate. One way ANOVA analysis of variance combined with Tukey posthoc test was used for multiple comparisons in cell culture experiments. In vivo data were analysed by the Kruskal–Wallis one way analysis of variance, followed by the Dunnett's posthoc test for multiple comparisons between groups (n = 6 for each group). Statistical differences are presented at probability levels of p b 0.05, p b 0.01 and p b 0.001. Calculations were performed with standard statistical software (GraphPad Prism 4). 3. Results 3.1. c-Jun expression and nuclear translocation following glutamate toxicity in vitro c-Jun activation and mRNA increase in neuronal primary cultures were evaluated at different time points (15 min, 30 min, 1 h, 3 h and 6 h), by immunocytochemistry and QRT-PCR, following excitotoxic damage mediated by acute exposure to glutamate. c-Jun mRNA levels increased significantly within the first 30 min after glutamate exposure, reaching a peak at 30 min–1 h after the insult and then decreasing to basal levels (Fig. 1A). In the same time window, a similar timedependent increase in the number of cells showing c-Jun nuclear translocation was observed (Fig. 1B and C). At 30 min following glutamate exposure, 70% of neurons showed c-Jun nuclear labelling (Fig. 1B and C, panels c, f and i) which strongly correlates with the increase in c-Jun mRNA observed at this time point. No significant modifications in nuclear morphology and cytoskeleton were observed at early time points (Fig. 1C, panels a, b, c, j and k), but a degeneration of neuronal terminals and tubulin filaments was evident, starting at 6 h after glutamate insult (Fig. 1C, panel l). These results suggest that c-Jun nuclear translocation is an early event of neuronal degeneration following excitotoxic damage and precedes nuclear shrinkage and cytoskeleton fragmentation.

2.12. Western blot analysis

3.2. Neuroprotection following in vitro Tf-lipoplex-mediated c-Jun silencing

Protein extracts were obtained from neuronal primary cultures or from hippocampal tissue samples homogeneized at 4 °C in lysis buffer (50 mM NaCl, 50 mM EDTA, 1% Triton X-100) containing a protease inhibitor cocktail (Sigma), 10 μg/ml DTT and 1 mM PMSF. Protein content was determined using the Bio-Rad Dc protein assay (Bio-Rad) and 20 µg of total protein was resuspended in a loading buffer (20% glycerol, 10% SDS, 0.1% bromophenol blue), incubated for 2 min at 95 °C and loaded onto a 10% polyacrylamide gel. After electrophoresis the proteins were blotted onto a PVDF membrane according to standard protocols. After blocking in 5% non-fat milk, the membrane was incubated with the appropriate primary antibody (anti-c-Jun 1:500 and anti-GFAP 1:1000) overnight at 4 °C, and with the appropriate secondary antibody (1:20000) (Amersham, Uppsala, Sweden) for 2 h at room temperature. Equal protein loading was shown by reprobing the membrane with an anti-α-tubulin antibody (1:10000) (Sigma) and with the same secondary antibody. After this incubation period, the blots were washed several times with saline buffer (TBS/T — 25 mM Tris–HCl, 150 mM NaCl, 0.1% Tween and 5 mg/ml non-fat powder milk) and incubated with ECF (alkaline phosphatase substrate; 20 µl of ECF/cm2 of membrane) for 5 min at room temperature and then submitted to fluorescence detection at 570 nm using a Storm-860 (Molecular Dynamics, CA, USA). For each membrane, the analysis of band intensity was performed using the Quantity One software (Bio-Rad).

In a previous study [15], c-Jun silencing mediated by Tf-lipoplexes containing anti-c-Jun siRNAs was found to result in a neuroprotective effect in HT-22 cells, a neuronal cell line sensitive to glutamate damage. Aiming at examining the possible neuroprotective role of c-Jun knockdown in primary neurons, cortical cultures (day 13 after isolation) were transfected with Tf-lipoplexes containing anti-c-Jun or Mut (nonsilencing) siRNAs and exposed to glutamate or OGD 24 h later. Results in Fig. 2 illustrate the c-Jun mRNA and protein knockdown 24 h following glutamate or OGD insults, as well as the neuronal cell viability at this time point. A significant decrease in c-Jun mRNA (70% reduction) (Fig. 2A) and protein levels (50% reduction) (Fig. 2B and C) was observed 48 h after transfection. Gene silencing was found to be specific, since no significant alteration of c-Jun levels was observed after delivery of Mut siRNAs. These results demonstrate that efficient c-Jun silencing by Tf-lipoplexes persists also in neurons exposed to the excitotoxic damage. In accordance with our previous results in a neuronal cell line, c-Jun knockdown significantly increased neuronal viability after glutamate (Fig. 2D) or OGD (Fig. 2E) exposure, by 25% and 30% respectively, when compared to untreated neuronal cultures exposed to these two insults. No positive effect on cell viability was observed following delivery of Mut siRNAs, which indicates that the observed neuroprotective effect is directly related to c-Jun silencing and is not an indirect consequence of the transfection process.

Please cite this article as: A.L.C. Cardoso, et al., Tf-lipoplex-mediated c-Jun silencing improves neuronal survival following excitotoxic damage in vivo, J. Control. Release (2009), doi:10.1016/j.jconrel.2009.11.004

A.L.C. Cardoso et al. / Journal of Controlled Release xxx (2009) xxx–xxx

3.3. Efficient and non-toxic Tf-lipoplex-mediated c-Jun knockdown in vivo We have previously demonstrated that association of Tf to lipoplexes promotes siRNA delivery and efficient knockdown of reporter

5

genes in vivo [16]. Therefore, in this work we evaluated whether this strategy would also result in efficient c-Jun silencing in the mouse brain. For this purpose, C57/BL6 mice were injected with Tf-lipoplexes prepared at 6/1 lipid/siRNA charge ratio and containing 2 µg of anti-cJun siRNAs or Mut siRNAs. The injections were performed near the CA3 region of the hippocampus of the right hemisphere (ipsilateral) of each animal and c-Jun knockdown was evaluated 1, 3 or 5 days after Tf-lipoplex delivery, by QRT-PCR, immunohistochemistry and Western blot. Fig. 3 provides evidence that stereotactic injection of Tf-lipoplexes leads to efficient c-Jun silencing in vivo, as observed by the significant decrease of c-Jun mRNA (Fig. 3A) and protein levels (Fig. 3C and D) following siRNA delivery. Around 60% decrease in c-Jun mRNA was observed in the ipsilateral hemisphere of animals treated with anti-cJun siRNAs, when compared to animals treated with Mut siRNAs (Fig. 3A), 1 day after Tf-lipoplex injection. As expected from a non-viral delivery system, c-Jun mRNA silencing mediated by siRNAs was found to be transient and c-Jun mRNA returned to basal levels 5 days after Tflipoplex injection (data not shown). Although no differences were found in c-Jun protein levels at this same time point, a significant reduction (45%) was observed starting at day 3 (Fig. 3B, C and D) and silencing of the c-Jun gene lasted at least 5 days (Fig. 3A and D). Immunohistochemistry images indicate that protein knockdown occurs mainly in CA3 region (Fig. 3B, panel b), where loss of nuclear and cytoplasmic c-Jun labelling can be observed in the pyramidal neurons. In order to evaluate Tf-lipoplex biocompatibility, counterstaining of brain slices with cresyl violet was performed in animals injected with Tf-lipoplexes containing Mut or anti-c-Jun siRNAs (Fig. 4, panels a–d). In parallel experiments, immunohistochemical labelling of the specific cell markers GFAP (Fig. 4, panels e and f) and CD11b (Fig. 4, panels g and h) was performed in order to investigate a possible inflammatory reaction, which could lead to gliosis and microglia activation. No relevant signs of toxicity were found throughout the ipsilateral hippocampus where, similarly to observations in the contralateral hemisphere, neurons presented a light violet colour and large cell body characteristic of healthy cells (Fig. 4, panel d). The presence of vacuolization or large number of apoptotic cells was not detected in any of the injected animals and only a small number of pycnotic cells were observed in the tissue surrounding the injection site. Moreover, no differences in the number and morphology of astrocytes and microglia cells were observed between the contralateral and ipsilateral hemispheres (Fig. 4, panels e–h). Results were similar in animals receiving anti-c-Jun or Mut siRNAs. These results indicate that Tf-lipoplexes are well tolerated and do not seem to induce a significant inflammatory reaction following local application. 3.4. Contribution of c-Jun silencing to neuroprotection following in vivo KA-induced excitotoxic lesion Neurotoxicity mediated by overstimulation of glutamate receptors and massive calcium influx is known to play an important role in

Fig. 1. c-Jun mRNA upregulation and nuclear translocation following glutamate-mediated excitotoxicity. C-Jun mRNA quantification by QRT-PCR or immunocytochemistry experiments took place 0 min, 15 min, 30 min, 1 h , 3 h or 6 h after the onset of lesion (20 min with 125 µM glutamate). (A) c-Jun mRNA levels are expressed as fold increase above c-Jun mRNA levels in control cells. In order to study c-Jun nuclear translocation, neurons were labelled with Hoechst 33342 (blue), anti-c-Jun antibody (red) and antitubulin antibody (green). (B) The number of c-Jun positive nuclei was determined for each time point and is expressed as the percentage of total number of nuclei. (C) Fluorescence microscopy images were acquired at 200× magnification. Representative images for each time point are presented separately for the red and blue channels and as a merged image. Results in (A) and (B) are presented as mean values ± SD and are representative of three independent experiments, each performed in triplicate. ⁎p b 0.5, ⁎⁎p b 0.01 and ⁎⁎⁎p b 0.001 compared to control cells.

Please cite this article as: A.L.C. Cardoso, et al., Tf-lipoplex-mediated c-Jun silencing improves neuronal survival following excitotoxic damage in vivo, J. Control. Release (2009), doi:10.1016/j.jconrel.2009.11.004

GENE DELIVERY

ARTICLE IN PRESS

GENE DELIVERY

ARTICLE IN PRESS 6

A.L.C. Cardoso et al. / Journal of Controlled Release xxx (2009) xxx–xxx

several pathologies such as stroke, epilepsy, trauma and some neurodegenerative diseases. In order to evaluate the potential therapeutic effect of c-Jun silencing in an excitotoxic context, a selective lesion was induced in the mouse brain by administering kainic acid, an agonist of both kainate and AMPA subtypes of glutamate receptors. Intracerebroventricular injection of 0.1 µg of this excitotoxic substance (Fig. 5A) in the left brain hemisphere was applied to establish a well defined lesion in the hippocampus according to a previously developed epilepsy model [21,22]. Tf-lipoplexes containing anti-c-Jun or Mut siRNAs were injected near the CA3 region of the right hemisphere (Fig. 5A) 3 days before or immediately after KA injection, aiming at evaluating the neuroprotective role of c-Jun silencing in both a pre- and post-lesion situation. The loss of neuronal integrity was investigated 3 days after KA injection, by both cresyl violet staining (Fig. 5B and panels a–j) and NeuN immunohistochemistry (Fig. 5B and panels k–t). Fig. 5 shows clear signs of neuronal loss after KA injection, in both hemispheres of Mut siRNA-treated animals and in contralateral hemispheres of animals receiving c-Jun siRNAs. Between 60–70% of neuronal death was observed, following cresyl violet counterstaining (Fig. 5C), as well as significant loss (75%) of the neuronal marker NeuN (Fig. 5D), indicating the presence of neurodegeneration and apoptosis and providing evidence of KA-induced lesion in these brain regions. On the contrary, no lesion could be observed in the hemispheres (right) of animals injected with Tf-lipoplexes containing anti-c-Jun siRNAs, both pre- and post-lesion (Fig. 5, panels f, j, p and t), where few pycnotic cells were detected and NeuN loss was not significant (10–20%), similarly to what was observed in all sham-operated animals (Fig. 5, panels a, b, k and l). These results indicate a clear neuroprotective effect of c-Jun silencing in this excitotoxic model of brain injury. Concerning seizure activity, all animals were observed continuously for 4 h following KA injection for the onset and extent of seizures. Seizures were rated according to a previously defined scale [20]. Almost all control animals (injected with KA only) and animals treated with Mut siRNAs, pre- (Mut + KA) or post-lesion (Mut/KA) achieved the expected stage 5 in the Racine's scale (80%), presenting spontaneous blackouts and continuous rearing and falling (Fig. 6). Moreover, a small percentage of death animals (10–30%) were also observed in these groups (Fig. 6). In contrast, no deaths were registered in the animals treated with anti-c-Jun siRNAs pre- (c-Jun + KA) or post-lesion (c-Jun/KA), and in the group that received pre-treatment most animals only achieved stages 3 and 4 of status epilepticus, with only 15% of the animals reaching stage 5. Overall, these data imply that c-Jun silencing, when performed previously to the onset of seizure activity, can effectively moderate the amount and severity of seizures. 3.5. Reduction of KA-induced inflammation following Tf-lipoplex-mediated c-Jun silencing KA administration is known to induce dramatic changes in the number and morphology of both astrocytes and microglia, leading to the production and release of inflammatory cytokines. In order to evaluate the effect of c-Jun knockdown in the modulation of the inflammatory reaction induced by KA injection, the protein levels of GFAP, a specific marker of astrocytes, were analysed by Western blot (Fig. 7A and B) and immunohistochemistry (Fig. 7C, panels k–t),

Fig. 2. Recovery of neuronal viability following Tf-lipoplex-mediated c-Jun silencing in vitro. Twenty-four hours following transfection with Tf-lipoplexes containing c-Jun siRNA or Mut siRNA, neurons were incubated with glutamate (125 µM–20 min) or exposed to oxygen–glucose deprivation (OGD) for 4 h. (A) Quantification of c-Jun mRNA by QRT-PCR, (B) and (C) Western blot analysis of c-Jun protein levels or (D) and (E) cell viability analysis using the MTT assay were performed 18 h after lesion. Results in (A), (B), (D) and (E) are expressed as a percentage of control and are presented as mean values ± SD. All results are representative of three independent experiments, each performed in triplicate. ⁎⁎p b 0.01 and ⁎⁎⁎p b 0.001 compared to cells exposed to glutamate or OGD.

Please cite this article as: A.L.C. Cardoso, et al., Tf-lipoplex-mediated c-Jun silencing improves neuronal survival following excitotoxic damage in vivo, J. Control. Release (2009), doi:10.1016/j.jconrel.2009.11.004

A.L.C. Cardoso et al. / Journal of Controlled Release xxx (2009) xxx–xxx

7

Fig. 4. Tf-lipoplex biocompatibility in vivo, following injection in the mouse hippocampus. Mice (n = 6) were injected, in the CA3 region of the right hippocampus (ipsilateral) with Tf-lipoplexes (Mut siRNAs); no injection was performed in the left hippocampus (contralateral). Animals were sacrificed 3 days postinjection. (a–d) Biocompatibility of Tf-lipoplexes was evaluated in brain sections surrounding the injection site by cresyl violet staining. Representative light microscopy images are presented at (a,b) 50× and (c,d) 200× magnifications. A possible inflammatory response to the delivery of Tflipoplexes was evaluated by immunohistochemistry. Sections were labelled with (e,f) an anti-GFAP antibody (green) which specifically labels astrocytes and with (g,h) an anti-CD11b antibody (red) which labels microglia cells. Representative fluorescence microscopy images are presented at 200× magnification. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 3. Tf-lipoplex-mediated c-Jun silencing in the mouse hippocampus. Mice were injected in the CA3 region of the right hippocampus (ipsilateral), with Tf-lipoplexes (2 µg c-Jun siRNAs or Mut siRNAs); no injection was performed in the left hippocampus (contralateral). Animals were sacrificed 1, 3 or 5 days postinjection (D1, D3 and D5). (A) c-Jun mRNA levels were evaluated by QRT-PCR. Results are presented as a percentage of mRNA levels in sham-operated animals. (B) c-Jun protein knockdown was investigated by immunohistochemistry (green — anti-c-Jun antibody). (a–d) Representative fluorescence microscopy images of the contralateral hemispheres and ipsilateral hemispheres injected with anti-c-Jun or Mut siRNAs (day 3) are presented at 200× magnification. Protein knockdown was assessed by Western blot. (C) Representative gel showing c-Jun protein levels at day 3 following Tf-lipoplex delivery. Two bands, corresponding to the ipsilareral (ips) and contralateral (ct) hemispheres are presented for each animal (#). (D) Quantification of c-Jun protein knockdown, corrected for individual α-tubulin levels. Results are expressed as a percentage of c-Jun levels in the contralateral hemispheres. Results in (A) and (D) are presented as mean values± SD (n = 6). ⁎⁎p b 0.01 and ⁎p b 0.05 compared to animals injected with Mut siRNAs. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3 days after KA administration. In parallel immunohistochemistry experiments, microglia activation was followed using the specific cell marker CD11b (Fig. 7C, panels a–j). A significant increase in GFAP levels was observed in animals lesioned with KA and treated with Mut siRNAs (right hemisphere) and in those lesioned with KA (lateral ventricle) without further treatment, when compared to control animals, which received a single intracereõbroventricular injection of PBS. Similar results were obtained when Tf-lipoplexes were administered both pre- (c-Jun + KA and Mut + KA) and post-lesion (c-Jun/KA and Mut/KA). In contrast, GFAP levels in the right hemisphere of animals injected with anti-c-Jun siRNAs were equal to those of the control group, independently of whether Tf-lipoplexes had been administered 3 days before or immediately after KA injection (Fig. 7A). Significant changes in the number of astrocytes could also be observed in the hemispheres that did not receive anti-c-Jun siRNAs (Fig. 7C, panels m, n, o, q, r and s) as well as a notorious alteration in the

Please cite this article as: A.L.C. Cardoso, et al., Tf-lipoplex-mediated c-Jun silencing improves neuronal survival following excitotoxic damage in vivo, J. Control. Release (2009), doi:10.1016/j.jconrel.2009.11.004

GENE DELIVERY

ARTICLE IN PRESS

GENE DELIVERY

ARTICLE IN PRESS 8

A.L.C. Cardoso et al. / Journal of Controlled Release xxx (2009) xxx–xxx

Fig. 5. Neuroprotection mediated by c-Jun silencing in vivo, following excitotoxic lesion. (A) Mice were injected in the CA3 region of the right hippocampus (ipsilateral), with Tf-lipoplexes (c-Jun siRNAs or Mut siRNAs); 3 days after (c-Jun + KA; Mut + KA) or immediately before (c-Jun/KA; Mut/KA) Tf-lipoplex injection, KA (0.1 nmol–3 μl) was injected in the lateral ventricle. Sham-operated animals (PBS) received a single injection of PBS (3 μl). Animals were sacrificed 3 days postinjection. (B) Neuronal death was evaluated by cresyl violet staining and immunohistochemistry. (a–j) Representative light microscopy images of both hemispheres stained with cresyl violet are shown at 200× magnification. (k–t) Sections were labelled with anti-NeuN antibody (red), a specific marker of neurons. Representative fluorescence microscopy images are shown at 200× magnification. (C) Quantification of neuronal death. Results are expressed as the percentage of intact neurons with respect to the total CA3 neurons per hemisphere. (D) Quantification of NeuN loss. Results are expressed as a percentage of NeuN positive neurons in sham-operated animals. Results in (C) and (D) are presented as mean values ± SD obtained from cell counts performed in every six section (n = 6). ⁎⁎⁎p b 0.001 compared to animals injected with KA in the absence of Tf-lipoplex treatment.

morphology and activation state of microglia (Fig. 7C, panels c, d, e, g, h and i). An increase in GFAP and CD11b labelling was observed in the CA3 region of these hemispheres, specially surrounding the area where neuronal loss was most evident. On the other hand, GFAP (Fig. 7C, panels p and t) and CD11b (Fig. 7C, panels f and j) labelling in the hemispheres treated with anti-c-Jun siRNAs remained similar to that of the sham-

operated group (Fig. 7C, panels a, b, k and l), with very few CD11b positive cells and a small number of astrocytes. Regarding cytokine production, the mRNA levels of some relevant inflammatory cytokines (IL-1β, IL-6 and TNF-α) were analysed by QRT-PCR, following Tf-mediated treatment and KA administration (Fig. 8).

Please cite this article as: A.L.C. Cardoso, et al., Tf-lipoplex-mediated c-Jun silencing improves neuronal survival following excitotoxic damage in vivo, J. Control. Release (2009), doi:10.1016/j.jconrel.2009.11.004

A.L.C. Cardoso et al. / Journal of Controlled Release xxx (2009) xxx–xxx

Fig. 6. Decrease in seizure activity mediated by c-Jun silencing, following kainic acid injection. Immediately before (c-Jun/KA; Mut/KA) or 3 days after (c-Jun + KA; Mut + KA) Tf-lipoplex injection in the CA3 region of the right hippocampus, mice were injected with KA (0.1 nmol–3 μl) in the lateral ventricle. All mice were monitored for 4 h for the onset and extent of seizure activity, 1 day after KA injection. Seizures were rated according to the previously described Racine's scale. The number of animals reaching each stage of status epilepticus was determined for all experimental conditions. Results are expressed as a percentage of the total number of animals per experimental condition (n= 12).

Cytokine mRNA levels were measured in the right hemisphere of animals treated with Tf-lipoplexes carrying anti-c-Jun or Mut siRNAs, 1 day after KA injection and the results were compared to the cytokine mRNA levels of animals lesioned with KA in the absence of Tflipoplex-mediated treatment (control group). In parallel experiments, the mRNA levels of IL-1β, IL-6 and TNF-α were also evaluated in sham-operated animals, in order to determine cytokine levels in undamaged tissue. While cytokine mRNA levels of animals treated with Mut siRNAs, pre- or post-lesion, remained similar to those of the control group, a significant reduction of the tested cytokines was found in the animals treated with anti-c-Jun siRNAs, which presented IL-1β and TNF-α mRNA levels similar to those observed in the shamoperated group. It is interesting to note that, while IL-6 levels found in PBS injected animals are similar to those of the control group, indicating that IL-6 is not upregulated 1 day after KA-induced lesion, c-Jun silencing was able to induce a reduction of IL-6 transcription. Overall, these data suggest that the neuroprotective effect of Tflipoplex-mediated c-Jun silencing is related not only to a decrease in neuronal cell death, but also to the modulation of the inflammatory reaction that follows an excitotoxic lesion. 4. Discussion Gene therapy based on RNA interference has been considered a promising therapeutic approach to a variety of neurological disorders for which current treatments are largely unsatisfactory [23]. Recent reports support the use of RNAi both as a research tool, to uncover the complexity of intracellular signalling pathways involved in cell death, and as a new therapeutic strategy for genetic and sporadic neurodegenerative diseases [24–27]. Nevertheless, few studies have been performed using RNAi applications in acute brain disorders, such as stroke, trauma and epilepsy. In this work, using a lipid-based delivery vector in a model of excitotoxic brain lesion we were able to demonstrate the potential of RNA interference to identify and validate a new promising target for the treatment of acute brain injury in vivo. Several studies have implicated the JNK protein family and its downstream target, the transcription factor c-Jun, in the neurodegenerative events resulting from overstimulation of glutamate receptors [11,14,28]. c-Jun is known to activate the transcription of several immediate early genes encoding proteins [29], which include pro-apoptotic mithochondrial proteins and inflammatory cytokines [30]. Moreover, evidence suggests that the AP-1 transcription factor, composed of c-Jun and c-Fos dimers, is able to induce the transcription of c-Jun itself, leading to a perpetuation of its activity. In the

9

present work, we investigated the time-course of c-Jun activation in the context of excitotoxic injury in neurons. In this model system, early translocation of c-Jun to the nucleus (Fig. 1B and C) and a significant increase in c-Jun mRNA were observed, as soon as 15 min after glutamate exposure (Fig. 1A). These data strongly correlate with previous reports on JNK and c-Jun activation after excitotoxic lesion and suggest that c-Jun nuclear translocation and upregulation are early events of the signalling cascades that occur within a few minutes after overstimulation of glutamate receptors. Similarly to other MAPK family members, the role of JNKs and c-Jun can be neuroprotective or neurotoxic, depending on the cell type and pathological context of activation. The direct inhibition of JNKs in neurons has provided substantial protection against various neurodegenerative stimulus including ischemia and excitotoxicity [28,31,32], suggesting a pro-apoptotic role for these enzymes after acute brain damage. In order to clarify the relevance of c-Jun activation in this context, we evaluated the effect of RNAi-mediated c-Jun silencing in neuronal viability following glutamate insult or oxygen–glucose deprivation, two well established models of excitotoxic damage. Aiming at achieving efficient c-Jun knockdown in neuronal primary cultures, a nonviral strategy previously developed in our laboratory was employed to mediate efficient delivery of siRNA molecules. Tf-lipoplexes, generated from complexation of Tf-associated cationic liposomes with anti-c-Jun siRNAs, where shown to promote efficient c-Jun knockdown in primary neuronal cultures, similarly to that demonstrated for reporter genes, such as luciferase [16]. Following demonstration of both mRNA and protein knockdown in vitro mediated by Tf-lipoplexes (Fig. 2A, B and C), neuronal viability was determined 18 h after both excitotoxic insults (Fig. 2D and E) and neuronal cell death was found to be significantly decreased when cells were pre-treated with anti-c-Jun siRNAs, but not when cells were treated with Mut siRNAs. These results correlate with indirect observation made in previous studies [14] and strongly suggest that c-Jun is directly involved in neuronal death after excitotoxic damage. Kainic acid is known to cause epileptic seizures and to increase the levels of extracellular glutamate, thereby mimicking some of the events that occur after excitotoxic lesion in several neurodegenerative diseases and acute brain disorders. Here, we used the kainate lesion model to evaluate the efficiency of c-Jun siRNA Tf-lipoplexes to provide neuroprotection and attenuate inflammatory responses in vivo. Considering the difficulties associated with nucleic acid delivery in vivo, we first evaluated whether Tf-lipoplexes could lead to successful c-Jun knockdown after local brain injection. Our results clearly confirm the potential of this strategy to promote in vivo siRNA delivery and c-Jun knockdown in the mouse brain (Fig. 3). In fact, we observed an impressive 70% reduction of c-Jun mRNA levels 1 day after Tf-lipoplex injection, accompanied by a 45% decrease in protein levels 3 days after siRNA administration mediated by Tf-lipoplexes, which could be sustained for at least 5 days. This knockdown was found to be specific, since no decrease in c-Jun levels was observed upon treatment with Mut siRNAs. Moreover, Tf-lipoplexes did not induce any kind of toxicity or inflammatory response in the mouse brain (Fig. 4), showing their high biocompatibility. In the presently applied kainate lesion model, a clear and reproducible neurodegeneration of CA3 pyramidal neurons could be observed in the hippocampus of both hemispheres as soon as 3 days following KA injection. In order to evaluate the neuroprotective effect of anti-c-Jun siRNAs, when delivered by Tf-lipoplexes, Tf-lipoplex injection was performed near the CA3 region, 3 days before or immediately after KA injection (Fig. 5). An impressive reduction of lesion size could be observed in all hemispheres treated with anti-c-Jun siRNAs (Fig. 5), independently of whether the siRNA was delivered 3 days before or immediately after the lesion. In addition, a strong decrease in seizure activity was observed in the animals pre-treated with anti-c-Jun siRNAs (Fig. 6), which translated into a significant decrease in the number of death animals due to KA-induced convulsions. These findings indicate that c-Jun silencing results in a neuroprotective and repairing effect in

Please cite this article as: A.L.C. Cardoso, et al., Tf-lipoplex-mediated c-Jun silencing improves neuronal survival following excitotoxic damage in vivo, J. Control. Release (2009), doi:10.1016/j.jconrel.2009.11.004

GENE DELIVERY

ARTICLE IN PRESS

GENE DELIVERY

ARTICLE IN PRESS 10

A.L.C. Cardoso et al. / Journal of Controlled Release xxx (2009) xxx–xxx

Fig. 7. Anti-inflammatory potential of c-Jun silencing mediated by Tf-lipoplexes. Immediately before (c-Jun/KA; Mut/KA) or 3 days after (c-Jun + KA; Mut + KA) Tf-lipoplex delivery in the right hippocampus, mice were injected with KA (0.1 nmol–3 μl) in the lateral ventricle. In parallel experiments, animals received a single injection of KA (0.1 nmol–3 μl) or PBS in the lateral ventricle. Animals were sacrificed 3 days postinjection. The presence of an inflammatory reaction was investigated in all experimental groups. (A, B) GFAP levels were analysed by Western blot. Results from the quantification of GFAP levels corrected for individual α-tubulin levels are presented in (A) and are expressed as fold increase above GFAP levels in sham-operated animals. Results correspond to mean values ± SD (n = 6). ⁎⁎p b 0.01, ⁎⁎⁎p b 0.001 and n.s. (without statistical significance) compared to the sham-operated group. A representative gel showing GFAP levels for the most relevant experimental conditions is shown in (B). (C) Microglia activation and gliosis were assessed by immunohistochemistry. Sections were labelled with (a–j) anti-CD11b antibody (red) or (k–t) anti-GFAP antibody (green). Representative fluorescence microscopy images are presented at 200× magnification.

this model of excitotoxic injury, which seems to be specific for this target, since it could not be achieved with a Mut siRNA sequence. Kainic acid-induced lesion was also found to be accompanied by a strong inflammatory reaction, characterized by astrogliosis and activation of microglia. While in a healthy brain, resting microglia and astrocytes contact neuronal cells in order to survey any changes in the neuronal environment, in damaged brains astrocytes proliferate at

a high rate and microglia cells are activated and may damage neurons. Within 1 to 2 days after KA injection, resident microglia cells change their morphology and migrate into injured regions. After KA injection, the number of activated microglia cells and astrocytes increases during 3 and 4 days due to proliferation, and a high number of these cells can still be observed one month after the onset of injury [33]. Moreover, production of several cytokines can also be detected in the

Please cite this article as: A.L.C. Cardoso, et al., Tf-lipoplex-mediated c-Jun silencing improves neuronal survival following excitotoxic damage in vivo, J. Control. Release (2009), doi:10.1016/j.jconrel.2009.11.004

A.L.C. Cardoso et al. / Journal of Controlled Release xxx (2009) xxx–xxx

11

this study was shown to last at least five days, could reduce neuronal apoptosis in the crucial period that follows stroke or an epileptic seizure. This approach prevents loss of neuronal function in the affected areas and, therefore, represents a therapeutic alternative to the use of JNK and AP-1 systemic inhibitors, which although somewhat effective may lead to serious side-effects in other organs. Acknowledgements Fig. 8. Reduction of inflammatory cytokine expression following Tf-lipoplex-mediated c-Jun silencing. Immediately before (c-Jun/KA; Mut/KA) or 3 days after (c-Jun + KA; Mut + KA) Tf-lipoplex injection in the right hippocampus (ispilateral), mice were injected with KA (0.1 nmol–3 μl) in the lateral ventricle. In parallel experiments, animals received a single injection of KA or PBS in the lateral ventricle. Animals were sacrificed 1 day postinjection. The mRNA levels of IL-1β, IL-6 and TNF-α in the right (ipsilateral) hemispheres were assessed by QRT-PCR and are expressed as fold reduction with respect to cytokine levels in the KA group. Results are presented as mean values± SD (n = 6), representative of three independent experiments. ⁎⁎⁎p b 0.001, ⁎⁎p b 0.01 and ⁎p b 0.05 compared to animals injected with KA in the absence of Tf-lipoplex treatment.

lesioned tissue, such as TNF-α, IL-6 and IL-1β, which are key players in cerebral inflammation and neurodegeneration [34]. Upon Tf-lipoplexmediated c-Jun silencing, an attenuation of microglia activation and astrocyte proliferation could be observed in the treated hemispheres, when compared with the untreated or Mut siRNA-treated hemispheres (Fig. 7). Moreover, a significant decrease in TNF-α, IL-6 and IL-1β was also observed in the hemispheres showing a decrease in microglia activation (Fig. 8). Our findings indicate that c-Jun silencing not only protects neurons from cell death but may also attenuate microglia activation and astrogliosis, thereby extending the therapeutic relevance of c-Jun silencing to anti-inflammatory activity. Although this activity could be attributed simply to the decrease in neuronal cell death due to c-Jun silencing in neurons, which would be sufficient per se to reduce microglia activation, the observed effect can also be partially related to c-Jun knockdown in microglial cells. It has been recently demonstrated that JNK inhibition in microglia significantly reduces TNF-α and IL-6 production [30] and that c-Jun/AP-1 consensus sequences can be found in the promoters of the genes coding for these interleukins. It is therefore possible that c-Jun silencing in microglia cells will directly lead to a reduction in c-Jun/AP-1-mediated expression of these proinflammatory molecules, contributing to the anti-inflammatory effect observed after Tf-lipoplex delivery of anti-c-Jun siRNAs. Overall, our work illustrates how RNA interference technology can help to identify the most relevant proteins responsible for the progression of cell death following excitotoxic injury, and to silence these same proteins, achieving a therapeutic effect. In this study we show that Tf-lipoplexes promote siRNA delivery and siRNA-mediated protein silencing in the brain, following stereotactic injection, with high efficiency and minimum toxicity. Although this kind of administration is not as easy to perform as oral or systemic delivery, stereotactic injection of a non-viral siRNA delivery system in a specific brain region is feasible in humans and can be accomplished using common neurosurgical procedures. Moreover, Tf-lipoplexes can be further optimized for systemic delivery, since Tf receptors are present not only in neuronal cells, but also in the endothelial cells that constitute the BBB. This opens the possibility of developing new therapeutic strategies to neurological disorders based on the RNA interference technology. Once the major barrier of brain delivery is surpassed, the chances of success of this kind of therapy increase, especially if long-term silencing is not required, such as in the case of acute brain disorders where neuronal loss occurs in a short period of time following injury. The results presented in this work also provide evidence that the transcription factor c-Jun is a promising therapeutic target, whose silencing leads to significant neuroprotection after excitotoxic injury, mediated by both anti-apoptotic and anti-inflammatory effects. This suggests that c-Jun silencing by siRNAs, which in

A.L.C. Cardoso is the recipient of a fellowship from the Portuguese Foundation for Science and Technology (SFRH/BD/17216/2004). This work was partially financed by a grant from the Portuguese Foundation for Science and Technology (PTDC/BIO/65627/2006). References [1] A.K. Liou, R.S. Clark, D.C. Henshall, X.M. Yin, J. Chen, To die or not to die for neurons in ischemia, traumatic brain injury and epilepsy: a review on the stress-activated signaling pathways and apoptotic pathways, Progress In Neurobiology 69 (2003) 103–142. [2] D.W. Choi, S.M. Rothman, The role of glutamate neurotoxicity in hypoxic-ischemic neuronal death, Annual Review Of Neuroscience 13 (1990) 171–182. [3] P. Nicotera, M. Ankarcrona, E. Bonfoco, S. Orrenius, S.A. Lipton, Neuronal necrosis and apoptosis: two distinct events induced by exposure to glutamate or oxidative stress, Advances In Neurology 72 (1997) 95–101. [4] S.L. Mehta, N. Manhas, R. Raghubir, Molecular targets in cerebral ischemia for developing novel therapeutics, Brain Research Reviews 54 (2007) 34–66. [5] M.A. Hossain, Molecular mediators of hypoxic-ischemic injury and implications for epilepsy in the developing brain, Epilepsy & Behavior 7 (2005) 204–213. [6] T. Sugino, K. Nozaki, N. Hashimoto, Activation of mitogen-activated protein kinases in gerbil hippocampus with ischemic tolerance induced by 3-nitropropionic acid, Neuroscience Letters 278 (2000) 101–104. [7] B.H. Han, D.M. Holtzman, BDNF protects the neonatal brain from hypoxicischemic injury in vivo via the ERK pathway, Journal of Neuroscience 20 (2000) 5775–5781. [8] J. Harada, M. Sugimoto, An inhibitor of p38 and JNK MAP kinases prevents activation of caspase and apoptosis of cultured cerebellar granule neurons, Japanese Journal of Pharmacology 79 (1999) 369–378. [9] Z. Xia, M. Dickens, J. Raingeaud, R.J. Davis, M.E. Greenberg, Opposing effects of ERK and JNK-p38 MAP kinases on apoptosis, Science (New York, N.Y.) 270 (1995) 1326–1331. [10] S. Brecht, R. Kirchhof, A. Chromik, M. Willesen, T. Nicolaus, G. Raivich, J. Wessig, V. Waetzig, M. Goetz, M. Claussen, D. Pearse, C.Y. Kuan, E. Vaudano, A. Behrens, E. Wagner, R.A. Flavell, R.J. Davis, T. Herdegen, Specific pathophysiological functions of JNK isoforms in the brain, The European Journal Of Neuroscience 21 (2005) 363–377. [11] D.D. Yang, C.Y. Kuan, A.J. Whitmarsh, M. Rincon, T.S. Zheng, R.J. Davis, P. Rakic, R.A. Flavell, Absence of excitotoxicity-induced apoptosis in the hippocampus of mice lacking the Jnk3 gene, Nature 389 (1997) 865–870. [12] T. Borsello, P.G. Clarke, L. Hirt, A. Vercelli, M. Repici, D.F. Schorderet, J. Bogousslavsky, C. Bonny, A peptide inhibitor of c-Jun N-terminal kinase protects against excitotoxicity and cerebral ischemia, Nature Medicine 9 (2003) 1180–1186. [13] M. Ameyar, M. Wisniewska, J.B. Weitzman, A role for AP-1 in apoptosis: the case for and against, Biochimie 85 (2003) 747–752. [14] J. Whitfield, S.J. Neame, L. Paquet, O. Bernard, J. Ham, Dominant-negative c-Jun promotes neuronal survival by reducing BIM expression and inhibiting mitochondrial cytochrome c release, Neuron 29 (2001) 629–643. [15] A.L. Cardoso, S. Simoes, L.P. de Almeida, J. Pelisek, C. Culmsee, E. Wagner, M.C. Pedroso de Lima, siRNA delivery by a transferrin-associated lipid-based vector: a non-viral strategy to mediate gene silencing, The Journal Of Gene Medicine 9 (2007) 170–183. [16] A.L. Cardoso, S. Simoes, L.P. de Almeida, N. Plesnila, M.C. Pedroso de Lima, E. Wagner, C. Culmsee, Tf-lipoplexes for neuronal siRNA delivery: a promising system to mediate gene silencing in the CNS, Journal Of Controlled Release 132 (2008) 113–123. [17] M.J. Campbell, Lipofection reagents prepared by a simple ethanol injection technique, BioTechniques 18 (1995) 1027–1032. [18] C. Culmsee, C. Zhu, S. Landshamer, B. Becattini, E. Wagner, M. Pellecchia, K. Blomgren, N. Plesnila, Apoptosis-inducing factor triggered by poly(ADP-ribose) polymerase and Bid mediates neuronal cell death after oxygen–glucose deprivation and focal cerebral ischemia, Journal Of Neuroscience 25 (2005) 10262–10272. [19] G.F. Paxinos, K.B.J., The Mouse Brain in Stereotaxic Coordinatessecond edition ed, Elsevier Academic Press, San Diego, 2004. [20] R. Racine, V. Okujava, S. Chipashvili, Modification of seizure activity by electrical stimulation. 3. Mechanisms, Electroencephalography and Clinical Neurophysiology 32 (1972) 295–299. [21] H.K. Lee, S.S. Choi, E.J. Han, J.Y. Lee, M.S. Kwon, E.J. Shim, Y.J. Seo, H.W. Suh, Role of nicotinic acetylcholine receptors in the regulation of kainic acid-induced hippocampal cell death in mice, Brain Research Bulletin 64 (2004) 309–317. [22] E.J. Shim, Y.J. Seo, M.S. Kwon, Y.O. Ham, O.S. Choi, J.Y. Lee, S.M. Choi, H.W. Suh, The intracerebroventricular kainic acid-induced damage affects animal nociceptive behavior, Brain Research Bulletin 73 (2007) 203–209. [23] V.M. Miller, H. Xia, G.L. Marrs, C.M. Gouvion, G. Lee, B.L. Davidson, H.L. Paulson, Allele-specific silencing of dominant disease genes, Proceedings of the National Academy of Sciences of the United States of America 100 (2003) 7195–7200.

Please cite this article as: A.L.C. Cardoso, et al., Tf-lipoplex-mediated c-Jun silencing improves neuronal survival following excitotoxic damage in vivo, J. Control. Release (2009), doi:10.1016/j.jconrel.2009.11.004

GENE DELIVERY

ARTICLE IN PRESS

GENE DELIVERY

ARTICLE IN PRESS 12

A.L.C. Cardoso et al. / Journal of Controlled Release xxx (2009) xxx–xxx

[24] T.M. Fountaine, R. Wade-Martins, RNA interference-mediated knockdown of alphasynuclein protects human dopaminergic neuroblastoma cells from MPP(+) toxicity and reduces dopamine transport, Journal of Neuroscience Research 85 (2007) 351–363. [25] P. Gonzalez-Alegre, N. Bode, B.L. Davidson, H.L. Paulson, Silencing primary dystonia: lentiviral-mediated RNA interference therapy for DYT1 dystonia, Journal of Neuroscience 25 (2005) 10502–10509. [26] S.Q. Harper, P.D. Staber, X. He, S.L. Eliason, I.H. Martins, Q. Mao, L. Yang, R.M. Kotin, H.L. Paulson, B.L. Davidson, RNA interference improves motor and neuropathological abnormalities in a Huntington's disease mouse model, Proceedings of the National Academy of Sciences of the United States of America 102 (2005) 5820–5825. [27] V.M. Miller, C.M. Gouvion, B.L. Davidson, H.L. Paulson, Targeting Alzheimer's disease genes with RNA interference: an efficient strategy for silencing mutant alleles, Nucleic Acids Research 32 (2004) 661–668. [28] Y. Gao, A.P. Signore, W. Yin, G. Cao, X.M. Yin, F. Sun, Y. Luo, S.H. Graham, J. Chen, Neuroprotection against focal ischemic brain injury by inhibition of c-Jun N-terminal kinase and attenuation of the mitochondrial apoptosis-signaling pathway, Journal of Cerebral Blood Flow and Metabolism 25 (2005) 694–712. [29] J.C. Sng, H. Taniura, Y. Yoneda, A tale of early response genes, Biological & Pharmaceutical Bulletin 27 (2004) 606–612.

[30] V. Waetzig, K. Czeloth, U. Hidding, K. Mielke, M. Kanzow, S. Brecht, M. Goetz, R. Lucius, T. Herdegen, U.K. Hanisch, c-Jun N-terminal kinases (JNKs) mediate proinflammatory actions of microglia, Glia 50 (2005) 235–246. [31] D. Bozyczko-Coyne, T.M. O'Kane, Z.L. Wu, P. Dobrzanski, S. Murthy, J.L. Vaught, R.W. Scott, CEP-1347/KT-7515, an inhibitor of SAPK/JNK pathway activation, promotes survival and blocks multiple events associated with Abeta-induced cortical neuron apoptosis, Journal of Neurochemistry 77 (2001) 849–863. [32] A.C. Maroney, M.A. Glicksman, A.N. Basma, K.M. Walton, E. Knight Jr., C.A. Murphy, B.A. Bartlett, J.P. Finn, T. Angeles, Y. Matsuda, N.T. Neff, C.A. Dionne, Motoneuron apoptosis is blocked by CEP-1347 (KT 7515), a novel inhibitor of the JNK signaling pathway, Journal of Neuroscience 18 (1998) 104–111. [33] S. Hasegawa, M. Yamaguchi, H. Nagao, M. Mishina, K. Mori, Enhanced cell-to-cell contacts between activated microglia and pyramidal cell dendrites following kainic acid-induced neurotoxicity in the hippocampus, Journal of Neuroimmunology 186 (2007) 75–85. [34] T. Owens, H. Wekerle, J. Antel, Genetic models for CNS inflammation, Nature Medicine 7 (2001) 161–166.

Please cite this article as: A.L.C. Cardoso, et al., Tf-lipoplex-mediated c-Jun silencing improves neuronal survival following excitotoxic damage in vivo, J. Control. Release (2009), doi:10.1016/j.jconrel.2009.11.004