SEARCH

SEARCH BY CITATION

Keywords:

  • AP-1;
  • c-Jun;
  • glutamate;
  • neuroprotection;
  • phylomer peptides;
  • TAT

Abstract

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

J. Neurochem. (2010) 112, 258–270.

Abstract

Neuronal cell death caused by glutamate excitotoxicity is prevalent in various neurological disorders and has been associated with the transcriptional activation of activator protein-1 (AP-1). In this study, we tested 19 recently isolated AP-1 inhibitory peptides, fused to the cell penetrating peptide TAT, for their efficacy in preventing cell death in cortical neuronal cultures following glutamate excitotoxicity. Five peptides (PYC19D-TAT, PYC35D-TAT, PYC36D-TAT, PYC38D-TAT, PYC41D-TAT) displayed neuroprotective activity in concentration responses in both l- and retro-inverso d-isoforms with increasing levels of neuroprotection peaking at 83%. Interestingly, the D-TAT peptide displayed a neuroprotective effect increasing neuronal survival to 25%. Using an AP-1 luciferase reporter assay, we confirmed that the AP-1 inhibitory peptides reduce AP-1 transcriptional activation, and that c-Jun and c-Fos mRNA following glutamate exposure is reduced. In addition, following glutamate exposure the AP-1 inhibitory peptides decreased calpain-mediated α-fodrin cleavage, but not neuronal calcium influx. Finally, as neuronal death following glutamate excitotoxicity was transcriptionally independent (actinomycin D insensitive), our data indicate that activation of AP-1 proteins can induce cell death via non-transcriptional pathways. Thus, these peptides have potential application as therapeutics directly or for the rational design of small molecule inhibitors in both apoptotic and necrotic neuronal death associated with AP-1 activation.

Abbreviations used:
BSS

balance salt solution

DIV

day in vitro

DMEM

Dulbecco’s modified Eagle medium

MTS

3-(4,5,dimethyliazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium salt

NB

neurobasal

PBS

phosphate-buffered saline

Neuronal cell death caused by glutamate excitotoxicity is associated with cerebral ischemia, traumatic brain injury, epilepsy and neurodegenerative disorders including Parkinson’s disease, Alzheimer’s disease and amyotrophic lateral sclerosis (Faden et al. 1989; Bradford 1995; Mattson et al. 1999; Coyle 2006; Lizasoain et al. 2006). Glutamate over-stimulation of NMDA, α-amino-3-hydroxy-5-methylisoxazole-4-propionate and kainic acid (KA) receptors results in an influx of extracellular calcium and sodium and the release of calcium from intracellular stores. Increased intracellular calcium initiates a range of cell damaging processes involving phospholipases, proteases, kinases, phosphatases and ROS/RNS that can trigger a host of neuronal cell death pathways. Depending on the intensity of glutamate receptor over-stimulation, the resulting neuronal death can display a variety of features ranging from mostly necrotic to mostly apoptotic (Choi et al. 1987; Lipton and Rosenberg 1994). Moreover, regardless of the neuronal cell death outcome, glutamate receptor stimulation activates the c-Jun N-terminal kinase (JNK) and activator protein-1 (AP-1) transcription factor complex pathways, which are usually associated with apoptosis (Schwarzschild et al. 1997; Zhang et al. 2006). While the inhibition of JNK and JNK/AP-1 pathways have been shown to inhibit both glutamate-induced necrotic and apoptotic cell death respectively, no study has assessed the effect of AP-1 inhibition on neuronal death using a glutamate necrotic excitotoxic model.

Apoptotic cell death involving the AP-1 complex is almost invariably associated with the transcriptional activation of the AP-1 protein family member c-Jun and to a lesser extent c-Fos. Activation requires phosphorylation of c-Jun by the stress induced mitogen-activated protein kinase, JNK (Hibi et al. 1993; Karin 1995). Phosphorylated c-Jun can homodimerize or heterodimerize with c-Fos, and subsequently promote AP-1 complex transcription of pro-apoptotic cell death proteins (e.g Fas-L, p53, Bax) (Mandal et al. 2001; Lauricella et al. 2006). As c-Jun and c-Fos are only transcriptionally active in their dimeric form, repression of dimerization presents an attractive target to inhibit cell death. To this end, following forward and reverse yeast two-hybrid screens using a Phylomer peptide library (Watt 2006), we identified 19 peptides (referred to hereafter as AP-1 inhibitory peptides) that bind to the c-Jun dimerization domain and down-regulate AP-1 transcription (unpublished data). In the present study, we assessed the ability of these 19 peptides to inhibit AP-1 activation and to provide neuroprotection in primary cortical neuronal cultures using a necrotic cell death glutamate excitotoxicity model (Schwarzschild et al. 1997; Eminel et al. 2008).

Materials and methods

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Primary cortical neuronal cultures

Establishment of cortical cultures was as previously described (Meloni et al. 2001). Briefly, cortical tissue from E18-E19 Sprague-Dawley rats was dissociated in Dulbecco’s modified Eagle medium (DMEM; Invitrogen, Melbourne, Australia) supplemented with 1.3 mM l-cysteine, 0.9 mM NaHCO3, 10 units/mL papain (Sigma, St Louis, MO, USA) and 50 units/mL Dnase (Sigma) and washed in cold DMEM/10% horse serum. Neurons were resuspended in Neurobasal (NB; Invitrogen) containing 2% B27 supplement (B27; Invitrogen). Before seeding, culture vessels [96-well plastic plate, 6-well plastic plate, or 24-well plastic plate with 13 mm glass coverslips (ProSciTech, Thuringowa, Australia)], were coated with poly-d-lysine (50 μg/mL; 70–150 K; Sigma) and incubated overnight at 21°C. Excess poly-d-lysine solution was removed and replaced with NB (containing 2% B27; 4% fetal bovine serum; 1% horse serum; 62.5 μM glutamate; 25 μM 2-mercaptoethanol; and 30 μg/mL streptomycin and 30 μg/mL penicillin). Neurons were plated to obtain approximately 10 000 viable neurons for each well of a 96-well plate, 200 000 viable neurons per well of a 24-well plate, or 1 500 000 viable neurons per well of a 6-well plate on day in vitro (DIV) 11. Neuronal cultures were maintained in a CO2 incubator (5% CO2, 95% air balance, 98% humidity) at 37°C. On DIV 4 one-third of the NB culture medium was removed and replaced with fresh NB/2% B27 containing the mitotic inhibitor, cytosine arabinofuranoside (Sigma) at 1 μM. On DIV 8 one half of the culture medium was replaced with NB/2% B27. Cultures were used on DIV 11 or 12 and consisted of > 95% neurons (Meloni et al. 2001).

Glutamate excitotoxicity model

To induce excitotoxicity in the cortical neuronal cultures (96-well plate format) 50 μL of conditioned media (media in which the neurons were previously maintained) containing glutamate at 200 μM concentration was added to culture wells containing 50 μL of conditioned media (100 μM final glutamate concentration). Cultures were incubated at 37°C in the CO2 incubator for 5 min after which time the media was replaced with 100 μL of 50% NB/2% N2 and 50% balance salt solution (NB/N2:BSS). In the majority of the experiments peptides were added to wells 15 min prior to glutamate exposure. However, for the administration, time-course experiment peptides were added either prior to (15 min) or at different times post-glutamate exposure (0, 15, 30, 45 or 60 min), while in the calcium influx experiments peptide was added both prior to and post-glutamate exposure. A non-peptide positive control consisting of the glutamate receptor blockers 5 μM MK801/5 μM 6-cyano-7-nitroquinoxaline (MK801/CNQX) was added in a similar manner to the peptides, either prior to or post-glutamate exposure. The glutamate-treated controls and untreated controls received media additions with and without glutamate respectively.

AP-1 inhibitory peptides and control peptides

The peptides were synthesized and HPLC purified by Mimotopes Pty Ltd (Clayton, Australia), or where indicated by GenScript Corporation (Piscataway, NJ, USA) or Auspep (Parkville, Australia) (Table 1). Peptides were fused to a truncated HIV-1 TAT(48–57) transduction domain peptide (TAT) at the amino or carboxyl terminus to allow peptides to enter cells (Vives et al. 1997). Peptides were initially assessed in the l-isoform at 5 μM, with a selection of peptides later being evaluated in the protease resistant d-retro-inverso form, synthesized from d-amino acids in reverse sequence (referred to hereafter as d-isoforms) (Brugidou et al. 1995). A TAT-fused JNK inhibitory peptide (JNKI-1D-TAT) in the d-isoform was used as a positive control (Borsello et al. 2003). Other controls consisted of the d-isoform TAT peptide (D-TAT), L-isoform c-Jun leucine zipper-TAT peptide (c-JunZIPL-TAT), and PYC35D and PYC36D peptides without TAT and in a scrambled peptide sequence fused to TAT (PYC35DScram-TAT, PYC35DScram-TAT). All peptides were prepared as 100× stocks (500 μM) in normal saline.

Table 1.   AP-1 inhibitory and control peptide amino acid sequences
Peptide identificationAmino acid sequence*
  1. *At the N-terminus, H indicates free amine, at the C-terminus NH2 indicates amide and OH indicates free acid. Underlined single letter code indicates d-isoform of the amino acid. Transduction domain is indicated by bold lettering.

PYC8L-TATH-GRKKRRQRRRGAYQSIIANEEEREKNFASSKKDGSYTDLL-NH2
PYC16L-TATH-GRKKRRQRRRGVYQSIRYIKSLLKKSVIYMHMHRL-NH2
PYC16D-TATH-LRHMHMYIVSKKLLSKIYRISQYVGRRRQRRKKRG-NH2
PYC19L-TATH-GRKKRRQRRRGLINQSYAYPYIY-NH2
PYC20L-TATH-GRKKRRQRRRGKIYSSILWGTKWCVLLVITP-NH2 (GenScript)
PYC22L-TATH-GRKKRRQRRRGKDSIRRRGENISSQEVEAVLMSHPEVVNAAVYPVRGDLPGD-NH2 (GenScript)
PYC24L-TATH-GRKKRRQRRRGLYQSLLTATKELLFVAPVAKAFTSCD-NH2 (GenScript)
PYC32L-TATH-GRKKRRQRRRGNTPHSSDGHNNP-NH2
PYC34L-TATH-GRKKRRQRRRGAYQSINSSPVLPENSQELSLHQHVTKS-NH2 (Auspep)
PYC35L-TATH-GRKKRRQRRRGAYQSIRSGGIESSSKRER-NH2 (Auspep)
PYC35D-TATH-RERKSSSEIGGSRISQYAGRRRQRRKKRG-OH
PYC35L-FMH-AYQSIRSGGIESSSKRERGAAVLLPVLLAAP-NH2
PYC35D-FMH-PAALLVPLLVAAGRERKSSSEIGGSRISQYA-NH2
PYC35DH-RERKSSSEIGGSRISQYAG-OH
PYC35DScram-TATH-KIERSEGISQSARSRGYSGRRRQRRKKRG-OH
PYC36L-TATH-GRKKRRQRRRGGLQGRRRQGYQSIKP-NH2 (Auspep)
PYC36D-TATH-PKISQYGQRRRGQLGGRRRQRRKKRG-OH
PYC36L-FMH-GLQGRRRQGYQSIKPGAAVLLPVLLAAP-NH2
PYC36D-FMH-PAALLVPLLVAAGPKISQYGQRRRGQLG-NH2
PYC36DH-PKISQYGQRRRGQLG-OH
PYC36DScram-TATH-KRRGGILRYGQPQSQGRRRQRRKKRG-OH
PYC38L-TATH-GRKKRRQRRRGGLQGRRQPGQQPQGRWSGRALPAHR-NH2
PYC41L-TATH-GRKKRRQRRRGVSINQEHHRLLPL-NH2
PYC45L-TATH-GRKKRRQRRRGAYQSIGSIWNSCQCMSFWCAFVRSCYGPGRGWMKPKRRRVPGLKS-NH2 (GenScript)
PYC54L-TATH-GRKKRRQRRRGKAYQSFLLAGYHGDTSRTFLVGSVSATARKLVEATQETMIDYT-NH2 (GenScript)
PYC58L-TATH-GRKKRRQRRRGSIMAVAAQQPVAFLVGRQRRRGQVGIDSGDQHLRTPLFHEL-NH2 (GenScript)
PYC60L-TATH-GRKKRRQRRRGANQLPKILAGRLIFIKC-NH2
PYC60D-TATH-CKIFILRGALIKPLQNAGRRRQRRKKRG-NH2
PYC66L-TATH-GRKKRRQRRRGAYQSIIGAGKSTLIKALTGVYHADRGTIWLEGQAISPKNTAHAQQ-NH2 (GenScript)
PYC67L-TATH-GRKKRRQRRRGVINQSTVFLKIIRFF-NH2
PYC71L-TATH-GRKKRRQRRRGPFFWRLNKIFYFIDSRVAAAILN-NH2 (Auspep)
PYC71D-TATH-NLIAAAVRSDIFYFIKNLRWFFPGRRRQRRKKRG-NH2
JNKI-1D-TATH-TDQSRPVQPFLNLTTPRKPRPPRRRQRRKKRG-NH2
D-TATH-GRRRQRRKKRG-NH2
c-JunZIPL-TATH-GRKKRRQRRRGRIARLEEKVKTLKAQNSELASTANMLREQVAQLKQKVMNH-NH2 (Auspep)

Neuronal viability assessment and statistical analysis

Neuronal cultures were examined by light microscopy for qualitative assessment of neuronal cell viability 18–24 h after glutamate exposure, and quantitatively at 24 h by 3-(4,5,dimethyliazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium salt (MTS) assay (Promega, Sydney, Australia). For propidium iodide staining (0.05 μg/mL final concentration) the dye was added 15 min prior to visualization by fluorescence microscopy. Image acquisition was performed using an Olympus IX70 inverted microscope fitted with a digital camera (DP70; Olympus, Center Valley, PA, USA) under software control (DP controller; Olympus). The MTS assay measures the mitochondrial conversion of the tetrazolium salt to a water-soluble brown formazan salt, which is detected spectrophotometrically at 495 nm. MTS absorbance data was converted to reflect proportional cell viability relative to both the untreated (taken as 100% viable) and glutamate-treated controls. Viability data was analyzed by anova, followed by post-hoc Fischer’s PLSD test, with < 0.05 values considered statistically significant, and presented as mean ± SEM. All assays were performed in quadruplicate with sister neuronal cultures and repeated a minimum of four times independently.

Luciferase assay using HEK293 AP-1 luciferase cell line

The human embryonic kidney 293 AP-1 luciferase stable cell line (HEK293 AP-1 Luc), containing three copies of the AP-1 consensus sequence upstream of a luciferase gene (Panomics, Fremont, CA, USA) was maintained in 75 cm2 flasks in DMEM, supplemented with 10% fetal bovine serum, 5 mM l-glutamine, 100 units/mL penicillin, 100 μg/mL streptomycin and 100 μg/mL hygromycin B. The day prior to the luciferase assay, cells were seeded into 24-well plates at 1 × 105 cells per well in 500 μL DMEM, without hygromycin. AP-1 inhibitory peptides (d-isoforms), D-TAT peptide and c-JunZIPL-TAT peptide were added directly to culture wells (5 μM final concentration) and incubated at 37°C for 15 min before the addition of phorbol myristate acetate (10 ng/mL final concentration) to induce luciferase expression. Plates were incubated overnight and cells subsequently lysed in 100 μL of passive lysis buffer (Promega) for 15 min and 10 μL lysate samples from each well-transferred into duplicate wells of a white 96-well plate. Following the addition of 50 μL LARII reagent (Promega) to each well, luminescence was measured using a PolarStar Luminometer (Offenburg, Germany) 10 times, every 0.2 s, with 0.2 s intervals. The average luminescence reading was taken and percentage of luciferase induction was calculated in comparison to cells transduced with the negative control D-TAT peptide (100% luciferase induction). These experiments were repeated independently three times and data analyzed by anova, followed by post-hoc Fischer’s PLSD test, with < 0.05 values considered statistically significant, and presented as mean ± SEM.

Cell viability assays of HEK293 AP-1 Luc cells transduced with peptides were performed in 96-well plates. Wells were seeded with 5 × 104 cells in 100 μL DMEM, without hygromycin. The following day cells were transduced with AP-1 inhibitory peptides, D-TAT or c-JunZIPL-TAT peptides (5 μM final concentration), incubated for 15 min and then luciferase induced with the addition of phorbol myristate acetate (10 ng/mL final concentration) overnight. MTS assays were performed after a one-hour incubation time and absorbance data analyzed as previously described.

c-Jun and c-Fos mRNA real-time-PCR

Total RNA was isolated from primary rat cortical neurons cultured in 6-well plates with the use of Trizol (Invitrogen) at designated time points. Extraction was performed with 1 mL Trizol and included a glycogen carrier step, according to the manufacturer’s instructions. RNA integrity was confirmed visually, following agarose gel electrophoresis and ethidium bromide staining. Further analysis and quantitation was performed by Nanodrop spectrophotometer. Dnase treatment of 1 μg of RNA was performed with 1 unit RQ1 Rnase-free Dnase (Promega) as per manufacturer’s instructions. Reverse transcription of the Dnase-treated RNA into cDNA was achieved using 200 units of M-MLV Rnase H-minus reverse transcriptase (Promega), 1 μg oligo dT primer (Ambion, Austin, TX, USA) and 10 mM dNTPs (Promega), as per manufacturer’s instructions. cDNA was amplified in ABI Prism 7900HT with 0.5 μM primers, designed using Primer Express Software, specific for c-Jun (Fwd: 5′-GGCT-AACCCCGCGTGAA-3′; Rvs: 5′-AAGGTCGTTTCCATCTTTGCA-3′; amplifies a 57-bp sequence), c-Fos (Fwd: 5′-CCCCTCGCCGAGCTTT-3′; Rvs: 5′-GCGTTGAAACCCGAGAACAT-3′; amplifies a 53-bp sequence), or GAPDH (Fwd: 5′-CCTGGAGAAACCTGCCAAGTAT-3′; Rvs: 5′-CTCGGCCGCCTGCTT-3′; amplifies a 57-bp sequence) with SYBR green mix (Finnzymes) in conditions set out by manufacturer. Standard curves for each primer set were produced to facilitate analysis of samples by standard curve analysis method (Bookout and Mangelsdorf 2003). These experiments were performed in duplicate and were repeated three times.

Neuronal intracellular calcium measurements

Cortical neurons grown on glass coverslips were loaded with the fluorescent calcium ion indicator fura-2AM (1 μM) in 300 μL NB/N-2:BSS, 0.3% pluronic F-127, for 45 min at 37°C. The cells were then washed once with NB/N2:BSS before the addition of 300 μL of fresh NB/N2:BSS containing peptide (5 μM) or MK801/CNQX (5 μM/5 μM) and incubated for 20 min at 37°C. Control cultures received 300 μL of NB/N2:BSS only. Coverslips were transferred to the recording chamber of the microscope in 900 μL physiological rat saline (composition in mM: 138 NaCl, 2.7 KCl, 1.8 CaCl2, 1.06 MgCl2, 12.4 N-2-hydroxyethyl-piperazine-N′-2-ethanesulphonic acid (HEPES) and 5.6 glucose; pH 7.3) only, or containing peptide (5 μM) or MK801/CNQX (5 μM/5 μM) and incubated for a further 10 min at 21°C. A diaphragm device located on the microscope optically isolated a group of five to seven neurons for each measurement. Intracellular calcium ion levels were recorded for 30 s before the addition of 100 μL of 1 mM glutamate (100 μM final concentration), and for a further 210 s after the addition. Measurements of calcium ions were performed using an inverted epifluorescence microscope (Nikon TE2000, Tokyo, Japan) connected to a spectrophotometer (Cairn, Faversham, UK). The ratio (R) of fluorescence emission (emission wave-length: 510 nm) at 340 nm and 380 nm excitation (F340/F380), was collected at 10 Hz, stored and analyzed using the Cairn software package (Cairn). As the primary aim of these experiments was to determine the relative change in calcium ions compared with control measurements, changes in the fluorescence ratio were used as an indicator of changes in the intracellular calcium ions. These experiments were performed in duplicate with sister neuronal cultures and repeated a minimum of three times independently. Fluorescence ratio data were analyzed by anova, followed by post-hoc Fischer’s PLSD test, with < 0.05 values considered statistically significant, and presented as mean ± SEM.

Western blot for α-fodrin cleavage

For protein extraction, cultured cells were lysed in buffer [50 mM Tris–HCl, pH 7.5, 100 mM NaCl, 20 mM EDTA, 0.1% sodium dodecyl sulfate, 0.2% deoxycholic acid, containing Complete protein inhibitor (Roche, Indianapolis, IN, USA)], vortexed briefly, and clarified by centrifugation (14 000 g) at 4°C. Protein concentrations were determined by the Bradford assay (Bio-Rad, Hercules, CA, USA). Equivalent amounts of protein (5–10 μg per lane) were loaded and separated on 10% sodium dodecyl sulfate-polyacrylamide Bis–Tris minigels (Invitrogen) and transferred to a polyvinylidene difluoride membrane. Membranes were blocked in phosphate-buffered saline/0.5% Tween 20 (PBS/T) containing 5% skim milk for 1 h at 21°C before washing in PBS/T and PBS. Membranes were incubated at 4°C overnight in PBS/T containing 1 mg/mL ovalbumin and primary antibody, washed, and incubated in blocking solution containing horseradish peroxidase-conjugated secondary antibody for 1 h at 21°C. Protein bands were detected using ECL Plus (Amersham, Buckinghamshire, UK), visualized by exposure to X-ray film (Hyperfilm; Amersham), and scanned and quantified in Image J. As required, membranes were incubated for 10 min at 21°C in stripping solution (Chemicon, Temecula, CA, USA) prior to immunodetection of control proteins. Primary antibodies used were mouse monoclonal anti-α fodrin (1 : 500; MP Biomedicals, Solon, OH, USA) antibody and mouse monoclonal anti-β-tubulin (1 : 1,000; BD Pharmingen, San Jose, CA, USA). Secondary antibody was sheep anti-mouse IgG (1 : 10,000; Amersham). These experiments were repeated three times independently and data analyzed by anova, followed by post-hoc Fischer’s PLSD test, with < 0.05 values considered statistically significant, and presented as mean ± SEM.

Fluorescence and luminescent caspase assays

Cortical neurons were either exposed to 100 μM glutamate, 5 μM staurosporine (positive control), or pre-incubated for 15 min with 100 μM Z-VAD-FMK (carbobenzoxy-valyl-alanyl-aspartyl-[O-methyl]-fluoromethylketone) and then exposed to 5 μM staurosporine (negative control). Caspase 3 and 7 activation were assessed at 1, 3 and 6 h post-glutamate or staurosporine exposure using Apo-ONE homogenous caspase 3/7 assay (Promega) as per manufacturers instructions. Caspase 8 and 9 activation were assessed at the same time points using Caspase-Glo 8/9 Assays (Promega) as per manufacturers instructions. Data were analyzed by anova, followed by post-hoc Fischer’s PLSD test, with < 0.05 values considered statistically significant, and presented as mean ± SEM.

Calpain I inhibition assay

Active Calpain I (5 μg) (Sigma) was combined with the AP-1 inhibitory peptides (5, 10 or 20 μM final concentration) or calpain inhibitor Z-LLY-FMK (20 μM final concentration; Abcam, Cambridge, UK) in calpain activity reaction buffer (50 μL final volume; Abcam) and incubated at 21°C for 5 min. Calpain substrate Ac-LLY-AFC (2.5 μL : 50 μM final concentration; Abcam) was added and incubated, in the dark, for 1 h before measuring fluorescence of free AFC at 505 nm with a PolarStar fluorescence plate reader. Fluorescence data were converted to reflect proportional fluorescence relative to the positive control (100%). Data were analyzed by anova, followed by post-hoc Fischer’s PLSD test, with < 0.05 values considered statistically significant, and presented as mean ± SEM.

Results

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Morphological characteristics of neuronal cell death following glutamate exposure

Light microscopy revealed that from 15 min following glutamate exposure, neuronal cell bodies appeared rounded and swollen and their nuclei were easily visible, shrunken (fried egg appearance) and stained with propidium iodide (Fig. 1a). From 2 h following glutamate exposure, neuronal processes and cell bodies had begun to degenerate in most neurons, while other neurons appeared rounded and swollen, and stained with propidium iodide. Twenty hours after glutamate exposure few remaining neurons appeared intact and normal (1–5%). The presence of the mRNA transcription inhibitor actinomycin D (1 μM), during and post-glutamate exposure had no effect in inhibiting cell death (data not shown). The morphological characteristic of the neuronal cell death and its insensitivity to actinomycin D inhibition in our glutamate model is generally accepted as indicative of necrosis.

image

Figure 1.  (a) Time-course photomicrographs of neurons following glutamate exposure under light and fluorescence microscopy; morphological features and propidium iodide staining of dying neurons is typical of necrosis. (b) Photomicrographs from cultures 18 h after exposure to glutamate and treated with AP-1 inhibitory peptides (5 μM), D-TAT peptide (5 μM) or glutamate receptor blockers. Control neurons that have not been exposed to glutamate are also shown (representative photomicrographs of = 3).

Download figure to PowerPoint

Initial screen of l-isoform AP-1 inhibitory peptides

The initial screen of the 19 l-isoform peptides (5 μM) added to neuronal cultures 15 min prior to glutamate exposure resulted in five of the peptides (PYC19L-TAT, PYC35L-TAT, PYC36L-TAT, PYC38L-TAT, PYC41L-TAT) displaying statistically significant neuroprotective activity (Fig. 2a and Table 2). The increase in neuronal viability afforded by the five neuroprotective peptides ranged from 32–83% (< 0.0001 for all five AP-1 inhibitory peptides). The remaining peptides had either no effect (PYC8L-TAT, PYC20L-TAT, PYC32L-TAT, PYC34L-TAT, PYC45L-TAT, PYC58L-TAT, PYC60L-TAT, PYC67L-TAT, PYC71L-TAT) (Fig. 1 for PYC71L-TAT, or data not shown) or appeared to increase neuronal death (PYC26L-TAT, PYC22L-TAT, PYC24L-TAT, PYC54L-TAT, PYC66L-TAT) (data not shown). The positive control peptide JNKI-1D-TAT (5 μM) increased neuronal viability to 69% (< 0.0001). The control peptide D-TAT conferred significant neuroprotection at 5 μM with 20% neuronal survival (= 0.001; Fig. 3). However, at the 5 μM concentration the scrambled control peptides PYC35DScram-TAT and PYC36DScram-TAT did not provide significant neuroprotection (Fig. 2b).

image

Figure 2.  The comparative neuroprotective activity of AP-1 inhibitory peptides, scrambled controls and over an administration timecourse. (a) The comparative neuroprotective activity of L-isoform AP-1 inhibitory peptides (5 μM) 24 h following glutamate exposure. (b) Neuroprotective activity of scrambled peptides PYC35DScram-TAT and PYC36DScram-TAT (5 μM) 24 h following glutamate exposure. (c) Efficacy of PYC35D-TAT, PYC36D-TAT and JNKI-1 D-TAT peptides (5 μM) when administered 15 min prior to, immediately post (0 min) or at 15 min post-glutamate exposure. Neuronal viability with the administration of glutamate receptor blockers (Blckrs) is also shown. MTS data were expressed as percentage neuronal viability with no insult control taken as 100% viability (mean ± SEM; = 4; *< 0.05, **< 0.0001).

Download figure to PowerPoint

Table 2.   Percentage neuronal viability and IC50 values of AP-1 inhibitory, JNKI-1D-TAT and D-TAT peptides in glutamate model
PeptideIC50 (μM)Viability: 5 μM (%)Viability: 10 μM (%)
PYC19L-TAT> 103243
PYC19D-TAT2.76875
PYC35L-TAT> 103842
PYC35D-TAT8.54858
PYC36L-TAT1.56457
PYC36D-TAT1.38276
PYC38L-TAT3.26983
PYC38D-TAT2.88361
PYC41L-TAT> 103845
PYC41D-TAT> 104040
JNKI-1D-TAT2.17589
D-TAT> 102025
image

Figure 3.  Concentration response graphs for neuroprotective AP-1 inhibitory peptides (PYC19-TAT, PYC35-TAT, PYC36-TAT, PYC38-TAT, PYC41-TAT), in both l- and d-isoforms, D-TAT, and JNKI-1D-TAT, 24 h following glutamate exposure. Note: 2 μM concentration bars are shaded as a reference point. MTS data were expressed as percentage neuronal viability with no insult control taken as 100% viability (mean ± SEM; = 4; *< 0.05; **< 0.0001).

Download figure to PowerPoint

At 1 h post-glutamate exposure, light microscopic examination of neuronal cultures revealed that neurons in untreated or control peptide-treated cultures (5 μM: D-TAT, PYC35D-TATScram, PYC36D-TATScram, PYC35D, PYC36D) displayed cellular rounding (data not shown), and, by 18 h post-exposure, few neurons appeared intact and viable (Fig. 1b). In contrast, most neurons in cultures treated with neuroprotective AP-1 inhibitory peptides (PYC19D-TAT, PYC35D-TAT, PYC36D-TAT, PYC38D-TAT, PYC41D-TAT), JNKI-1D-TAT, or glutamate receptor blockers, appeared intact and viable, and survived glutamate exposure (Fig. 1b).

Dose concentration experiments using l- and d-isoform neuroprotective AP-1 inhibitory peptides

The five l- and d-isoform AP-1 inhibitory peptides increased neuronal survival in a concentration-dependent fashion, with respective IC50 values ranging from 1.5 –≥ 10 μM and 1.3–8 μM (Fig. 3 and Table 2) and maximal cell viability levels ranging from 32 to 83% and 40–83%. d-isoforms peptides were more potent than l-isoforms, except for PYC41-TAT, which had comparable levels of neuroprotection in both isoforms (Fig. 3 and Table 2). At the 2, 5 and 10 μM concentrations all five AP-1 inhibitory peptides provided significant neuroprotection, with diverse neuroprotection at the lower 0.1 or 0.5 μM concentrations. The JNKI-1D-TAT peptide achieved a peak viability of 89% at 10 μM and generated an IC50 of 2.1 μM (Fig. 3 and Table 2). As mentioned above, D-TAT increased neuronal survival to 20% at 5 μM (Fig. 3 and Table 2).

Based on the results using the d-isoform peptides showing increased potency it was hypothesized that l-isoform peptides that did not display neuroprotective activity may do so in the d-isoform. To test this hypothesis, three peptides (PYC16-TAT, PYC60-TAT, PYC71-TAT) were randomly selected, synthesized in the d-isoform and assessed in the glutamate model at 5 μM. None of the three peptides in the d-isoform provided neuroprotection (data not shown).

Peptides (PYC35D-TAT, PYC36D-TAT) are only neuroprotective when present prior to glutamate exposure

An administration time-course of peptides PYC35D-TAT, PYC36D-TAT, and JNKI-1D-TAT (5 μM) when administered prior to glutamate exposure or after glutamate exposure (0, 15, 30, 45 or 60 min) showed only neuroprotective efficacy of peptides when administered prior to glutamate exposure and not at any timepoint after glutamate exposure (Fig. 2c). In contrast, glutamate receptor blockers were effective when added prior to, or immediately after glutamate exposure, and to a lesser extent when administered 15 min after glutamate exposure (21% viability; = 0.041; Fig. 2c).

AP-1 inhibitory peptides down-regulate AP-1 driven luciferase reporter expression

All five AP-1 inhibitory peptides (PYC19D-TAT, PYC35D-TAT, PYC36D-TAT, PYC38D-TAT, PYC41D-TAT) significantly reduced constitutive AP-1 complex-mediated luciferase transcription in HEK293 AP-1 Luc cells, compared with the control cells (mean reductions: 27%, 32%, 33%, 38%, 27%; = 0.024, 0.009 0.005, 0.003, 0.040, respectively) (Fig. 4a). It should be noted that there was no significant difference in luminescence between cells that had not been administered with peptide and cells that had been administered with the control D-TAT peptide. The positive control peptide, c-JunZIPL-TAT, reduced luminescence by 33% (= 0.004). To ensure that luminescence data was not influenced by peptide toxicity, the MTS assay was performed and showed no difference in cell viability between cells transduced with the different peptides (data not shown).

image

Figure 4.  Effect of AP-1 inhibitory peptides on transcription with (a) AP-1 down-regulation by AP-1 inhibitory peptides (5 μM) in luciferase assay. Relative light unit readings were normalized as a percentage against the HEK293 AP-1 Luc cells as a negative control (100% luminescence). (b) Real-time-PCR for c-Jun and c-Fos mRNA levels in cortical neuronal cultures treated with AP-1 inhibitory peptides or JNKI-1D-TAT (5 μM) and exposed to glutamate. Total RNA was extracted at 60 min post-glutamate exposure. mRNA levels for c-Jun and c-Fos were normalized against GAPDH mRNA and compared with glutamate-treated controls for statistical analysis (mean ± SEM; = 3; *< 0.05).

Download figure to PowerPoint

AP-1 inhibitory peptides down-regulated c-Jun and c-Fos mRNA transcription

One hour following glutamate exposure, cortical neuronal cultures displayed a 100% increase in c-Jun and c-Fos mRNA, which was significantly reduced by all five AP-1 inhibitory peptides and the JNKI-1D-TAT inhibitory peptide (Fig. 4b). Peptides PYC19D-TAT, PYC35D-TAT, PYC36D-TAT, PYC38D-TAT, PYC41D-TAT and JNKI-1D-TAT reduced c-Jun/c-Fos mRNA levels respectively by 28/43%, 59/56%, 56/63%, 34/19%, 29/37% and 35/28%. There was no significant reduction in c-Jun or c-Fos mRNA levels in cells that had been administered with the D-TAT peptide and exposed to glutamate compared with control neuronal cultures (data not shown).

AP-1 inhibitory peptides (PYC35D-TAT, PYC36D-TAT) do not inhibit calcium entry into neurons

Neuronal cultures exposed to glutamate with or without PYC35D-TAT or PYC36D-TAT peptide treatment underwent a rapid increase in intracellular calcium (untreated mean amplitude: 496.0; mean amplitudes: PYC35D-TAT, 409.3; PYC36D-TAT, 423.5; Fig. 5a, c, and d). In contrast, neuronal cultures treated with glutamate blockers (MK801/CNQX) exhibited only a minor increase in intracellular calcium after glutamate exposure (mean amplitude: 41.5) (Fig. 5b and d).

image

Figure 5.  A comparison of calcium influx in cortical neurons exposed to glutamate (30 s time point; arrow) and treated with: (a) no treatment, (b) glutamate receptor blockers, and (c) PYC36D-TAT (5 μM). (d) Measurement of amplitude between basal calcium ion level and peak calcium ion level for untreated and treated cultures (mean ± SEM; = 8; *< 0.05).

Download figure to PowerPoint

Neuronal death following glutamate excitotoxicity involves calpain (but not caspase) activation, which can be down-regulated by AP-1 inhibitory peptides

The cleavage of the cytoskeletal protein α-fodrin to produce 145 and 150 kDa protein fragments at 2, 3 and 6 h after glutamate exposure is indicative of cleavage by activated calpain (Fig. 6a). There were no detectable 120 kDa protein fragments at any time point, indicating that cleavage of α-fodrin by caspase 3 did not occur. Similarly, caspase 9, 8 or 7 assays at 1, 3 and 6 h post-glutamate exposure did not reveal any significant caspase activation (Fig. 6b, c and d). The addition of AP-1 inhibitory peptides (5 μM final concentration: PYC19D-TAT, PYC35D-TAT, PYC36D-TAT, PYC38D-TAT, PYC41D-TAT, JNKI-1D-TAT) and the JNK inhibitory peptide (5 μM) to neuronal cultures inhibited induced calpain driven α-fodrin cleavage (43%, 33%, 8%, 13%, 45%, 12%, respectively) following glutamate exposure, which was statistically significant when compared with the glutamate control cultures, but not significantly different from the no insult control cultures (Fig. 6e).

image

Figure 6.  Assessment of α-fodrin cleavage in cortical neuronal cultures a) over a time-course from 15 min to 6 h following glutamate exposure. Calpain cleavage of α-fodrin results in bands at 145 and 150 kDa, caspase cleavage of α-fodrin results in bands at 120 kDa (none evident and not shown) and 145 kDa. The lack of caspase activation in cortical neuronal cultures 1, 3, and 6 h following glutamate exposure was confirmed with fluorescence assays. Caspase assays show a lack of (b) caspase 3/7, (c) caspase 8 and d) caspase 9 activation. Assessment of α-fodrin cleavage in cortical neuronal cultures 2 h following glutamate exposure. (e) Fold increase of 145/150 kDa band intensity, normalized against β-tubulin, and compared against glutamate-treated controls is shown graphically. (f) Calpain I assay showing a lack of direct inhibition of calpain by AP-1 inhibitory peptides or JNKI-1D-TAT (5 μM), compared with positive control of active calpain I (5 μg). Inhibitor control is calpain inhibitor Z-LLY-FMK (20 μM). (Mean ± SEM; = 4; *< 0.05; **< 0.0001).

Download figure to PowerPoint

AP-1 inhibitory peptides do not directly inhibit calpain activation

Using a fluorometric calpain I assay, the AP-1 inhibitory peptides did not significantly inhibit calpain I activation when assessed at 5, 10 and 20 μM (5 μM; Fig. 6f). This was in contrast to the calpain inhibitor peptide (Z-LLY-FMK; 20 μM), which abolished nearly all calpain activation. Interestingly, the PYC35D-TAT peptide appeared to increase calpain I activation. The increased fluorescent signal obtained with the PYC35D-TAT peptides was not seen when the calpain substrate (Ac-LLY-AFC) was omitted from the assay (data not shown).

Discussion

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

Following the screening of 19 AP-1 inhibitory peptides in a necrotic glutamate excitotoxicity model, we identified five peptides with neuroprotective activity. The inability of all 19 AP-1 inhibitory peptides to provide neuroprotection may be due to differences in peptide half-life, target binding affinity, and/or cellular uptake. The importance of peptide stability is supported by the findings of increased potency of d-isoform versus l-isoform peptides. With respect to AP-1 inhibitory peptide target binding efficacy, this is governed by the peptides: (i) differential binding affinity to c-Jun homodimers versus other c-Jun heterodimer moieties, such c-Fos and ATF (Smeal et al. 1989) and; (ii) differential c-Jun homodimer formation inhibition and disruption ability. Finally, as the transduction efficacy of cell penetrating peptides has been shown to vary in relation to the cargo, it is possible some of the AP-1 inhibitory peptides may have disrupted TAT’s transduction ability (Fischer et al. 2002; Peitz et al. 2002; Mueller et al. 2008).

The ability of the five neuroprotective peptides to inhibit AP-1 activation was confirmed using an AP-1 luciferase expression assay. Additionally, the AP-1 inhibitory peptides were shown to suppress c-Jun and c-Fos mRNA expression in cortical neuronal cultures following glutamate exposure; as transcription of c-Jun is directly up-regulated by AP-1 activation (Angel et al. 1988), inhibition of c-Jun will suppress c-Jun mRNA expression. Our findings are in line with previous studies demonstrating decreases in c-Jun and c-Fos mRNA are positively correlated with neuronal survival in glutamate excitotoxicity and glucose/oxygen deprivation models (Dong et al. 2005; Fernandez et al. 2005).

With the exception of PYC19D-TAT, the remaining four AP-1 inhibitory peptides and JNK inhibitory peptide displayed a strong linear correlation relation between peak levels of neuronal death inhibition/IC50 values, and HEK AP-1 driven luciferase inhibition, and following glutamate exposure, inhibition of c-Jun/c-Fos mRNA expression and calpain α-fodrin cleavage. The lack of correlation with the PYC19D-TAT peptide may indicate that this peptide may have a different neuroprotective mechanistic action, which may simply involve enhancing the neuroprotective capability of TAT. Also of interest, was the inability of three of the AP-1 inhibitory peptides and the JNK inhibitory peptide to show significant protection when used at lower concentrations (< 0.5 μM), which could be attributable to these peptides altering the concentration of TAT required to achieve transduction (Binder and Lindblom 2003; Hallbrink et al. 2004). To this end, we predict that any improvements in transduction efficiency are likely to increase peptide potency.

While it is beyond the scope of this study to identify all cell death pathways activated following glutamate exposure, we have shown that in our model, the AP1 complex is elicited (c-Jun/c-Fos transcription), intracellular calcium influx is induced, calpain is activated and caspases are not, cell death is insensitive to transcriptional inhibition and that cell death morphology is indicative of necrosis. This raises the question of how the AP-1 inhibitory peptides are working in this glutamate model. There are three main possibilities: (i) activation of AP-1 has transcriptional independent cell death functions; (ii) peptides inhibit other targets involved in necrotic cell death and; (iii) AP-1 inhibitory peptides potentiate TAT’s neuroprotective mechanism. At present, we favor the first possibility.

The AP-1 inhibitory peptides maintained their neuroprotective efficacy in the presence of the transcriptional inhibitor actinomycin D following glutamate excitotoxicity, similar to that of TAT-TIJIP (TAT-fused Truncated Inhibitor of JNK Interacting Protein; a 9 amino acid truncated version of JNKI-1) (data not shown & Arthur et al. 2007). Thus, while the AP-1 inhibitory peptides and JNK inhibitors down-regulate transcription of c-Jun and c-Fos, the effect of AP-1 activation may not be limited to inducing expression of pro-apoptotic genes, but may also have post-translational effects elsewhere in the cell, for example, BAX mitochondrial translocation, or calpain activation. Hence the down-regulation of α-fodrin calpain cleavage by the neuroprotective peptides may lie in the antagonism of AP-1 post-translational effects on calpain activation directly or indirectly, although this is speculative. However, this is similar to the transcription factor p53 which, in addition to inducing pro-apoptotic gene expression, can directly antagonize proteins necessary for cell survival (e.g. replication protein A) (Caelles et al. 1994; Miyashita and Reed 1995).

While possible, it is unlikely that five neuroprotective peptides with diverse amino acids sequences would each inhibit an additional target involved in necrosis, except of course if the major neuroprotective mechanism is via the TAT peptide (Wei et al. 2008; Vaslin et al. 2009). While the neuroprotection provided by D-TAT was significant it did not exceed that afforded by the five neuroprotective AP-1 inhibitory peptides or JNKI-1D-TAT peptide. Consequently, although we cannot rule out that the TAT-fused AP-1 or JNK inhibitory peptides, while inhibiting their specific targets (c-Jun/JNK), are not merely potentiating TAT’s neuroprotective mechanism, the low neuroprotective potency of TAT at high concentrations would argue against this. While the mechanism of TAT’s neuroprotective action has not been elucidated and requires investigation, at present it could only be considered a beneficial property with respect to its use in neuroprotective drug-delivery.

The use of peptide and chemical JNK inhibitors have and are being investigated as potential treatments to inhibit neuronal cell death associated with glutamate excitotoxicity in vitro and following cerebral ischemia in animals and humans (Harding et al. 2001; Borsello et al. 2003). However, besides phosphorylating c-Jun, JNK phosphorylates multiple additional proteins including ATF1, NFAT4 and Elk-1 which have roles other than apoptosis (Chow et al. 1997). In addition, it also appears that c-Jun, which can be phosphorylated by other kinases besides JNK, has a major involvement in both apoptotic (Behrens et al. 1999; Fernandez et al. 2005) and necrotic neuronal cell death (current study; Arthur et al. 2007). Therefore, inhibiting activation of AP-1 proteins represents a more specific target for preventing cell death. In addition, it appears that at least one of our AP-1 inhibitory peptides (PYC36D-TAT) is a more potent neuroprotective agent than the JNK inhibitory peptide (JNKI-1D-TAT), which is currently in a Phase 1 clinical stroke trial (Wiegler et al. 2008). Finally, while there was no neuroprotection afforded by post-insult administration of either AP-1 inhibitory peptides or the JNKI-1D-TAT peptide in our glutamate model, evidence of in vivo neuroprotection for the JNKI-1D-TAT following cerebral ischemia leads us to believe that the AP-1 inhibitory peptides will also follow this trend (Borsello et al. 2003; Esneault et al. 2008).

Acknowledgments

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References

The authors would like to thank Dr Nadia Milech, Ms Kate Thomas, Ms Joanne Chieng and Mr Tom Hamilton for their technical support and Dr Kym Campbell for expert editing. This work was supported by a Biotechnology Innovation Fund Grant from AusIndustry, Neurotrauma Research Project Grant, The University of Western Australia and Phylogica Ltd. Phylogica was not involved in either the experimental design or the analysis of data arising from this study.

References

  1. Top of page
  2. Abstract
  3. Materials and methods
  4. Results
  5. Discussion
  6. Acknowledgments
  7. References