Mesenchymal stem cell-secreted superoxide dismutase promotes cerebellar neuronal survival

Authors

  • Kevin Kemp,

    1. Multiple Sclerosis and Stem Cell Group, Institute of Clinical Neurosciences, Clinical Sciences North Bristol, University of Bristol, Bristol, UK
    Search for more papers by this author
  • Kelly Hares,

    1. Multiple Sclerosis and Stem Cell Group, Institute of Clinical Neurosciences, Clinical Sciences North Bristol, University of Bristol, Bristol, UK
    Search for more papers by this author
  • Elizabeth Mallam,

    1. Multiple Sclerosis and Stem Cell Group, Institute of Clinical Neurosciences, Clinical Sciences North Bristol, University of Bristol, Bristol, UK
    Search for more papers by this author
  • Kate J. Heesom,

    1. Proteomics Facility, Department of Biochemistry, School of Medical Sciences, University of Bristol, Bristol, UK
    Search for more papers by this author
  • Neil Scolding,

    1. Multiple Sclerosis and Stem Cell Group, Institute of Clinical Neurosciences, Clinical Sciences North Bristol, University of Bristol, Bristol, UK
    Search for more papers by this author
  • Alastair Wilkins

    1. Multiple Sclerosis and Stem Cell Group, Institute of Clinical Neurosciences, Clinical Sciences North Bristol, University of Bristol, Bristol, UK
    Search for more papers by this author

Address correspondence and reprint requests to Kevin Kemp, MS labs, 1st floor, Burden Centre, Frenchay Hospital, Bristol BS16 1JB, UK. E-mail: kevin.kemp@bristol.ac.uk

Abstract

J. Neurochem. (2010) 114, 1569–1580.

Abstract

It has been postulated that bone marrow-derived mesenchymal stem cells (MSCs) might be effective treatments for neurodegenerative disorders either by replacement of lost cells by differentiation into functional neural tissue; modulation of the immune system to prevent further neurodegeneration; and/or provision of trophic support for the diseased nervous system. Here we have performed a series of experiments showing that human bone marrow-derived MSCs are able to protect cultured rodent cerebellar neurons, and specifically cells expressing Purkinje cell markers, against either nitric oxide exposure or withdrawal of trophic support via cell-cell contact and/or secretion of soluble factors, or through secretion of soluble factors alone. We have demonstrated that MSCs protect cerebellar neurons against toxic insults via modulation of both the phosphatidylinositol 3-kinase/Akt and MAPK pathways and defined superoxide dismutase 3 as a secreted active antioxidant biomolecule by which MSCs modulate, at least in part, their neuroprotective effect on cerebellar cells in vitro. Together, the results demonstrate new and specific mechanisms by which MSCs promote cerebellar neuronal survival and add further evidence to the concept that MSCs may be potential therapeutic agents for neurological disorders involving the cerebellum.

Abbreviations used
DETCA

diethyldithiocarbamic acid

DMEM

Dulbecco’s Modified Eagles Medium

JNK

c-Jun N-terminal kinase

MNC

mononuclear cell

MSC

mesenchymal stem cell

NH

non hematopoietic

NO

nitric oxide

PAGE

polyacrylamide gel electrophoresis

PI3K

phosphatidylinositol 3-kinase

ROS

reactive oxygen species

SDS

sodium dodecyl sulfate

SOD

superoxide dismutase

With improved characterization of mesenchymal stem cells (MSCs), techniques are now available to isolate human MSCs and manipulate their expansion in vitro under defined culture conditions without loss of phenotype or function (Devine and Hoffman 2000). MSCs have therefore generated a great deal of interest in many clinical settings including that of regenerative medicine, immune modulation and tissue engineering. Specifically stem cell therapies hold therapeutic promise in a variety of neurodegenerative diseases, including degenerative ataxias, which are characterised by degeneration of cerebellar neuronal populations. Human MSC transplantation has been shown to improve outcome in a variety of animal models of neurological disease including experimental autoimmune encephalomyelitis, stroke and spinal cord injury (Zhang et al. 2004, 2006; Himes et al. 2006). Although these reports are encouraging, precise mechanisms of action or ways in which the therapeutic potential can be enhanced are still required.

It has been postulated that MSCs might be effective treatments in neurodegenerative disorders either by replacement of lost cells by differentiation into functional neural tissue; modulation of the immune system to prevent further neurodegeneration; and/or provision of trophic support for the diseased nervous system (Parr et al. 2007). Bone marrow-derived stem cells are able to evade the allogeneic immune system, as well as suppressing immune responses directed against third-party cells following intravenous infusion (Bartholomew et al. 2002; Le Blanc et al. 2003; Maitra et al. 2004; Beyth et al. 2005). In addition, when infused into the circulation, MSCs have the capacity to migrate specifically to sites of brain injury, thus targeting sites for neural repair (Chopp and Li 2002; Mahmood et al. 2003, 2005). MSCs have been shown to generate cells with a neuronal morphology and phenotype in vitro (Woodbury et al. 2000; Black and Woodbury 2001), however recent evidence seems to indicate this is not a major mechanism promoting neuronal recovery and the neuroprotective role of bone marrow-derived stem cells is more likely through their ability to elaborate various potentially neuroprotective factors in vitro including brain-derived neurotrophic factor or more recently through antioxidant and neuroprotective actions (Arnhold et al. 2006; Crigler et al. 2006; Parr et al. 2007; Hokari et al. 2008; Lanza et al. 2009; Wilkins et al. 2009).

If human stem cell trials for ataxic conditions are to be performed, a precise understanding of their effects on cerebellar neurons and other cells in vitro and in vivo is required. Here we have performed a series of experiments showing that human bone marrow-derived stem cells are able to protect cerebellar neurons, and specifically cells expressing a Purkinje cell marker, in vitro, via cell-cell contact and/or secretion of soluble factors, or through secretion of soluble factors alone. We have also defined a specific active antioxidant biomolecule secreted by MSCs and intracellular mechanisms underlying these neuroprotective effects. These findings suggest that bone marrow-derived stem cells are potential therapeutic agents for cerebellar ataxia and other ataxic conditions and, in time, transplantation of bone marrow-derived stem cells may offer an effective therapy for neurological injury.

Materials and methods

Establishment of mesenchymal cultures

Bone marrow samples were obtained by an Orthopaedic surgeon at Southmead Hospital, Bristol with informed written consent and hospital ethic committee approval. Bone marrow was taken at the time of total hip replacement surgery from the femoral shaft and placed into a sterile 50 mL tubes containing 1000 IU heparin. Patients with a history of malignancy, immune disorders or rheumatoid arthritis were excluded from the study. Femoral shaft bone marrow donors were healthy apart from osteoarthritis, and were not receiving drugs known to be associated with myelosuppression or bone-marrow failure.

Femoral shaft marrow samples were broken up with a scalpel and washed with Dulbecco’s Modified Eagles Medium (DMEM) (Sigma-Aldrich, Gillingham, UK) until remaining material (bone) looked white at the bottom of the 50 mL tube. All washings were pipetted into a new 50 mL tube and kept for centrifugation. The suspension was centrifuged and re-suspended in DMEM and overlaid onto an equal volume of Lymphoprep™ (Axis-Shield, Dundee, UK; density 1.077 ± 0.001 g/mL) and centrifuged at 600 g for 35 min at 24°C to separate the mononuclear cells (MNCs) from neutrophils and red cells. The MNC layer was harvested and washed twice in DMEM.

MSC culture

Isolated MNCs were centrifuged and re-suspended in MSC medium consisting of DMEM with 10% foetal calf serum which selected for the growth of MSCs (StemCell Technologies, London, UK), and 1% Penicillin and Streptomycin (Sigma-Aldrich). Vented flasks (25 cm2) containing 10 mL of MSC medium were seeded with 1 × 107 cells for primary culture. Flasks were incubated at 37°C in a humidified atmosphere containing 5% CO2 and fed every week with MSC medium by half medium exchange to remove non-adherent hematopoietic cells until the adherent fibroblast-like MSCs reached approximately 70% confluence.

On reaching confluence the adherent cells were re-suspended using 0.25% trypsin (Sigma-Aldrich) and re-seeded at 2.25 × 105 cells per (75 cm2) flask into first passage. Cultures were then incubated, fed every week with MSC medium by half medium exchange, and again trypsinized, a cell count taken and re-seeded at 2.25 × 105 cells per flask (75 cm2).

MSC characterization

Cells harvested from femoral shaft marrows displayed all the typical characteristics of MSC in culture. To ensure a homogenous population of MSC had been cultured immunophenotyping of surface markers was carried out using flow cytometry according to previous reports (Pittenger et al. 1999). Cells were examined at third passage using APC-conjugated anti-CD105, APC-Cy7-conjugated anti-CD45 (eBioscience, San Diego, CA, USA), PE-conjugated anti-CD166, FITC-conjugated anti-CD90 (BD Biosciences, Oxford, UK), and with PE-Cy7-conjugated anti-CD44 (Serotec, Oxford, UK). Mesenchymal stem cells were induced into adipogenic, osteoblastic and chondrogenic differentiation by culturing MSC, at third passage, in non hematopoietic (NH) Adipodiff medium, NH Osteodiff medium and NH Chondrodiff medium (Miltenyi Biotec, Woking, UK) respectively according to the manufacturers instructions. Adipogenic, osteogenic and chondrogenic differentiation were characterized using immunofluorescent detection by indirect labelling using a anti-lipoprotein lipase (Abcam, Cambridge, UK), anti-Alkaline phosphatase (Abcam) and anti-human aggrecan (4F4) (Santa Cruz Biotechnology, Santa Cruz, CA, USA) respectively.

Neuronal cell culture

Neuronal cultures were prepared from the cerebella of E18 rat embryos. In brief, the pregnant female (time mated), was killed and embryos removed. The cerebella were dissected out and the meninges removed. Following enzymatic and mechanical dissociation, cells were counted and plated onto poly-l-lysine coated 13 mm coverslips at 250 000cells/coverslip and cultured in Dulbecco’s modified eagles medium supplemented with 2% B27 (Gibco, Paisley, UK) and 1% Penicillin and Streptomycin (B27 medium). Five days after plating neurons (identified by β-tubulin III staining) represented 97.9 ± 1.2% (= 3) of the total cells, and within this population 20.1 ± 2.3% (= 3) were Purkinje cells [identified by Calbindin D28K staining, a calcium binding protein that is expressed by Purkinje cells in the cerebellum (Nordquist et al. 1988)]. The remaining cells were predominantly glial fibrillary acidic protein expressing astrocytes and GalC expressing oligodendrocytes.

At 5 days in vitro culture, cerebellar neuronal cells were exposed to experimental conditions. The base medium for all experiments was ‘minimal’, which consisted of Dulbecco’s modified eagles medium supplemented with insulin-free Sato (containing 100 μg/mL bovine serum albumin, 100 μg/mL transferrin, 0.06 μg/mL progesterone, 16 μg/mL putrescine, 0.04 μg/mL selenite, 0.04 μg/mL thyroxine, 0.04 μg/mL triiodothryonine).

The pathway inhibitors PD98059 (30 μM) and SP600125 c-Jun N-terminal kinase (JNK) inhibitor II (20 μM) were obtained from Calbiochem (San Diego, CA, USA), and LY294002 (10 μM) and the p38 inhibitor SB 203580 (20 μM) were from Sigma-Aldrich. Superoxide dismutase (SOD) was derived from human erythrocytes and used at 250 U/mL (Sigma-Aldrich). The Cu/Zn SOD inhibitor diethyldithiocarbamic acid (DETCA) was used at a concentration of (3 μM) (Sigma-Aldrich). A stock solution (50 mM in 10 mM NaOH) of (Z)-1-[2-(aminoethyl)-N-(2-ammonioethyl)amino]daizen-1-ium-1,2-diolate (DETANONOate; Alexis Biochemicals, Nottingham, UK) was prepared immediately before use.

Preparation of mesenchymal stem cell conditioned medium

Confluent mesenchymal stem cell cultures, at third passage, were washed in DMEM and cultured for 24 h in minimal base medium (as described above in the neuronal cell culture section). Media was then removed and stored at −20°C prior to being used within culture experiments.

Preparation of mesenchymal stem cell transwell cultures

Confluent mesenchymal stem cell cultures, at third passage, were washed in DMEM, trypsinised and cultured in transwell inserts (Millipore, Watford, UK) at 30 000 cells per well for 72 h in MSC medium. Prior to use in the cell survival assay transwells containing MSCs were washed several times in minimal base medium.

Assays for cell survival

At 5 days in vitro culture, cerebellar neuronal cells were exposed to experimental conditions. Media was removed from all wells and cells were washed twice in DMEM. 1 mL of B27 medium, minimal media or MSC-conditioned medium was added to appropriate wells. Bone marrow-derived stem cells were added in direct co-culture (third passage MSCs, 3000 cells/well) or in transwells (30 000 cells/well) in minimal media. Specific pathway inhibitors or SOD, at the concentrations described above were then added. The effect of withdrawal of trophic factors on cerebellar neuronal cultures was achieved through culturing in serum free minimal media for 72 h. The effects of nitric oxide on cerebellar neuronal cultures was achieved through exposure to DETANONOate (0.1 mM or 0.2 mM) for 24 h. All cell survival experiments concerning the neuroprotective effects of MSC-conditioned media, and bone marrow-derived stem cells either in direct co-culture or in transwells, with or without specific pathway inhibitors were carried out within the same plates.

Evaluation of cerebellar neuronal cell survival was carried out using immunocytochemistry and morphological examination of neuronal cultures using the following methods.

Immunocytochemistry

Immunocytochemisty was used to identify cell phenotypes and allowed examination of cellular morphology and SOD3 expression. Neuronal cultures were stained after fixation with 4% paraformalaldehyde. Primary antibodies against intracellular markers were used after treatment with 100% methanol at −20°C for 10 min. These were β-tubulin III (1 : 500) (Promega, Southampton, UK), Calbindin D28K (1 : 500) (Sigma-Aldrich) and mouse anti-SOD3 (1 : 250) (Abcam). Species specific (1 : 500) Alexa Fluor® 488 and 555 conjugated secondary antibodies (Invitrogen, Paisley, UK) were used to visualize primary antibody staining. DAPI Vectasheid™ was used for nuclear identification.

Cell survival assay

Neurons were identified by β-tubulin III expression, and Calbindin D28K used for Purkinje cell identification. In addition, nuclear staining of cells enabled a morphological assessment of apoptosis. Neuronal and Purkinje cell survival was assessed using counts of β-tubulin III expression and Calbindin D28K stained cells respectively, taking five random fields per culture and at least three cultures per treatment. In all cases, control cultures grown throughout the experimental period in DMEM/2% B27 were analyzed and values for the experimental conditions divided by this value, in order to standardize results between experiments.

Immunoblotting for cell signalling proteins

Neurons were cultured at high density (2 × 106 cells/well) in a 6-well plate for 5 days before exposure to test conditions. At set time points cells were lysed using Beadlyte cell signalling universal lysis buffer (Millipore, Watford, UK). The Total protein Kit™ (Sigma-Aldrich) was then used to quantify the concentration of total protein within each cell lysate sample according to manufacturers’ instructions to ensure equal loading of cell lysates. Lysates were heated to 95°C for 5 min with Laemmli 2× sample buffer (Invitrogen) and run on Tris–HCl 10–20% ready gels (Bio-Rad, UK). After transfer to nitrocellulose membrane (Bio-Rad) and blocking in 5% w/v powdered milk, membranes were incubated overnight in primary antibody at 4°C (in Tris-buffered saline/5% bovine serum albumin). Antibodies used were Phospho-Akt (Ser473) and Akt (pan) (C67E7; Cell Signaling Technology, Beverly, MA, USA), Phospho-p38 MAPK (Thr180/Tyr182) and p38 MAPK (Cell Signalling Technology, Beverly, MA, USA), and Phospho-stress-activated protein kinase (SAPK)/JNK (Thr183/Tyr185) and SAPK/JNK (Cell Signalling Technology). Immunoblotting for SOD3 in concentrated MSC conditioned medium was carried out using a mouse anti-SOD3 (1 : 5000; Abcam).

Immunoreactivity was detected using secondary anti-rabbit horseradish peroxidase conjugated antibodies (Abcam) (in Tris-buffered saline/5% bovine serum albumin) and specific protein expression patterns were visualized by chemiluminescence using an Amersham ECL Plus™ Western Blotting Detection System (GE Healthcare, Chalfont St Giles, UK).

Two-dimensional polyacrylamide gel electrophoresis of MSC conditioned medium

Proteins within the MSC conditioned medium were concentrated using an Amicon Ultra-15, 10 kDa centrifugal filter unit (Millipore), and albumin removed using ProteoExtract® Albumin Removal Kit Maxi (Calbiochem) according to the manufacturer’s instructions.

Protein clean-up to remove contaminants carried over from sample preparation was achieved using trichloroacetic acid-acetone precipitation and the protein pellet was re-suspended into 450 μL of rehydration solution (7 M urea, 2 M thiourea, 4% CHAPS, 0.5% IPG buffer pH3-11NL, 1% dithiothreitol, 0.002% bromophenol blue, and 1% ASB-14). Resuspended samples were loaded onto 24 cm immobiline DryStrip first dimension IPG strips pH3-11 nonlinear (GE Healthcare) by passive rehydration. Following rehydration overnight, isoelectric focussing was performed using the Ettan IPGphor 3 (GE Healthcare) according to the manufacturers instructions (in brief, 500 V for 1 h, 1000 V for 1 h and 8000 V for 10.5 h). The focussed IPG strips were then incubated for 15 min in sodium dodecyl sulfate (SDS) equilibration buffer [50 mM Tris–HCl, pH8.8, 6 M urea, 30% (v/v) glycerol, 2% SDS, 0.002% bromophenol blue] containing 1% (w/v) dithiothreitol and for a further 15 min in SDS equilibration buffer containing 2.5% (w/v) iodoacetamide. The equilibrated strips were applied to the surface of vertical 12.5% SDS–polyacrylamide gels (PAGE) and proteins separated in the second dimension using the Ettan DALT 6 separation unit (GE Healthcare) at 5 mA/gel for 1 h, 8 mA/gel for 1 h and then 20 W/gel until completion.

Following SDS–PAGE, gels were fixed for 1 h in a solution of 7% glacial acetic acid, 50% methanol (v/v) and then incubated with Sypro Ruby Protein gel stain (Invitrogen) for 12 h on an orbital shaker. Gels were then washed with 7% acetic acid, 10% (v/v) methanol for 30 min and visualised at an excitation level of 450 nm and an emission level of 610 nm using a Typhoon 9400 Variable Mode Imager (GE Healthcare). Protein spots of interest were excised from the gel using the Investigator ProPic automated spot picker and digested with trypsin using the ProGest automated digestion unit (both from Perkin Elmer Life Sciences, Waltham, MA, USA). The resulting peptides were analysed using a 4700 MALDI-Tof/Tof mass spectrometer (Applied Biosystems, Warrington, UK) to give a peptide mass fingerprint and peptide sequence information, which was searched against the Mass Spec Database using the Mascot search engine from Matrix Science to identify the protein.

SOD ELISA

Mesenchymal stem cell-conditioned media (50 μL), derived from MSCs (2 × 105 cells conditioned 0.5 mL of chemically defined medium with no serum for 24 h) at varying passages, were analyzed by ELISA using the Human Cu/ZnSOD ELISA kit (Bender MedSystems, Vienna, Austria) according to the manufacturer’s instructions. All samples were analyzed in triplicate.

Statistical analysis

Counting data were analyzed using non-parametric tests (Kruskal–Wallis with post hoc Dunn’s testing between groups). Values are expressed as the mean ± SEM from at least five independent experiments, unless otherwise stated.

Results

Neuronal survival is increased in the presence of MSCs and MSC-derived factors

Cerebellar neurons were maintained in B27 supplemented DMEM for 5 days before exposure to injurious conditions. Both total neuronal counts and Purkinje cell counts were significantly decreased when compared to the controls after trophic factor withdrawal (through exposure to minimal base medium for 72 h), or nitric oxide exposure (through a nitric oxide donor DETANONOate 0.1 mM for 24 h; Fig. 1a and b). To investigate the neuroprotective properties of MSCs against both trophic factor withdrawal and nitric oxide (NO)-mediated neurotoxicity, neurons were exposed to these conditions in the presence of MSCs in direct co-culture, a transwell co-culture system (allowing for exchange of factors between MSCs and neurons, but no direct cell-cell contact) or in the presence of MSC-conditioned medium. Cultures were then fixed and stained for βIII tubulin, Calbindin D28K and the nuclear marker DAPI Vectashield. The numbers of viable neurons (determined by nuclear appearance) expressing βIII tubulin and/or Calbindin D28K were counted for each condition. In the presence of MSCs within direct co-culture, a transwell co-culture system or in the presence of MSC conditioned medium neuronal survival, and specifically Purkinje cell survival, was significantly increased during exposure to trophic deprivation or nitric oxide (Fig. 1a–c).

Figure 1.

 (a) The effect of withdrawal of trophic factors [culturing in serum free minimal media (MIN) for 72 h] on cerebellar neuronal/Purkinje cell survival in vitro, in the presence of bone marrow-derived stem cells in direct co-culture (MIN + CO-CUL), bone marrow-derived stem cells in transwells (MIN + TRANS) or MSC-conditioned media (MIN + CM) (number of β-tubulin cells per field expressed as percentage of cells grown in B27 medium; *< 0.05, **< 0.01, ***< 0.001 compared to MIN; = 8). (b) The effect of DETANONOate exposure (0.1 mM for 24 h) on cerebellar neuronal/Purkinje cell survival in vitro (MIN + NO; ***< 0.001 compared to serum free minimal media (MIN); and the effect of bone marrow-derived stem cells in direct co-culture (MIN + CO-CUL), bone marrow-derived stem cells in transwells (MIN + TRANS) or MSC-conditioned media (MIN + CM) on cerebellar neuronal/Purkinje cell survival (number of β-tubulin cells per field expressed as percentage of cells grown in B27 medium; *< 0.05, **< 0.01, ***< 0.001 compared to MIN + NO; = 8). (c) Rat E18 cerebellar neuronal cultures after exposure to B27 base medium (B27), nitric oxide (MIN + NO), or nitric oxide in the presence of bone marrow-derived conditioned medium (MIN + NO + CM) for 24 h. Stained for Calbindin (red), βIII tubulin (green) and Hoescht nuclear stain (blue).

MSC conditioned medium activates the Akt pathway

We investigated the importance of phosphatidylinositol 3-kinase (PI3K)/Akt signalling in the MSC-induced neuroprotection (Almeida et al. 2005; He et al. 2008). Cerebellar neuronal cells were cultured for 5 days before exposure to trophic factor withdrawal or nitric oxide exposure. Previous experiments had determined optimal time points for analysis of intracellular signalling molecules (Wilkins and Compston 2005; Wilkins et al. 2009). After 1-h exposure to test conditions, cells were lysed and immunoblotting performed for activated Phospho-Akt (Ser473). Phospho-Akt was detected at lower levels after exposure to either minimal base medium or DETANONOate for 1 h (Fig. 2a and b). Addition of MSC-conditioned media to neuronal cultures, exposed to trophic factor withdrawal or nitric oxide exposure, led to significant increases in levels phosphorylated Akt, and the presence of the PI3K/Akt signalling inhibitor LY294002 in the MSC-conditioned media abrogated this effect during insult by nitric oxide (Fig. 2a and b). It was noted that after 1-h exposure to trophic deprivation (minimal medium alone), there was some variability in the decrease in levels of Phospho-Akt detected, however in all cases a drop in phosphorylated Akt was evident after 1 h and the addition of MSC-conditioned media to neuronal cultures exposed to trophic factor withdrawal always led to significant increase in levels of phospho-Akt.

Figure 2.

 Immunoblotting of intracellular signalling molecules in cerebellar neurons after exposure to B27 base medium (B27), serum free minimal medium (MIN), DETANONOate (0.1 mM) (MIN + NO), DETANONOate (0.1 mM) plus bone marrow-derived stem cell conditioned medium (MIN + NO + CM) and DETANONOate (0.1 mM) plus bone marrow-derived stem cell conditioned medium plus LY294002 (MIN + NO + CM + LY). Upper panels correspond to phospho-activation of signalling molecules. Lower panel corresponds to appropriate control molecule. (a and b) Akt activation in cerebellar neurons exposed to test condition for 1 h. (c) JNK activation in cerebellar neurons exposed to test condition for 6 h. (d) p38 activation in cerebellar neurons exposed to test condition for 6 h.

MSC conditioned medium reduce JNK and p38 activation after trophic factor withdrawal or NO exposure

Recent evidence has highlighted the importance of MAPK signalling in NO-mediated cell death in a variety of models (Ghatan et al. 2000; Bossy-Wetzel et al. 2004; Wilkins and Compston 2005). We investigated p38 and JNK signalling in neurons exposed to DETANONOate or ‘minimal’ base medium for 6 h (which had previously been optimised as a time point for analysis of these pathways; Wilkins and Compston 2005; Wilkins et al. 2009). Phospho-p38 and JNK were detected at increased levels after exposure to DETANONOate or minimal base medium respectively, when compared to levels in cells exposed to B27 media (Fig. 2c and d). A large increase in p38 signalling was also evident in neurons after nitric oxide exposure when compared to cell cultured in minimal media alone, however this effect was not evident when observing JNK signalling (Fig. 2e and d). Addition of MSC-conditioned media to neuronal cultures exposed to trophic deprivation led to a significant decrease in levels phosphorylated JNK (Fig. 2c). Addition of MSC-conditioned media to neuronal cultures exposed to DETANONOate led to a significant decrease in levels phosphorylated p38 (Fig. 2d).

MSC conditioned medium survival effects require intact PI3K/Akt intracellular signalling

Using the pathway inhibitors to these specific cellular pathways, we investigated the functional significance of the pathways in MSC induced cerebellar neuroprotection. The addition of the PI3K/Akt pathway inhibitor (LY294002) to neurons exposed to DETANONOate or minimal basal medium with MSC condition medium resulted in the loss of the neuroprotective effect of MSC-condition medium on neuronal and Purkinje cell survival (Fig. 3a and b), suggesting that the effect of MSC-conditioned medium on neuronal cell survival requires intact PI3K/Akt intracellular signalling pathways. Inhibition of the mitogen-activated protein kinase/extracellular (MEK/ERK) pathways under the above conditions had no significant effect on whole neuronal or Purkinje cell survival.

Figure 3.

 (a) The effect of intracellular pathway inhibitors [p38 inhibitor: SB 203580 (20 μM; MIN + p38) and JNK II inhibitor: SP600125 (20 μM; MIN + JNK)] on cerebellar neuronal/Purkinje cell survival under conditions of trophic deprivation [culturing in serum free minimal media (MIN) for 72 h]; and the effect of intracellular pathway inhibitors on MSC-conditioned media effects [MIN + CM; PI3K/Akt inhibitor LY294002 (10 μM; MIN + CM + LY); or MAPK/ERK inhibitor PD 98059 (30 μM; MIN + CM + PD)] (number of β-tubulin cells per field expressed as percentage of cells grown in B27 medium; *< 0.05, **< 0.01, ***< 0.001 compared to MIN; = 8). (b) The effect of intracellular pathway inhibitors [p38 inhibitor: SB 203580 (20 μM; MIN + NO + p38) and JNK II inhibitor: SP600125 (20 μM; MIN + NO + JNK)] on cerebellar neuronal/Purkinje cell survival exposed to DETANONOate (MIN + NO; 0.1 mM for 24 h); and the effect of intracellular pathway inhibitors on MSC-conditioned media effects [MIN + NO + CM; PI3K/Akt inhibitor LY294002 (10 μM; MIN + NO + CM + LY); or MAPK/ERK inhibitor PD 98059 (30 μM; MIN + NO + CM + PD)] (number of β-tubulin cells per field expressed as percentage of cells grown in B27 medium; *< 0.05, **< 0.01, compared to MIN + NO; = 8).

The functional significance of modulation of p38 and JNK signalling on neuronal cell survival was assessed using the pathway inhibitors SB 203580 (p38 inhibitor) and SP600125 (JNK II inhibitor) respectively. p38 pathway inhibition led to a significant increase in neuronal cell survival after exposure to both trophic factor withdrawal and nitric oxide, when compared to toxic insult alone (Fig. 3a and b). No significant changes in neuronal cell survival were seen in the presence of SP600125 after exposure to both trophic factor withdrawal and nitric oxide (Fig. 3a and b).

Mesenchymal stem cells secrete superoxide dismutase 3

In order to define precisely the soluble factors which may be involved in cerebellar neuroprotection mediated by MSCs, MSC conditioned medium was subjected to two-dimensional (2D) SDS–PAGE and Sypro Ruby staining. Visualization of protein spots revealed the expression of large numbers of different proteins. Protein spots were then selected for mass spectrometry and identified using a Mass Spec Database specifically designed for peptide mass searches. Using this method SOD was identified as a protein present within the MSC conditioned medium. As SOD3 is the secreted form of SOD, we carried out SDS–PAGE gel electrophoresis and immunoblotting using a SOD3 specific antibody on samples of concentrated MSC-conditioned media to confirm the presence of SOD3 within the media (Fig. 4). A SOD ELISA also demonstrated significant amounts of SOD3 secreted by MSCs at a level of 0.046 pg/cell (±0.0123 SEM, = 3).

Figure 4.

 (a) Immunoblot of serum free minimal medium (MIN) and bone marrow-derived stem cell conditioned medium (CM) for SOD3. (b) and (c) Human MSC cultures immunocytochemically stained for SOD3 (red) and Hoescht nuclear stain (blue) (c); negative control (no primary antibody) (b).

Neuronal survival is increased in the presence of superoxide dismutase 3

In order to demonstrate the functional significance of SOD3 within MSC conditioned medium the Cu/Zn SOD inhibitor DETCA was used. Preliminary experiments determined that a concentration of 3 μM DETCA had no detrimental effect on neuronal survival under control conditions (data not shown). Neurons were exposed to DETANONOate at a higher concentration of 0.2 mM, which resulted in increased loss in neuronal and Purkinje cell survival evident after 24 h exposure, compared to 0.1 mM DETANONOate. Addition of DETCA to the MSC conditioned medium significantly reduced MSC-conditioned medium induced attenuation of neuronal and Purkinje cell death on nitric oxide exposure (Fig. 5).

Figure 5.

 The effect of DETANONOate (0.2 mM for 24 h; MIN + NO) on neuronal/Purkinje cell survival, in the presence of MSC-conditioned media (MIN + NO + CM), MSC-conditioned media with SOD inhibitor (DETCA 3 μM; MIN + NO + CM + SODIN), or human SOD (250 U/mL; MIN + NO + SOD). (number of β-tubulin cells per field expressed as percentage of cells grown in B27 medium; *< 0.05, compared to MIN + NO; = 4).

Next we determined whether a commercially available Cu/Zn SOD, derived from human erythrocytes, improved neuronal survival under conditions of nitric oxide exposure. The addition of SOD (250 U/mL) to neurons in the absence of nitric oxide had no effect on neuronal survival (data not shown). However, there was increased survival of both neurons and Purkinje cells exposed to SOD (250 U/mL) in the presence of nitric oxide when compared to those cultured with nitric oxide alone (Fig. 5).

SOD increases Akt pathway signalling in cerebellar neurons

Cerebellar neuronal cells were cultured for 5 days before nitric oxide exposure. After a 1-h exposure to test conditions, cells were lysed and immunoblotting performed for activated phospho-Akt (Ser473). As shown previously phospho-Akt was detected at lower levels after exposure to ‘minimal’ base medium and DETANONOate for 1 h. Addition of SOD (250 U/mL) to neuronal cultures exposed to nitric oxide exposure led to significant increases in levels phospho-Akt (Fig. 6a). The addition of MSC-conditioned media to neuronal cultures exposed to nitric oxide also led to significant increases in levels phospho-Akt, as previously shown, and further addition of the SOD inhibitor DETCA to the MSC-conditioned media resulting in a decrease in Akt signalling (Fig. 6b).

Figure 6.

 (a) Immunoblotting of Akt signalling in cerebellar neurons after exposure to B27 base media (B27), serum free minimal medium (MIN), DETANONOate (0.2 mM; MIN + NO), DETANONOate (0.2 mM) plus Superoxide dismutase (250 U/mL; MIN + NO + SOD) or bone marrow-derived stem cell conditioned medium (MIN + NO + CM) for 1 h. (b) Immunoblotting of Akt signalling in cerebellar neurons after exposure to B27 base media (B27), serum free minimal medium (MIN), DETANONOate (0.2 mM; MIN + NO), bone marrow-derived stem cell conditioned medium (MIN + NO + CM) or bone marrow-derived stem cell conditioned medium with SOD inhibitor (DETCA 3 μM; MIN + NO + CM + SODIN) or for 1 h. Upper panels correspond to phospho-Akt; lower panel corresponds to total Akt.

Discussion

Here we have performed a series of experiments showing that human bone marrow-derived mesenchymal stem cells (MSCs) are able to protect rodent cerebellar neurons, and specifically Purkinje cells, in vitro against both nitric oxide exposure and trophic factor withdrawal. In addition, we have demonstrated for the first time that MSCs secrete SOD3 which exerts neuroprotective effects. Furthermore we have demonstrated a central role for the PI3K/Akt pathway in mediating these effects.

Stem cell transplantation strategies hold therapeutic promise and are currently being investigated for a number of neurological disorders. To date, MSCs have been shown to improve outcomes in a variety of animal models of neurological disease including experimental autoimmune encephalomyelitis, stroke and spinal cord injury (Zhang et al. 2004, 2006; Himes et al. 2006). Following transplantation, stem cells may theoretically provide neuroprotection and neural repair through differentiation and the replacement of damaged neuronal cells. However, it may be more likely that the synthesis and secretion of active biomolecules or neurotrophic factors by MSCs will promote an environment conducive to neuroprotection and support neuronal cell survival.

Degeneration within the central nervous system during disease is a complex process induced by a wide variety of insults to neurons. The activation and suppression of intracellular signalling pathways play different roles in regulating cell growth, differentiation, survival and death, and mechanisms of neuronal cell death or survival are highly complex and include activation or suppression of several intracellular signalling pathways including PI3K/Akt and members of the MAPK superfamily (Du et al. 2008). Toxic metabolites and trophic factor withdrawal can both induce cell death, but evidence suggests that this is through activation of independent intracellular signalling cascades (Cao et al. 2004). For example, nitric oxide, whilst performing many physiological roles at low levels, has been shown to induce apoptosis in a variety of cultured peripheral and central neurons and be involved in degeneration during central nervous system inflammation (Estevez et al. 1998; Wang et al. 2003), and is associated with cytotoxicity acting through the p38 MAPK pathway (Wilkins and Compston 2005). The death of neurons induced by trophic factor withdrawal is also an important process. Trophically depleted areas of the nervous system can manifest as part of a pathological process and neurons may be rendered vulnerable to death directly as a result, in which context a variety of downstream pathways to cell death may act (Wilkins et al. 2001). In this study, we show that MSCs are able to protect cerebellar neurons, and specifically Purkinje cells, in vitro via cell-cell contact and/or secretion of soluble factors, or through secretion of soluble factors alone under conditions of trophic deprivation and nitric oxide exposure. Secreted protein factors, including neurotrophic factors and cytokines, have attracted much attention as they may have the ability to inhibit death-inducing pathways and also activate cell survival pathways (Goldberg and Barres 2000; Kaplan and Miller 2000).

In an attempt to identify the molecular mechanisms underlying the neuroprotective role of MSCs, we investigated PI3K, MEK/ERK, JNK and p38 MAPK pathway signalling in cerebellar neuronal cells during toxic insults. In neuronal cells the PI3K/Akt pathway has been shown to be an important regulator of neuronal survival (Zhao et al. 2005; Du et al. 2008), while JNK and p38 are often involved in mediating cell death (Wilkins and Compston 2005; Asada et al. 2007; Chen et al. 2007). Sustained activation of p38 MAPK has been associated with neuronal cell death/apoptosis and p38 MAPK specific pathway inhibitors have been shown to promote survival in a number of neuronal cell types (Jin et al. 2002; Park et al. 2002; Sun et al. 2008). p38 MAPK has been shown to be activated during NO mediated neuronal death and inhibitors of the pathway improve neuronal survival during exposure to NO (Lin et al. 2001; Ishikawa et al. 2003; Wilkins and Compston 2005). Reports have demonstrated the need for p38 signalling in neuronal death induced by other stimuli, such as cerebellar neurons undergoing apoptosis in response to low potassium (Yamagishi et al. 2003). The JNK signalling pathway has previously been shown to be activated by withdrawal of trophic factors in cultured neurons (Xia et al. 1995; Eilers et al. 1998). The current study demonstrates that the exposure of cerebellar neurons to either nitric oxide or trophic withdrawal results in activation of both p38 and JNK signalling respectively, however in the presence of MSCs conditioned medium p38 and JNK signalling was partially inhibited. The importance of blocking p38 signalling was demonstrated using the p38 pathway inhibitor SB 203580, as the addition of this blocker during toxic insult resulted in a significant increase in neuronal cell survival, when compared to toxic insult alone. Inhibiting JNK signalling did not result in any increase in cell survival suggesting that, in this system, it does not appear to have a direct role in mediating or attenuating nitric oxide or trophic withdrawal mediated neuronal cell death.

The PI3K/Akt pathway is known as one of the most prominent survival signalling cascades in neurons and is activated by a wide variety of cytokines and growth factors (Isele et al. 2007). The PI3K/Akt pathway has been implicated in the survival of several neuronal types including sensory neurons, sympathetic neurons, cerebellar granule cells and retinal ganglion cells (Crowder and Freeman 1998; Klesse and Parada 1998; Vaillant et al. 1999; Diem et al. 2001; Encinas et al. 2001). Exposure of cerebellar neurons to both nitric oxide and trophic withdrawal resulted in a decrease in Akt signalling and neuronal cell death, however in the presence of MSCs conditioned medium an increase in Akt signalling and cell survival was observed. The importance of intact Akt signalling during toxic insult to cerebellar neurons was also demonstrated using the Akt pathway inhibitor LY294002 to the conditioned medium.

Collectively we have found evidence that exposure of cerebellar neuronal cultures to nitric oxide or trophic withdrawal induces cell death through both common and independent intracellular signalling pathways. Western blot analysis suggests that JNK signalling is only activated during trophic withdrawal as there was no increase in JNK signalling when nitric oxide was added. Blocking p38 and Akt signalling in the cell survival experiments led to an increase or decrease in cell survival respectively, during exposure to either nitric oxide or trophic factor withdrawal. Thus p38 and Akt signalling both appear to be common pathways by which nitric oxide and trophic withdrawal act. Furthermore, western blot analysis revealed differences in the levels of signalling through these pathways after exposure to either nitric oxide or trophic factor withdrawal, suggesting that p38 signalling may play a more important role in nitric oxide-mediated cell death.

Superoxide dismutase-3 expression by MSCs was determined through screening the total protein content within MSC condition medium using 2D SDS–PAGE coupled with mass spectrometry, and confirmed using western blotting and ELISA techniques. SOD3 is a secreted member of the SOD protein family and is the only antioxidant enzymic scavenger of superoxide within the extracellular compartment (Enghild et al. 1999). SODs catalyse the conversion of superoxide (O2) to hydrogen peroxide (H2O2), the latter of which is converted to water and oxygen by catalases and peroxydases (Harris et al. 1991). In doing this, SOD3 prevents superoxide reacting with nitric oxide to deplete nitric oxide’s bioactivity to form the strong oxidant, peroxynitrite (Nozik-Grayck et al. 2005). Thus, SOD3 is a major antioxidant defence that protects tissues within the body from oxidative damage. In this study SOD3 has been shown to increase survival of both cerebellar neuronal cells and specifically Purkinje cells in vitro against nitric oxide mediated toxicity. In addition, we have also demonstrated for the first time that SOD3 activates PI3K/Akt pathways in cerebellar neurons.

A major role for reactive oxygen species (ROS) in the pathophysiology of neurodegeneration has been demonstrated in both pathological and experimental studies (Sayre et al. 2008). ROS include superoxide ions, hydrogen peroxide, nitric oxide and peroxynitrite, all of which are produced as part of the inflammatory response and have a potential role in causing tissue damage within the CNS. ROS are detoxified by a number of enzymes including those of the SOD family, and dysfunction of these enzymes have been associated with inherited abnormalities causing a variety of neurological disorders such as amyotrophic lateral sclerosis (Wijesekera and Leigh 2009). The secretion of active antioxidant molecules, such as SOD3, by MSCs may also explain the recent observations in which MSCs are shown to have a direct antioxidant activity that is conducive to neuroprotection in both experimental autoimmune encephalomyelitis and in vitro (Lanza et al. 2009). With antioxidants emerging as potential therapeutic agents for neurodegenerative disorders caused by oxidative stress (Christofidou-Solomidou and Muzykantov 2006; Ratnam et al. 2006), the secretion of SOD3 by MSCs therefore might lead to the development of new treatment strategies for neurological conditions in which oxidative damage is thought to be a central feature, and may have important implications in inflammatory conditions, such as multiple sclerosis.

Together, our results give evidence of a new antioxidative mechanism by which bone marrow-derived MSCs promote neuronal survival and suggest that MSC-based cellular therapies could be a viable therapeutic approach for cerebellar degeneration particularly those associated with oxidative stress.

Acknowledgements

This work was carried out using a project grant from Ataxia UK.

Ancillary