|Home | About | Journals | Submit | Contact Us | Français|
Oligodendrocyte (OLG) damage leads to demyelination, which is frequently observed in ischemic cerebrovascular diseases. In this study, we investigated the effect of bone marrow stromal cells (BMSCs) on OLGs subjected to oxygen-glucose deprivation (OGD). N20.1 cells (mouse OLG cell line) were transferred into an anaerobic chamber for 3 hr in glucose-free and serum-free medium. After OGD incubation, OLG cultures were divided into the following groups: 1) OGD alone, 2) OLG cocultured with BMSCs, 3) treatment with the phosphoinostide 3-kinase (PI3k) inhibitor LY294002, 4) LY294002-treated OLGs with BMSC cocultured, and 5) anti-p75 antibody-treated OLGs. After an additional 3 hr of reoxygenation incubation, OLG viability and apoptosis were measured. The mRNA expression in the BMSCs and OLGs was analyzed using quantitative real-time PCR (RT-PCR). Serine/threonine-specific protein kinase (Akt), phosphorylated Akt (p-Akt), p75, and caspase 3 protein expressions in OLGs were measured by Western blot. Our results suggest that BMSCs produce growth factors, activate the Akt pathway, and increase the survival of OLGs. BMSCs also reduce p75 and caspase 3 expressions in the OGD-OLGs, which leads to decreased OLG apoptosis. BMSCs participate in OLG protection that may occur with promoting growth factors/PI3K/Akt and inhibiting the p75/caspase pathways. Our study provides insight into white matter damage and the therapeutic benefits of BMSC-based remyelinating therapy after stroke and demyelinating diseases.
Oligodendrocytes (OLGs) generate the myelin sheaths that enwrap axons and thus play a pivotal role in the proper execution of neural function (Shibata et al., 2000). OLGs are very vulnerable to hypoxic and ischemic insults (Lyons and Kettenmann, 1998; Pantoni et al., 1996); damage to them leads to demyelination, which is frequently observed in ischemic cerebrovascular diseases and contributes to neurological functional impairment (Shibata et al., 2000; Li et al., 2005, 2006; Gresle et al., 2006). The precise mechanisms underlying the ischemia-induced death of OLGs are currently unknown (Shibata et al., 2000). Many factors determine the vulnerability of OLGs after ischemia (Dirnagl et al., 1999), such as caspases (Shibata et al., 2000), excitotoxicity (Matute et al., 2007), and oxidative stress (Imai et al., 2001). Therapeutic interventions are needed to protect OLGs as well as neurons and thereby to reduce neurological deficits after stroke. Our previous studies found that bone marrow stromal cells (BMSCs) improved neurological functional recovery after stroke in rodents (Chopp and Li, 2002; Li et al., 2002; Zhang et al., 2004) and protected neural cells (i.e., neurons, astrocytes) from ischemic injury (Li et al., 2000, 2002, 2005, 2006; Chopp and Li, 2002; Gao et al., 2005). However, the effect of BMSCs on ischemic OLGs has not been investigated.
Growth factors bind to the tyrosine kinase receptors of neural cells, induce the production of phosphatidylinositol 3,4,5 [PtdIns(3,4,5)P3] by phosphoinostide 3-kinase (PI3K; Vlahos et al., 1994), activate the Akt cascade, and thereby play an important neurotrophic role in the central nervous system (CNS; Cantley, 2002). The pan-neurotrophin receptor p75 is a deathinducing receptor that belongs to the tumor necrosis factor receptor superfamily. p75 is a marker of cell damage (Coulson et al., 2000; Casha et al., 2001; Wang et al., 2001; Beattie et al., 2002). Caspases are evolutionarily conserved executioners of programmed cell death in normal development and are also implicated in a variety of pathological conditions, including cerebral ischemia (Nicholson and Thornberry, 1997). p75 induces apoptosis through its death domain. Caspases (e.g., caspase 3) are activated by p75 expression, indicating the activation of the common effector pathway of apoptosis (Casha et al., 2001; Wang et al., 2001). Caspases are involved in hypoxia- and ischemia-induced OLG death (Shibata et al., 2000). In this study, we investigated the effect of BMSCs on OLGs in an oxygen-glucose deprivation (OGD) condition and the molecular mechanisms underlying their protective effects.
An immortalized mouse OLG cell line (N20.1, generously provided by Dr. Anthony Campagnoni, University of California at Los Angeles) was employed in this study. N20.1 cells were obtained from mouse primary cultures of OLGs conditionally immortalized by transformation with a temperature-sensitive large T-antigen (Verity et al., 1993). N20.1 cells grow constantly in Dulbecco’s modified Eagle’s medium (DMEM)/F12 with 10% fetal bovine serum (FBS) and G418 (100 μg/ml) at 34°C (permissive temperature), and they differentiate into mature OLGs in DMEM/F12/1% FBS and G418 at 39°C (nonpermissive temperature; Paez et al., 2004). Therefore, in the present experiments, the N20.1 cell lines were placed in DMEM/F12 high glucose (Invitrogen, Carlsbad, CA), with 3.6 g/liter dextrose anhydrous, 3.38 g/liter HEPES, 2.16 g/liter sodium bicarbonate, 90 mg/liter gentamicin, 1% FBS, and 100 μg/ml G418 at 39°C (nonpermissive temperature) for 7 days.
The mouse BMSCs were isolated from C57/Bl6 mice and cultured as previously described (Li et al., 2006). Briefly, the marrow was removed from hind legs. The cells were plated at a density of 9 × 105/cm2 with alpha modified Eagle’s medium supplemented with 10% FBS and penicillin/streptomycin. After 2 days, the medium containing nonadherent cells was aspirated. Approximately 14 days later, the adherent stromal cells become confluent and were designated passage 0 (P0). The BMSCs were expanded and passaged to P3, at which point they were harvested and employed in the present study.
OLG cultures were transferred into an anaerobic chamber (model 1025; Forma Scientific) for 3 hr in glucose-free and serum-free medium (OGD). After OGD incubation, OLG cultures were divided into: 1) the OGD-alone group (OGD-OLGs); 2) the BMSC coculture group (OLGs: BMSCs = 2:1); an insert (0.4 μm; BD Biosciences, San Jose, CA) was used to contain BMSCs. OLGs were plated on the base of the culture wells, and the upper transwell compartments were seeded with BMSCs (BMSC-OLGs); 3) the LY294002 [2-(4-morpholinyl)-8-phenyl-4H-1-benzopyran-4-one hydrochloride, an inhibitor of PI3k; Calbiochem, La Jolla, CA; No. 440202; 10 μm]-treated group (LY-OLGs); 4) the LY294002-treated coculture of BMSC-OLG group (LY-BMSC-OLGs); and 5) the anti-p75 antibody (Chemicon, Temecula, CA; AB1554; 1:1,000)-treated group (anti-p75-OLGs; n = 4 in each group). To mimic the in vivo conditions, where the blood flow and oxygen supply increase gradually after stroke, OGD-OLGs with or without treatment were incubated in the reoxygenation incubator in medium with reduced glucose (1.69 g/liter) for an additional 3 hr.
After 3 hr post-OGD incubation, OLG viability was measured with a LIVE/DEAD Kit (Invitrogen, Molecular Probes). The OLG cultures were incubated with viability assay solution containing 2 μm calcein-AM and 4 μm ethidium homodimer-1 at room temperature for 30 min. After incubation, the cells were examined under a fluorescent microscope (Nikon, eclipse, TE2000-U). The extent of cell survival was calculated by counting in 10 random fields in each well with three wells per group. The results are presented as a percentage (survival cell divided by total cells).
After 3 hr post-OGD incubation, cell apoptosis assays were performed with Hoechst 33342/propidium iodide (PI) double staining (Chemicon; Gao et al., 2005) and TUNEL method (Chemicon; ApopTag Fluorescein In Situ Apoptosis Detection Kit), respectively, according to the manufacturer’s instructions. The OLG cultures were incubated with Hoechst 33342 (1 μg/ml) and PI (1.25 μg/ml) for 5 min. In contrast to normal cells, the nuclei of apoptotic cells have highly condensed chromatin that is uniformly stained by Hoechst 33342. This can take the form of crescents around the periphery of the nucleus, or the entire nucleus can appear to be one or a group of featureless, bright, spherical beads. These morphological changes in the nuclei of apoptotic cells are visualized by fluorescence microscopy. Costaining of the cells with PI allows the discrimination of necrotic cells from apoptotic cells (Dive et al., 1992). The apoptotic cells were calculated from Hoechst 33342-stained cells with shrunken and highly condensed nuclei in 10 random fields in each well with three wells per group.
For the TUNEL method, briefly, cells were fixed in 1% paraformaldehyde and postfixed in ethanol:acetic acid (2:1) for 5 min at -20°C. After application of equilibration buffer, cells were incubated with TdT enzyme at 37°C for 1 hr. Incubated antidigoxigenin was conjugated with the cells for 30 min at room temperature. After washing with PBS, DAPI was applied to stain nuclei. Cell were rinsed with PBS and observed under a fluorescent microscope (Nikon eclipse, TE2000-U). The apoptotic cells were calculated by counting in 10 random fields in each well with three wells per group. The results are presented as a percentage (apoptotic cells divided by total cells).
After BMSCs were cocultured with OGD-OLGs for 3 hr, mRNA expression in BMSCs and OLGs was measured by using the SYBR Green RT-PCR method (Wang et al., 2005; Zhang et al., 2005). Total RNA was isolated from BMSCs and OLGs using the RNeasy Micro Kit (Qiagen, Valencia, CA). One microgram of RNA from each sample was used to produce cDNA, following the standard protocol supplied with the SuperScript III RTase (Invitrogen). qRT-PCR was performed on an ABI 7000 PCR instrument (Applied Biosystems, Foster City, CA). Specificity of the produced amplification product was confirmed by examination of dissociation reaction plots. A distinct single peak indicated that a single DNA sequence will be amplified during PCR. PCR products were run on 2% agarose gels to confirm that correct molecular sizes. Each sample was tested in triplicate using qRT-PCR. Data analysis used the 2-ΔΔCT method (Livak and Schmittgen, 2001). The following primers for RT-PCR were designed as follows: glyceraldehyde-3-phosphate dehyrogenase (GAPDH; FWD AGAACATCATCCCTGCATCC, REV CACATT GGGGGTAGGAACAC), nerve growth factor (NGF; FWD CAAGGACGCAGCTTTCTATACTG, REV CTTCAGG GACAGAGTCTCCTTCT); glial cell-derived neurotrophic factor (GDNF; FWD GATATTGCAGCGGTTCCTGT, REV AACATGCCTGGCCTACTTTG), brain-derived neurotrophic factor (BDNF; FWD TACTTCGGTTGCATGAA GGCG, REV GTCAGACCTCTCGAACCTGCC), insulinlike growth factor (IGF-1; FWD GGCATTGTGGATGA GTGTTG, REV GTCTTGGGCATGTCAGTGTG), Akt (FWD GACCATGAACGAGTTTGAGTACCT, REV CTT GAGGATCTTCATGGCATAGTA), caspase 3 (FWD ATG GGAGCAAGTCAGTGGAC, REV CGTACCAGAGCGAG ATGACA), p75 (FWD CATCTCTGTGGACAGCCAGA, REV CAGCTTCTCGACCTCCTCAC), and tyrosine kinase receptor A (TrkA; FWD CGTCATGGCTGCTTTTATGG, REV ACTGGCGAGAAGG AGACAG).
After 3 hr of post-OGD treatment, proteins from OLG cultures were extracted in 200 μl RIPA lysis buffer. Equal amounts of proteins, as determined using the bicinchoninic acid (BCA) protocol (Pierce, Rockford, IL), were loaded on 12% Bis-Tris gels (Invitrogen) after being denatured. The proteins were then transferred to Invitrogen PVDF membranes (Invitrogen), which were blocked for 1 hr with 5% bovine serum albumin (BSA) in TBS-T (10 mM Tris-HCl, pH 7.6, and 150 mM NaCl, 0.1% Tween-20). Afterward, the membranes were incubated with primary antibodies against Akt (Cell Signaling, Beverly, MA; 9272; 1:1,000), phosphorylated Akt (p-Akt; Cell Signaling; 9271; 1:1,000), p75 (Chemicon; AB1554; 1:1,000), and caspase 3 (Cell Signaling; 9661; 1:1,000) in 3% BSA at 4°C overnight. The membranes were washed with TBS-T and incubated for 1 hr at room temperature with horseradish peroxidase (HRP)-conjugated secondary antibodies (Bio-Rad Laboratories, Hercules, CA). After washing, the immunoblots were detected using a SuperSignal West Pico Chemiluminescent Substrate kit (Pierce). The experiment was repeated in triplicate. β-Actin was used as the internal control, and the densities of bands were analyzed in Scion Image (Scion Corp., Frederick, MD).
The percentages of live cells and apoptotic cells were measured among four groups. Comparisons were made between treatment groups using analysis of variance (ANOVA). If the overall group difference was significant (P < 0.05), pairwise comparisons were made. Measures of NGF, GDNF, IGF-1, and BDNF mRNA expression were compared between normal BMSCs and BMSCs treated with OGD using two-sample t-tests. ANOVA was used to compare mRNA and protein expression among normal OLGs and OLGs treated with OGD and BMSCs. Pairwise comparisons were made if the overall treatment group difference was significant (P < 0.05). Data are presented as mean ± SD.
By using the LIVE/DEAD Kit to measure OLG viability, we found that approximately 20.8% of OLGs were dead in the OGD group after 3 hr of OGD injury and an additional 3 hr of post-OGD incubation (Fig. 1a). Cell survival was significantly increased in the BMSC-OLG group compared with the OGD-OLG group (P < 0.01; Fig. 1b), indicating the benefit of BMSCs on the survival of OLGs after OGD injury. The PI3K/Akt pathway is pivotal for cell survival (Cantley, 2002; Shah et al., 2005). In the present study, we employed LY294002, which is a potent and selective PI3K inhibitor (Vlahos et al., 1994). After LY294002 treatment, there was no significant decrease cell survival in LY-OLGs compared with the OGD-OLGs (Fig. 1c). Treatment with LY294002, blocked the BMSC survival benefits on OLGs after OGD injury (Fig. 1d,e), the cell survival of LY-BMSC-OLGs was significantly decreased compared with BMSC-OLGs, suggesting that the PI3K/Akt pathway plays an important role in the BMSC protective benefits on OLGs after OGD injury.
To test the BMSCs effect on the apoptosis of OLGs, we employed Hoechst 33342/PI double staining and TUNEL staining 3 hr post-OGD, respectively. With Hoechst 33342/PI double staining, approximately 51.9% of OLGs underwent apoptosis in the OGD group with shrunken or condensed bright blue nuclei, in the form of crescents around the periphery of the nucleus or the entire nucleus, which appear to be featureless, bright, spherical beads (Fig. 2a). After LY294002 treatment, there was no significant increase in cell apoptosis compared with the OGD group (Fig. 2b). Apoptotic OLGs were significantly decreased after BMSC coculture (Fig. 2c) and after p75 blocking antibody treatment (Fig. 2d) compared with the OGD group (P < 0.01; Fig. 2e). With the TUNEL method, similar results were obtained (Fig. 3a-e). The protection of OLGs from apoptosis by BMSCs was more effective than that of single p75 antibody treatment (P < 0.05), suggesting that the benefits of BMSCs on apoptotic OLGs may be partially mediated by p75.
The PI3K/Akt pathway is activated by growth factors that promote cell survival. Therefore, in the present study, we measured the growth factor expression in BMSCs and investigated the mechanisms underlying the BMSC benefits on OGD-OLGs. qRT-PCR data revealed that normal BMSCs expressed mRNA of NGF, GDNF, IGF-1, and BDNF. After coculture with the OGD-OLGs in the reduced glucose and serum medium, NGF mRNA level was significantly increased, GDNF mRNA level did not change, and the expression of IGF-1 and BDNF mRNA was significantly decreased in BMSCs compared with the normal BMSCs (Fig. 4A). These data revealed that the BMSC growth factor gene expression is sensitive to the medium condition and is altered when cocultured with damaged OLGs. Thus, there may be feedback mechanisms of gene expression in BMSCs that foster protection of OLGs after OGD injury.
Next, we measured the expression of TrkA (the high-affinity NGF receptor) and other important genes involved in the PI3K/Akt and p75/caspase pathways in normal OLGs, OGD-OLGs, and BMSC-OLGs. TrkA, p75, and caspase 3 mRNA expressions were significantly increased and Akt mRNA level was significantly decreased in the OGD-OLGs group compared with normal OLGs (P < 0.01; Fig. 4B). After treatment with BMSCs, TrkA and Akt mRNA levels in the OLGs were significantly increased, and p75 and caspase 3 mRNA levels were siginificantly decreased compared with OGD-OLGs.
To confirm our hypothesis that the protective effect of BMSCs on OGD-OLGs is mediated via PI3K/Akt and p75/caspase 3 pathways, we employed Western blots to measure p-Akt, p75, and caspase 3 expression. Western blots showed that p-Akt levels were significantly decreased and protein levels of p75 and caspase 3 were significantly increased in OLGs after 3 hr of OGD injury and an additional 3 hr post-OGD incubation. These data suggest that the ability of the OLG survival decreased, and apoptosis pathways are activated in OLGs after OGD. However, BMSC treatment increased p-Akt protein expression and inhibited increased p75 and caspase 3 expression levels in the OGD-OLGs (Fig. 5). These data are consistent with the cell viability and apoptosis data.
OLGs are important components in the CNS and are the only myelin-producing cells that allow rapid electrical conduction of impulses (Wilkins et al., 2003). The lack of oxygen and glucose and free radicals contribute to OLG damage after stroke (Beckman and Koppenol, 1996; Shibata et al., 2000). BMSC treatment of stroke in rodents improves neurological recovery and enhances reactive OLG- and astrocyte-related axonal remodeling (Li et al., 2006). The present study revealed that BMSC treatment increased the viability and reduced apoptosis of OLGs after 3 hr of OGD injury and an additional 3 hr post-OGD incubation. These benefits were associated with growth factors expressed by BMSCs and increased p-Akt and decreased p75 and caspase 3 in OLGs cocultured with BMSCs.
The OGD model provides a standard and widely accepted, simplified in vitro model for extracting mechanistic information on focal stroke (Mielke et al., 2006; Thompson et al., 2006; McCarran and Goldberg, 2007). After occlusion of the middle cerebral artery in the animal or stroke in the patient, cerebral blood flow and thereby glucose deprivation persist for hours. Hypoperfusion of the brain after ischemic injury is somewhat simulated in vitro by the delay in reinstituting glucose. Therefore, we used reduced-glucose (1.69 g/liter) medium post-OGD. Glucose added to the culture medium (in concentrations up to 25 mM) prevented hypoxiainduced cell death in a concentration-dependent manner (Yoshioka et al., 2000). Moreover, when we double the glucose level in the medium after OGD, OLG damage was reduced after 3 hr post-OGD injury (data not shown). Our objective was to provide a proof of principle that BMSCs reduce OLG damage in a stressed, ischemic environment, e.g., OGD, by multiple mechanisms. Simulating the in vitro situation, the complexity of the in vivo conditions of ischemia, reperfusion, and glucose levels, is beyond the goals of the present study. However, we demonstrate for the first time the neuroprotective effects of BMSCs on OLGs.
BMSCs are a mixed cell population, including multipotential mesenchymal stem and precursor cells, and appear to have broad therapeutic applications to neurological disease. Functional recovery was evident after BMSC treatment in rodents with stroke (Chen et al., 2001; Li et al., 2002), experimental autoimmune encephalomyelitis (EAE; Zappia et al., 2005; Zhang et al., 2005), traumatic brain injury (Mahmood et al., 2001), and Parkinson’s disease (Li et al., 2001). We also found that the BMSC treatment of stroke in rodents reduced apoptosis in the penumbral zone of the lesion (Li et al., 2002; Chen et al., 2003) and protected against astrocyte apoptosis after OGD and reperfusion injury in vitro (Gao et al., 2005). Expression of neurotrophic and growth factors, i.e., NGF (Mahmood et al., 2004; Zhang et al., 2006), BDNF (Wang et al., 2004), IGF-1 (Zhang et al., 2004), VEGF (Chen et al., 2002b), and bFGF (Chen et al., 2002a), stimulated by BMSCs likely contribute to the beneficial effects of the treatment (Chopp and Li, 2002).
Akt is a serine/threonine kinase and plays a critical regulatory role in diverse cellular processes (Brazil et al., 2002; Shiojima and Walsh, 2002). Akt is a major regulator of insulin signaling and glucose metabolism (Hajduch et al., 2001) and regulates cell growth and proliferation. Akt also mediates growth factor-associated cell survival (Shah et al., 2005) by inhibiting apoptosis through its ability to phosphorylate and inactivate several targets, including Bad and Forkhead transcription factors (Coffer et al., 1998; Brunet et al., 1999). Akt is activated downstream from PI3K (Cantley, 2002; Vivanco and Sawyers, 2002), which appears to be centrally involved with growth factor signal transduction (Vlahos et al., 1994; Ulrich et al., 1998; Vaillant et al., 1999), including NGF (Xie et al., 2000; Itakura et al., 2005; Rahbek et al., 2005; Lin et al., 2006), BDNF (Bhave et al., 1999; Dolcet et al., 1999; Foulstone et al., 1999; Rajagopal et al., 2004), IGF-1 (Kulik et al., 1997; Fujio et al., 2000; Mockridge et al., 2000; Politi et al., 2001; Yamaguchi et al., 2001; Itakura et al., 2005), and GDNF (Mograbi et al., 2001; Focke et al., 2003; Veit et al., 2004; Anitha et al., 2006; Hauck et al., 2006; Tsui et al., 2006; Villegas et al., 2006; Lee et al., 2007; Braydich-Stolle et al., 2007; Wang et al., 2007). p-Akt is activatedAkt (Shah etal., 2005). Akt is therefore an important therapeutic target for the treatment of stroke and neurodegenerative diseases. Our data show that BMSCs expressed growth factors, i.e., NGF, BDNF, IGF-1, and GDNF, although the expression of some growth factors in the medium with reduced glucose and serum was lower than that of normal medium. The mRNA level of TrkA, the high-affinity receptor of NGF, was significantly increased in BMSC-OLGs compared with OGD-OLGs. These growth factors bind to the tyrosine kinase receptors of the OLGs subjected to OGD and, thereby, activate the Akt cascade via PI3K, which induced p-Akt, leading to improved OLG survival and reduced the apoptosis of OGD-OLGs. The PI3K inhibitor LY 294002 blocked this survival benefit of BMSCs on OGD-OLGs, implying that BMSCs improved OLG survival via the PI3K/Akt cascade.
The pan-neurotrophin receptor p75 has been implicated in damage-induced cell death (Casaccia-Bonnefil et al., 1996; Yoon et al., 1998; Coulson et al., 2000; Syroid et al., 2000; Casha et al., 2001; Wang et al., 2001; Beattie et al., 2002). p75 is required for the death of neural cells after CNS injury (Lee et al., 2001; Beattie et al., 2002). OLGs undergoing apoptosis expressed p75, and the absence of p75 resulted in a decrease in the number of apoptotic OLGs and increased survival of OLGs (Beattie et al., 2002). Our results also show that a p75 blocking antibody decreased apoptotic OLGs. Caspase 3 is activated by p75 expression and induces the common apoptotic pathway (Wang et al., 2001). The present study revealed that, although OGD injury induced p75 and caspase 3, treatment with BMSCs significantly inhibited transcriptional activity of p75, caspase-3-like activation, and OLG death.
The present data in concert with our previous studies indicate that the BMSC capacity to increase expression of growth and trophic factors may be key to the neuroprotective and neurorestorative benefits found in neural cells after ischemic injury (Chen et al., 2002b; Chopp and Li, 2002; Zhang et al., 2004; Gao et al., 2005). Growth and trophic factors bind to the tyrosine kinase receptors of neural cells, activate the PI3K/Akt cascade (Vlahos et al., 1994), and thereby play an important neurorestorative role in the CNS (Cantley, 2002). Moreover, the present study also found that BMSC can inhibit p75/caspase pathways and reduce apoptosis of OLGs after OGD and reperfusion injury. These data provide insight into white matter damage and the therapeutic benefits of BMSC-based remyelinating therapy after stroke (Chopp and Li, 2002; Li et al., 2005, 2006) and demyelinating diseases (Zhang et al., 2005).
The authors thank Dr. Mark Katakowski and Qinge Lu for their technical assistance and Deborah Jewell for secretarial support.
Contract grant sponsor: Benson Ford Foundation; Contract grant sponsor: NIH; Contract grant number: PO1 NS42345; Contract grant number: RO1 NS45041.