Regulated release of BDNF by cortical oligodendrocytes is mediated through metabotropic glutamate receptors and the PLC pathway

Abstract

A number of studies suggest that OLGs (oligodendrocytes), the myelinating cells of the central nervous system, are also a source of trophic molecules, such as neurotrophins that may influence survival of proximate neurons. What is less clear is how the release of these molecules may be regulated. The present study investigated the effects of BDNF (brain-derived neurotrophic factor) derived from cortical OLGs on proximate neurons, as well as regulatory mechanisms mediating BDNF release. Initial work determined that BDNF derived from cortical OLGs increased the numbers of VGLUT1 (vesicular glutamate transporter 1)-positive glutamatergic cortical neurons. Furthermore, glutamate acting through metabotropic, and not AMPA/kainate or NMDA (N-methyl-d-aspartate), receptors increased BDNF release. The PLC (phospholipase C) pathway is a key mediator of metabotropic actions to release BDNF in astrocytes and neurons. Treatment of OLGs with the PLC activator m-3M3FBS [N-(3-trifluoromethylphenyl)-2,4,6-trimethylbenzenesulfonamide] induced robust release of BDNF. Moreover, release elicited by the metabotropic receptor agonist ACPD [trans-(1S,3R)-1-aminocyclopentane-1,3-dicarboxylic acid] was inhibited by the PLC antagonist U73122, the IP3 (inositol triphosphate 3) receptor inhibitor 2-APB (2-aminoethoxydiphenylborane) and the intracellular calcium chelator BAPTA/AM [1,2-bis-(o-aminophenoxy)ethane-N,N,N',N'-tetra-acetic acid tetrakis(acetoxymethyl ester)]. Taken together, these results suggest that OLG lineage cells release BDNF, a molecule trophic for proximate neurons. BDNF release is regulated by glutamate acting through mGluRs (metabotropic glutamate receptors) and the PLC pathway. Thus glutamate and BDNF may be molecules that support neuron-OLG interactions in the cortex.

Figure 1. OCM increases the number of VGLUT1(+) neurons

VGLUT1(+) neurons/mm² are increased when grown in cortical OCM for 5 days. The OCM effect is blocked by an anti-BDNF neutralizing antibody. Values are the percentage of the control±S.E.M (n = 4). In total, three cultures were assayed per condition in each experiment. *P<0.05, significantly different compared with the control. **P<0.05, significantly different compared with OCM alone. Data were analysed by ANOVA and Fisher’s test.

Figure 2. Cortical OLGs express vesicular proteins and respond to glutamate by the release of BDNF

(A) Cortical OLGs express vesicular proteins. Immunocytochemical analysis revealed the presence of VAMP2, secretogranin II and chromogranin A. The representative negative control did not receive primary antibody. Scale bar = 50 μm. (B) Glutamate (100 μM) elicits the release of BDNF after 10 min stimulation as determined by ELISA (n = 6). Values are the mean BDNF (pg/50 μl) levels expressed as a percentage of the control±S.E.M. *P<0.05, significantly different compared with the control. Data were analysed using a paired Student’s t test.

Figure 3. Cortical OLGs express glutamate receptors in culture

Immunocytochemical analysis revealed the presence of (A) ionotropic AMPA GluR1–4, kainate GluR6/7 and NMDA NR3A subunits and (B) mGluR1, mGluR2/3 and mGluR5 in OLGs. The representative negative control did not receive primary antibody. Scale bar = 50 μm.

Figure 4. BDNF release by cortical OLGs in response to glutamate is mediated by metabotropic, and not ionotropic, receptors

A 10 min stimulation with 200 μM kainate (A) or 100 μM NMDA (B) does not elicit BDNF release, as determined by ELISA (n = 5 for A and n = 3 for B). NMDA-treated cells and their vehicle control were co-treated with glycine (5 μM). (C) ACPD (10 μM) elicits BDNF release in the medium after 10 min stimulation, as determined by ELISA (n = 9). (D) ACPD (10 μM for 10 min) elicits an increase in BDNF release as determined by Western blot analysis. The histogram represents the densitometric analysis of three independent Western blot experiments. (E) Pretreatment with MCPG (300 μM for 4 h), inhibits the glutamate-induced BDNF release as determined by ELISA (n = 5). Glutamate (−MCPG) is compared with control (−MCPG), whereas glutamate (+MCPG) is compared with control (+MCPG) in the same experiment. Values represent BDNF (pg/50 μl) levels expressed as a percentage of the control±S.E.M. *P<0.05, significantly different compared with the control. Data were analysed using a paired Student’s t test.

Figure 5. Group I, but not Group II, mGluRs mediate BDNF release

(A) DHPG (100 μM) induces BDNF release into the medium after 10 min stimulation, as determined by ELISA (n = 6). *P<0.05, significantly different compared with the control. (B) DCG-IV (10 μM) does not induce BDNF release after 10 min stimulation, as determined by ELISA (n = 3). Values are BDNF (pg/50 μl) levels expressed as a percentage of the control±S.E.M. Data were analysed using a paired Student’s t test. *P<0.05, significantly different compared with the control.

Figure 6. mGluRs mediate BDNF release through the PLC pathway

(A) The PLC agonist m-3M3FBS (25 μM) induces BDNF release after 20 min stimulation, as determined by ELISA (n = 8). (B) The PLC inhibitor U73122 (1 μM with 15 min pretreatment) inhibits the ACPD-induced BDNF release in OLGs, as determined by ELISA (n = 3). (C) The IP₃ antagonist 2-APB (150 μM with 30 min pretreatment) inhibits the ACPD-induced BDNF release in OLGs, as determined by ELISA (n = 3). (D) The intracellular calcium chelator BAPTA-AM (80 μM with 3 h pretreatment) inhibits the ACPD-induced BDNF release in OLGs, as determined by ELISA (n = 3). Each treatment condition is compared with its own control, not treated with agonist. In all experiments the effects of ACPD in comparison with the control were run with or without inhibitor in the same experiment. Values represent the mean BDNF (pg/50 μl) levels expressed as a percentage of the control±S.E.M. *P<0.05, significantly different compared with the control. Data were analysed using a paired Student’s t test.

Figure 7. Treatment of OLGs with glutamate (100 μM) or ACPD (10 μM) for 10 min does not activate caspase 3 or affect cell number

(A) Western blot analysis of cell lysates at 24 h post-stimulation reveals no change in activated (Act.) caspase 3 levels in control vehicle (ctrl), glutamate (glut)- or ACPD-treated OLGs. Staurosporine (300 nM for 4 h) is used as a positive control (strspo). (B) Densitometric analysis of three independent Western blot experiments. Values are the means±S.E.M. (C) Total cell numbers evaluated at 24 h post-stimulation reveal no difference among control vehicle, glutamate- and ACPD-treated groups. Values are the mean cells/mm² expressed as a percentage of the control±S.E.M. Results represent three independent experiments and each condition within an experiment was performed with three cultures. Data were analysed using ANOVA.