G protein-coupled receptor 37 is a negative regulator of oligodendrocyte differentiation and myelination

Abstract

While the formation of myelin by oligodendrocytes is critical for the function of the central nervous system, the molecular mechanism controlling oligodendrocyte differentiation remains largely unknown. Here we identify G protein-coupled receptor 37 (GPR37) as an inhibitor of late-stage oligodendrocyte differentiation and myelination. GPR37 is enriched in oligodendrocytes and its expression increases during their differentiation into myelin forming cells. Genetic deletion of Gpr37 does not affect the number of oligodendrocyte precursor cells, but results in precocious oligodendrocyte differentiation and hypermyelination. The inhibition of oligodendrocyte differentiation by GPR37 is mediated by suppression of an exchange protein activated by cAMP (EPAC)-dependent activation of Raf-MAPK-ERK1/2 module and nuclear translocation of ERK1/2. Our data suggest that GPR37 regulates central nervous system myelination by controlling the transition from early-differentiated to mature oligodendrocytes.

Figure 1. Gpr37 is enriched in myelinating glia in the CNS.

(a–c) In situ hybridization. Sagittal-sections of adult rat brain were hybridized to GPR37 antisense probe. GPR37 mRNA is detected in white matter areas: corpus callosum (a; CC), hippocampal fimbria (b; Fi) and the cerebellar white matter tracks (c; WM). (d) RT–PCR of brain RNA isolated from P12 wild-type (WT) and oligodendrocyte-ablated mice (OL-). MAG and MBP were used to monitor genes that are expressed specifically in oligodendrocytes, while actin was used as a control for ubiquitously expressed genes. (e) β-gal staining in a sagittal brain section of adult mice carrying a LacZ allele in the Gpr37 locus: cerebellum (cb), brainstem (bs), cerebral peduncle (cp), corpus callosum (cc), hippocampal fimbria (fi), thalamus (tl), anterior commissure (ac) and optic tract (ot). (f,g) Higher magnification of the boxed areas in e. (h–j) Lac Z activity monitored in whole mount preparations of optic nerve (ON), sciatic nerve (SN), and spinal cord (SC) as indicated. Representative pictures are from P70 heterozygous mice (h) and P20 (i) or P15 (j) homozygous mice. GPR37 is absent in the unmyelinated part of the optic nerve (asterisk in h), as well as in sciatic nerve (i) and spinal nerve roots (arrowheads in j) emanating from the spinal cord (asterisk in j). (k–m) Immunolabelling of P12 mice caudate putamen using antibodies to βgal and Olig2. (n,o) Expression of LacZ in the cerebellum (n), and optic nerve (o) isolated from P7, P12, P15 and P20 heterozygous mice. Asterisk mark the location of the white matter. (p) RT–PCR analysis of GPR37 mRNA expression in mouse brain at the indicated postnatal days. Primers to actin were used as control. The expression of GPR37 at P105 was compared with Gpr37^−/− (KO) mice. (q) Relative mRNA levels of GPR37 and GPR17 determined by real-time PCR analysis of whole brain RNA. Scale bars, (a–c) 100 μm; (e) 1 mm; (f,g) 200 μm; (h–j) 50 μm; (k–m) 20 μm; (n–o) 100 μm.

Figure 2. Absence of Gpr37 results in faster differentiation of oligodendrocytes.

(a) Immunolabelling of wild-type (WT) or Gpr37^−/− OPCs co-cultured for 3 (DIV3) or 7 (DIV7) days with DRG neurons, using antibodies to Olig2, O4 and PLP. (b) The number of Olig2-positive cells (per field of view). (c) Percentage of proliferating cells (labelled for ki67) at DIV3 is comparable between WT and Gpr37^−/− cultures. Total 2111 and 2076 cells were investigated for WT and Gpr37^−/−, respectively. (d) Percentage of O4-positive oligodendrocytes among the total population of Olig2-positive cells (that is, early differentiation). (e) Percentage of PLP-positive-oligodendrocytes among the total population of O4-positive cells (that is, late differentiation). The number of cells already expressing PLP at DIV3 is significantly higher in Gpr37^−/− compared with wild-type oligodendrocytes (e; *P<0.05, n=3 different cultures per each genotype at the indicated time point). (f) Relative fluorescence intensity (arbitrary units) of PLP at DIV7 (*P<0.05, n=3 cultures per each genotype). (g) Immunolabelling of wild-type or Gpr37^−/− OPCs co-cultured for 9 (DIV9) days with DRG neurons, using antibodies to PLP, Caspr, and neurofilament (NF). Three different PLP-positive oligodendrocytes (label: a–c) are marked in each panel. In contrast to the wild-type co-culture, PLP-positive oligodendrocytes in Gpr37^−/− coculture are associated with intense clustering of Caspr at contact sites along the axons. (h) Quantitation of the results showing the per cent of PLP-positive cells that associate with axonal Caspr immunoreactivity (*P<0.05, n=3 different cultures per each genotype). (i) Advanced myelination by Gpr37^−/− oligodendrocytes. Co-cultures (DIV9) prepared using oligodendrocytes of each genotype together with wild-type mouse DRG neurons labelled with an antibody to PLP. High magnification images are shown on the right. (j) The number of PLP-positive myelin segments longer than 15 μm per cell is shown. Bars represent mean±s.e.m. *P<0.001, total of 100 cells each genotype were analysed). Scale bars, (a,i) 100 μm; (g) 40 μm; (i) (insets), 20 μm.

Figure 3. Absence of Gpr37 results in precocious myelination in vivo.

(a) Immunolabelling of WT and Gpr37^−/− P9 mice brains (delineated with a purple line) with an antibody to MBP. (b) Higher magnification of the corpus callosum (white rectangle in a). (c) Fluorescent images of longitudinal sections of optic nerves isolated from P12 PlpRed (upper panel) and PlpRed/Gpr37^−/− (lower panel) mice, immunolabelled with an antibody to APC (CC1). The CC1 (left) and red fluorescent (DsRed; right) signals are shown in separate panels. (d) Lower magnification images showing the distribution of DsRed in PlpRed (upper panel) and PlpRed/Gpr37^−/− (lower panel) optic nerves. (e) Fluorescence intensity of DsRed in optic nerves of the two genotypes (d) is shown per μm² (*P<0.01, n=12 sections from 4 mice per each genotype). (f,g) Electron microscope images showing cross-sections through the corpus callosum of P14 WT (upper panels) and Gpr37^−/− (lower panels) mice. (h) The number of myelinated axons per μm² was significantly higher in Gpr37^−/− than in WT (*P<0.05, 12 images from 3 WT and 8 images from 3 KO mice). (i,j) Gpr37^−/− exhibit significantly thicker myelin during development. G-ratio of myelinated axons in P14 corpus callosum is presented as a function of axon diameter (i), or as an average value (j). Gpr37^−/− exhibit a significantly lower g-ratio than WT mice (WT=155 axons, Gpr37^−/−=219 axons from three mice of each genotype, *P<0.001, Student's t-test). Bars represent mean±s.e.m. Scale bars, (a) 1 mm; (b) 200 μm; (c) 10 μm; (d) 500 μm; (f) 5 μm; (g) 0.2 μm.

Figure 4. Absence of Gpr37 results in hypermyelination of the corpus callosum.

(a,d,g) Electron micrographs of midsagittal sections of the corpus callosum from WT and Gpr37^−/− mice at the age of 2 months (a–c), 4 months (d–f) and 1.5 years (g–i). Representative higher magnification images are shown on the right columns. (b,e,h) g-ratio as a function of axonal diameter. GPR37^−/− shows lower g-ratio than WT. Total of 274 axons (WT) and 373 axons (KO) from three mice (for 2 month), 481 axons (WT) and 497 axons (KO) from 3 mice (for 4 month), 97 axons (WT)and 139 axons (KO) from 2 mice (for 1.5 year) of each genotype were analysed. (c,f,i) The averaged g-ratio of Gpr37^−/− is significantly lower than WT (*P<0.001). Scale bars, 1 μm (left two panels) and 0.2 μm (right two panels). Bars represent mean±s.e.m.

Figure 5. Gpr37-dependent inhibition of oligodendrocyte differentiation is mediated by ERK phosphorylation and nuclear translocation.

(a) Immunolabelling of sagittal brain sections of P12 brain stems from wild type (WT) and KO (Gpr37^−/−) mice using antibodies to phosphorylated ERK (pERK) and Olig2. (b) The number of pERK and Olig2 positive cells in Gpr37^−/− compared with WT brainstem (*P<0.05, n=5 images from two mice per genotype). (c) Western blot analysis of purified OPC cultures at DIV7 (three cultures per each genotype are shown). Blots were incubated with antibodies to pERK or general ERK (ERK) as indicated. (d) Relative value of pERK intensity normalized by general ERK (p=0.02, n=7 cultures each genotype). (e) pERK is present in the nucleus of oligodendrocytes lacking Gpr37) OPC cultures isolated from wild-type (WT) or KO (Gpr37^−/−) mice were fixed at DIV5 and immunolabelled using antibodies to pERK, PLP and Dapi. PLP and ERK immunoreactivity are shown in separate panels along with the Dapi signal. The location of the nucleus is marked with a yellow circle in the high magnification of the boxed area (insets). (f) Percentage of PLP-positive oligodendrocytes showing nuclear localization of pERK in wild-type (WT) and KO cultures. (*P<0.05, n=3 different primary cultures for each genotype; 100 PLP-positive oligodendrocytes were counted per each culture). (g,h) ERK signaling mediates Gpr37 effects on oligodendrocyte differentiation. (g) OPCs isolated from Gpr37^−/− or WT mice as indicated were grown with wild-type DRG neurons. Co-cultures were grown for 1 day in their growth medium and maintained for 6 days in a medium containing 10 μM EPE, 1 μM PLX4032 or DMSO as control before fixing and labelling with antibodies to Caspr and PLP. Nuclei were labelled with Dapi. (h) Quantitation of the results showing the fold change in the per cent of PLP-positive cells that associate with axonal Caspr immunoreactivity in all samples compare to the non-treated wild-type cultures. Gpr37^−/− oligodendrocytes treated with ERK inhibitors showed a significantly low number of Caspr-positive axons, when compared with DMSO-control (*P<0.05, **P<0.01, ***P<0.001; n=3 different primary cultures for each genotype, 100 PLP-positive cells were counted in each treatment). Bars represent mean±s.e.m. Scale bars, (a) 100 μm; (e) 50 μm; and (h) 20 μm.

Figure 6. GPR37 regulates oligodendrocyte myelination through MAPK signaling.

(a) Sequential expression and activity of GPCRs during differentiation of the oligodendrocyte lineage. GPR56 regulates OPC proliferation, while GPR17 and GPR37 negatively regulate two consecutive stages of oligodendrocytes differentiation. (b) A schematic model depicting GPR37 signalling. Relief of GPR37 inhibition results in increase in cAMP and Epac-dependent activation of MAPK cascade, resulting in translocation of phospho-ERK1/2 into the nucleus and myelination. Red lines mark the point of action of the various pharmacological inhibitors used.