Nerve Growth Factor Neutralization Promotes Oligodendrogenesis by Increasing miR-219a-5p Levels

Abstract

In the brain, the neurotrophin Nerve growth factor (NGF) regulates not only neuronal survival and differentiation, but also glial and microglial functions and neuroinflammation. NGF is known to regulate oligodendrogenesis, reducing myelination in the central nervous system (CNS). In this study, we found that NGF controls oligodendrogenesis by modulating the levels of miR-219a-5p, a well-known positive regulator of oligodendrocyte differentiation. We exploited an NGF-deprivation mouse model, the AD11 mice, in which the postnatal expression of an anti-NGF antibody leads to NGF neutralization and progressive neurodegeneration. Notably, we found that these mice also display increased myelination. A microRNA profiling of AD11 brain samples and qRT-PCR analyses revealed that NGF deprivation leads to an increase of miR-219a-5p levels in hippocampus and cortex and a corresponding down-regulation of its predicted targets. Neurospheres isolated from the hippocampus of AD11 mice give rise to more oligodendrocytes and this process is dependent on miR-219a-5p, as shown by decoy-mediated inhibition of this microRNA. Moreover, treatment of AD11 neurospheres with NGF inhibits miR-219a-5p up-regulation and, consequently, oligodendrocyte differentiation, while anti-NGF treatment of wild type (WT) oligodendrocyte progenitors increases miR-219a-5p expression and the number of mature cells. Overall, this study indicates that NGF inhibits oligodendrogenesis and myelination by down-regulating miR-219a-5p levels, suggesting a novel molecular circuitry that can be exploited for the discovery of new effectors for remyelination in human demyelinating diseases, such as Multiple Sclerosis.

Keywords: Nerve growth factor (NGF); demyelinating diseases; miR-219a-5p; microRNAs; myelin; oligodendrogenesis.

Figure 1

NGF neutralization increases myelination in vivo. Anti-myelin basic protein (MBP) immunostaining (in red) is higher in AD11 brain areas such as the corpus callosum (CC), the cerebral cortex (CTX), and the septum (SEP), as compared to control animals (VH). The histogram reports the quantification of the net fluorescence intensity in these areas. Nuclei labeled with 4,6-diamidine- 2-phenylindole dihydrochloride (DAPI) are in blue. Scale bar 75 µm; magnification 40×. The results are expressed as the mean ± standard error (SEM) from three animals for each genotype (quantifying six slices per animal). Statistical significance is calculated relative to the VH sample for each area. Student’s t-test: * p < 0.05, ** p < 0.01.

Figure 2

MicroRNA-219 (MiR-219) is up-regulated in AD11 mice. Relative Log2 expression ratio for a selected list of differentially expressed microRNAs in the hippocampus of AD11 anti-NGF mice at one month of age (AD11 vs. VH control, n = 4 vs. n = 4 samples). The list includes miR-219, which shows the largest fold change ratio. MicroRNA genes have been selected from a microarray experiment, using the following criteria: |Log2 FC| > 0.58 (FC > 1.5 in linear scale), heteroscedastic Student’s t-test p < 0.05.

Figure 3

Validated miR-219 target mRNAs are downregulated in AD11 brains. (A): Up-regulation of miR-219 in the cortex (CTX) and hippocampus (HP) of AD11 mice at 1 month of age, by qRT-PCR. The normalized Ct of AD11 mice (blue) and VH mice (green) are shown. Statistical significance by one-tail heteroscedastic Student’s t-test: * p < 0.05, ** p < 0.01. The Ct axis is in reverse order to highlight the up-regulation of miR-219. The results are expressed as the mean ± standard error (SEM) from independent experiments (n = 3–7). (B,C): Log2 ratio of differentially expressed miR-219 target mRNAs in AD11 hippocampus (B) and cortex (C) at one month of age by qRT-PCR (AD11/VH control). The results are expressed as the mean ± standard error (SEM) from independent experiments (n =7). Statistical significance is calculated relative to the control samples. Student’s t-test: * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 4

NGF neutralization promotes oligodendrogenesis in AD4 adult hippocampal progenitors. (A): Immunostaining for O4 (green) and MBP (red) in AD4 and in wild type (WT) neurosphere cultures upon oligodendrocyte (OL) differentiation. (B): Quantification of O4+ cells as in A (a total of 592 AD4 cells and 924 WT cells were examined). (C): AD4-derived OLs are more differentiated, in terms of number and length of branches and in terms of MBP expression, when compared to WT-OLs, as quantified in the histograms. Nuclei labeled with DAPI are in blue. Scale bars, 50 and 10 µm, magnification 20× and 40×, respectively. (n = 20). (D): Up-regulation of miR-219 in AD4 neurospheres by qRT-PCR from four independent experiments (n = 4). The results are expressed as the mean ± standard error (SEM). Fold change and Statistical significance are calculated relative to the WT sample. Student’s t-test: * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 5

NGF neutralization promotes OLs differentiation through miR-219 modulation. (A): MiR-219 down-regulation in AD4 neurospheres upon transduction with a decoy_miR-219 lentivirus. qRT-PCR results are reported as the ratio between miR-219 levels in decoy_miR-219 versus decoy_miR_cntr infected AD4 cells. The results are expressed as the mean ± standard error (SEM) from five independent experiments (n = 5). (B): Immunostaining for O4 (red) in AD4 progenitors infected with decoy_cntr or decoy_miR-219 lentiviruses. Inhibition of miR-219 in AD4 cells leads to poorly differentiated OL, without processes (arrow), compared to decoy_cntr infected cells (arrowheads), as quantified in the histograms (lower panel) (n = 20 O4+ cells). Scale bar 10 µm, magnification 60×. The nuclei labeled with DAPI are in blue. Statistical significance is calculated relative to the AD4 decoy_cntr sample. Student’s t-test: * p < 0.05, ** p < 0.01.

Figure 6

miR-219 overexpression promotes OL differentiation of WT neurospheres. (A): MiR-219 up-regulation in WT neurospheres driven by transduction with the plB_miR-219 lentivirus. Real-Time qRT-PCR results are expressed as fold-change of miR-219 levels in WT cells transduced with plB_miR-219 versus plB_cntr lentiviruses. Data are reported as the mean ± standard error (SEM) from three independent experiments (n = 3). (B,C): MiR-219 overexpression increases the number of O4+ cells and their differentiation. WT neurospheres were transduced with plB_cntr or plB_miR-219 lentiviruses (GFP positive) and differentiated in vitro. Data are expressed as the mean ± standard error (SEM) (B) Right panel: representative images of O4 immunostaining (in red) and DAPI (in blue). Scale bar 20 µm, 40× magnification. The inset in each panel represents a 1.5 magnification of transduced GFP-positive cells. Branch length is highlighted with arrows (short branches) or arrowheads (long branches). Right panel; histogram reporting the percentage of O4+ cells in transduced WT cultures (WT_plB_cntr n = 573; WT_plB_miR-219 n = 423). (C) Histograms reporting the number and length of O4+ branches in transduced WT cultures (n ≥ 20 O4+ cells for each condition). (A–C) Statistical significance is calculated relative to the WT_plB_cntr sample. Student’s t-test: * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 7

Modulation of miR-219 is NGF-dependent. (A) NGF neutralization in WT cells, by treatment with the anti-NGF αD11 antibody, induces a significant increase in miR-219 levels, which is not observed in control IgG-treated cells. The histogram reports qRT-PCR results of miR-219 levels measured in WT cells treated with αD11 or IgG as control (n = 3). (B): Representative immunofluorescence images for O4, showing enhanced differentiation of anti-NGF-treated WT oligodendrocytes (αD11) compared to IgG-treated WT oligodendrocytes, in terms of number and length of branches, as quantified in the histograms (lower panel). Nuclei labeled with DAPI are in blue. Scale bar, 10 µm, 20× magnification. (A,B) Statistical significance is calculated relative to the WT + IgG sample. Data are expressed as the mean ± standard error (SEM) (C): NGF treatment (45 min. long) of AD4 progenitors induces a down-regulation of miR-219 levels. qRT-PCR results are reported as ratio between miR-219 levels measured in treated versus untreated AD4 cells (n = 4). (D): Chronic NGF treatment reduces OL differentiation. Immunofluorescence for O4 shows less differentiated OL in NGF-treated AD4 cultures, compared to untreated cells. Scale bar, 10 µm, 20× magnification. The results were quantified (lower panel) both in terms of branch number and length (n≥ 20 O4+ cells for each condition). (C,D) Statistical significance is calculated relative to the AD4 untreated sample. The data are expressed as the mean ± standard error (SEM)). Student’s t-test: * p < 0.05, ** p < 0.01. (B,D). Branch length is highlighted with arrows (short branches) or arrowheads (long branches).

Figure 8

NGF neutralization increases OLs differentiation of primary OPC cultures. (A): miR-219 up-regulation by qRT-PCR in rat OPC cultures upon in vitro differentiation and αD11 antibody treatment (100 ng/mL), compared to IgG-treated cells. Statistical significance is calculated relative to the OPC undifferentiated sample. Data are expressed as the mean ± standard error (SEM) (n = 3). (B): immunostaining for MBP (red) and O4 (green) of rat OPC cultures upon in vitro differentiation. αD11 antibody treatment gives rise to more O4+ pre_OL (green cells) and O4 + MBP+ mature OLs (green and red cells), as quantified in the histogram (lower panel). Nuclei labeled with DAPI are in blue. Scale bar 10 µm, 40× magnification. Statistical significance is calculated relative to the untreated sample. The data are expressed as the mean ± standard error (SEM) (OL n =252; OL + αD11 n = 186). Student’s t-test: * p < 0.05, ** p < 0.01.