Rett syndrome astrocytes are abnormal and spread MeCP2 deficiency through gap junctions

Abstract

MECP2, an X-linked gene encoding the epigenetic factor methyl-CpG-binding protein-2, is mutated in Rett syndrome (RTT) and aberrantly expressed in autism. Most children affected by RTT are heterozygous Mecp2-/+ females whose brain function is impaired postnatally due to MeCP2 deficiency. While prior functional investigations of MeCP2 have focused exclusively on neurons and have concluded the absence of MeCP2 in astrocytes, here we report that astrocytes express MeCP2, and MeCP2 deficiency in astrocytes causes significant abnormalities in BDNF regulation, cytokine production, and neuronal dendritic induction, effects that may contribute to abnormal neurodevelopment. In addition, we show that the MeCP2 deficiency state can progressively spread at least in part via gap junction communications between mosaic Mecp2-/+ astrocytes in a novel non-cell-autonomous mechanism. This mechanism may lead to the pronounced loss of MeCP2 observed selectively in astrocytes in mouse Mecp2-/+ brain, which is coincident with phenotypic regression characteristic of RTT. Our results suggest that astrocytes are viable therapeutic targets for RTT and perhaps regressive forms of autism.

Figure 1.

Astrocytes express MeCP2. A, Hippocampal sections from the indicated mice were coimmunostained for MeCP2 (red nuclear stain) and the astrocytic marker GFAP (green cytoplasmic stain). Photomicrographs of corresponding CA1 region are shown in the upper row, and magnified fields from the stratum radiatum are shown in the lower row. Arrows point to examples of MeCP2+ astrocytes and dash-lined squares enclose examples of apparently MeCP2− astrocytes. B, The levels of Wt and mutant Mecp2 transcripts expressed by astrocytes derived from mice of indicated Mecp2 genotype were measured by quantitative RT-PCR. Shown are PCR products in gel. The PCR products were not amplified from genomic DNA because PCRs without reverse transcriptase (−RT control) yielded minimal products. C, Western blot analysis of cell lysates from astrocytes derived from mice of indicated Mecp2 genotype, analyzed by two distinct antibodies. D, Western blot analysis of lysates of primary neurons and primary astrocytes, to compare their levels of MeCP2 protein. The MeCP2 band in the Mecp2 ^−/+ astrocyte sample was barely detectable due to the limited ECL development to adequately compare between Wt neurons and Wt astrocytes. E, Astrocytes derived from mice of indicated Mecp2 genotypes were cultured and immunostained as in A and counterstained with DAPI (blue). Arrows point to examples of MeCP2+ astrocytes and arrowheads point to examples of MeCP2− astrocytes.

Figure 2.

MeCP2-deficient astrocytes are abnormal in regulation of Bdnf, cytokines, and p38MAPK. A, Bdnf mRNA levels were measured by quantitative RT-PCR and normalized to the expression of β-actin. Shown are PCR products in gel. B, Astrocyte conditioned media (ACM) were analyzed by Western blots for released BDNF, nerve growth factor (NGF), and Apolipoprotein E (ApoE). C, The indicated cytokines in ACM from cultures treated with LPS were measured by ELISA. n = 3, *p < 0.05 and **p < 0.001 compared with Wt. Without LPS stimulation, the basal levels of all three cytokines were minimal, and there were no significant differences between groups. D, The activation states of p38MAPK, ERK, and JNK in unstimulated astrocytes were evaluated by Western blot using antibodies for their respective phosphorylated epitopes. An antibody for p38MAPK was used to quantify the total p38MAPK level. The activation of p38MAPK is represented by the band intensity of phospho-p38MAPK normalized to that of total p38MAPK. The bar graph shows the fold increase of p38MAPK activation in Mecp2 ^−/+ and Mecp2 ^−/y astrocytes compared with Wt-As. n = 3, *p < 0.05 compared with Wt. E, The activation of p38MAPK in astrocytes following 24 h LPS stimulation. As in D, without stimulation, both Mecp2 ^−/y and Mecp2 ^−/+ astrocytes showed ∼1.5-fold increase in p38MAPK phosphorylation. After LPS stimulation, there was no difference of phosphorylation level between groups.

Figure 3.

MeCP2-deficient astrocytes support less dendritic growth from Wt neurons. E18 Wt hippocampal neurons were plated onto culture dishes without astrocytes or onto confluent monolayers of astrocytes of indicated genotype. The dendrites were allowed to grow for 24 and 72 h. A, Representative phase-contrast images of 72 h cocultures. B, Dendrites were demonstrated by immunostaining for dendritic marker MAP2 (red). Shown are representative photomicrographs of 72 h cocultures. C, Total dendritic length per neuron in each culture condition was quantified by Neurolucida and Neuroexplorer software. Dendrites from the numbers of individual neurons (indicated in the bars) over four independent experiments were imaged and analyzed. Average dendritic length was significantly different between the four groups (p < 0.001). In post hoc pairwise comparisons adjusted for multiple comparisons, the mean dendritic length induced by Wt-As was significantly higher than each of the other groups (*), but there was no significant difference between groups with Mecp2 ^−/+ and Mecp2 ^−/y. D, The 24 h dendritic length data were further stratified into average length in each branch order. E, Branch point complexity was evaluated by counting the number of branch points per order. Mecp2 ^−/+ and Mecp2 ^−/y astrocytes induced smaller dendrite length and branch point number than Wt-As in branch order 1–4 (p < 0.02), but there was no significant difference between Mecp2 ^−/+ and Mecp2 ^−/y (p = 0.1 for dendritic length and p = 0.86 for branch point number). F, The number of neurons was obtained by counting NeuN-immunoreactive nuclei and normalized by the neuron number in the “neuron-only” group in each experiment.

Figure 4.

A non-cell-autonomous effect reduces the MeCP2 expression of Wt-As in Mecp2 ^−/+ cultures. A, Western blot analysis of MeCP2 levels in cultured astrocytes with indicated time in vitro. B, GFP-labeled Wt-As (green) were cocultured with unlabeled Wt-As or Mecp2 ^−/+ astrocytes and subsequently stained with anti-MeCP2 (red) and DAPI (blue). The arrowhead in the left panel points to a GFP-labeled Wt-As that retained a high level of MeCP2 expression in the vicinity of MeCP2+ astrocytes. The arrow in the right panel points to a GFP-labeled Wt-As that lost its MeCP2 expression in the vicinity of MeCP2− astrocytes. C, MeCP2 immunoreactivities in GFP-labeled Wt-As were classified into three tiers: high, low, and none. Shown are average percentages of cells in these three tiers of MeCP2 expression. Totally ∼350 GFP-labeled cells per group in four independent experiments were analyzed. A shift in the distribution of high versus low expression (see Material and Methods) was observed between the groups (p < 0.01). D, The time-dependent changes in the MeCP2 expression of GFP-labeled Wt-As cocultured with Mecp2 ^−/+ astrocytes. n = 3.

Figure 5.

Mecp2 ^−/+ mice in the immediately presymptomatic or early symptomatic stage show a pronounced reduction of astrocytic MeCP2. Hippocampal sections from indicated mice were coimmunostained for either MeCP2 (red nuclear stain)/NeuN (green nuclear stain) (A) or MeCP2/GFAP (green cytoplasmic stain) (B). In A, arrows point to examples of MeCP2+ non-neuronal cells in the stratum radiatum, which are readily found in 7-month-old Mecp2 ^+/+ mice (two left panels) but rarely found in 7-month-old Mecp2 ^−/+ mice (two right panels). In B, arrows point to examples of MeCP2+ astrocytes and dash-lined squares enclose examples of apparently MeCP2− astrocytes. A large majority of astrocytes in 7-month-old Mecp2 ^−/+ mice appeared to be MeCP2−.

Figure 6.

The non-cell-autonomous effect between MeCP2− and MeCP2+ astrocytes is mediated at least partly by GJs. GFP-labeled Wt-As were cocultured with unlabeled Wt-As, Mecp2 ^−/+ astrocytes (−/+ A), and Mecp2 ^−/+ astrocytes with GJ inhibitors CBX and GA, respectively, and cultured for 24 h. In additional two groups, the inhibition of GJs by inhibitors was specifically prevented by tolbutamide (TB). In further two groups, the Cx-43-specific siRNA to achieve GJ inhibition and the control siRNA was used, respectively. The MeCP2 immunoreactivities in GFP-labeled Wt-As in each culture were analyzed and presented as described in Figure 4 C. The average percentage of cells with high (and therefore low) MeCP2 expression differed between the groups (p < 0.01). Post hoc pairwise comparisons, adjusted for multiple comparisons indicated that there was a significant shift from high to low expression in the following: (1) from “w/Wt-A” to “w/−/+A” to “w/−/+A+CBX+TB” and to “w/−/+A+GA+TB”; (2) from “w/−/+A+GA” to “w/−/+A” and to “w/−/+ A+GA+TB”; (3) from “w/−/+A+CBX” to “w/−/+A” and to “w/−/+ A+CBX+TB”; (4) from “w/Wt-A” to “w/−/+A+control siRNA”; and (5) from “w/−/+A+Cx43 siRNA” to “w/−/+A+control siRNA.”