A small-molecule screen reveals novel modulators of MeCP2 and X-chromosome inactivation maintenance

Abstract

Background: Rett syndrome (RTT) is a neurodevelopmental disorder caused by mutations in the X-linked methyl-CpG binding protein 2 (MeCP2) gene. While MeCP2 mutations are lethal in most males, females survive birth but show severe neurological defects. Because X-chromosome inactivation (XCI) is a random process, approximately 50% of the cells silence the wild-type (WT) copy of the MeCP2 gene. Thus, reactivating the silent WT copy of MeCP2 could provide therapeutic intervention for RTT.

Methods: Toward this goal, we screened ~ 28,000 small-molecule compounds from several libraries using a MeCP2-luciferase reporter cell line and cortical neurons from a MeCP2-EGFP mouse model. We used gain/increase of luminescence or fluorescence as a readout of MeCP2 reactivation and tested the efficacy of these drugs under different drug regimens, conditions, and cellular contexts.

Results: We identified inhibitors of the JAK/STAT pathway as XCI-reactivating agents, both by in vitro and ex vivo assays. In particular, we show that AG-490, a Janus Kinase 2 (JAK2) kinase inhibitor, and Jaki, a pan JAK/STAT inhibitor, are capable of reactivating MeCP2 from the inactive X chromosome, in different cellular contexts.

Conclusions: Our results suggest that inhibition of the JAK/STAT pathway is a new potential pathway to reinstate MeCP2 gene expression as an efficient RTT treatment.

Keywords: AG490; JAK/STAT; Janus Kinase; Janus Kinase inhibitors; MeCP2; PI3K/ATK pathways; Rett syndrome; X-chromosome inactivation.

Fig. 1

Luminescence-based screen identifies AG490 and Jaki as potential reactivators of genes on the inactive X chromosome. a Dot plot indicating the percentage of luciferase activation per drug using the Xi8 fibroblast cell line reporter for MeCP2 activation over the control (normalized values; see the “Material and methods” section). In red, we show AG490, in blue 5-Aza, and in green compounds previously reported to reactivate MeCP2 using different protocols (LY294002, OSU-03012, MLN8237, BX-912, K02288, VX-680; see the “Introduction” and the “Results” sections for more details). b Top: Chemical structure of 5-azacytidine (5-Aza), a known reactivator of MeCP2, and AG490. Bottom: Dose response of 3-day treatment of 5-Aza or AG490 in Xi8 fibroblast cells (n = 3, one representative graph shown here from three independent assays). c Left: Xi8 clone showing MeCP2 and MeCP2-luciferase genes on the inactive chromosome. Right: qRT-PCR data showing the effect of the indicated drug treatments on luciferase (Luc), Mecp2, and Xist mRNA levels; mean (± SEM) are shown, *statistical significance, p ≤ 0.05 (n = 3 per compounds, the average of three biological replicates are shown). d Left: Xa3 clone showing MeCP2 and MeCP2-luciferase genes on the active chromosome, respectively. Right: qRT-PCR data showing the effect of the indicated drug treatments on luciferase (Luc), Mecp2, and Xist mRNA levels; mean (± SEM) are shown, *statistical significance, p ≤ 0.05 (n = 3 per compounds, the average of three biological replicates are shown)

Fig. 2

AG490 effectively upregulates MeCP2 in the cultured cortical neurons. a Representative fluorescent immunostaining images of cortical neurons after treating with AG490 or DMSO vehicle twice for 3 days each. The top row shows labeling of nuclei, and the middle row shows GFP labeling (of MeCP2-GFP). The bottom row shows the zoom-in of nuclear and MeCP2-GFP labeling from the boxed area depicted in the middle row. Arrows indicate GFP-positive cells. b Quantitative analysis of the number of GFP-positive cells (**p ≤ 0.01) (n = 4 per dose, the average of three biological replicates are shown). c GFP western blot analysis (left) and WB quantification (right) show MeCP2 upregulation in AG490- and Jaki-treated cells, *statistical significance, p ≤ 0.05 (n = 3 per compounds, the average of two biological replicates are shown)

Fig. 3

AG490 upregulates MeCP2 in different cellular contexts. a qRT-PCR analysis of AG490 treatments in a humanized Xi-containing line (THX88) is shown. Dots indicate individual experimental points, samples are color-coded (see legend), and concentrations are indicated. ***Statistical significance, p ≤ 0.001, one-way ANOVA. The experiments were done using 2 (5-Aza), 3 (Jaki), and 4 (AG490) biological replicates, using 2 to 3 technical replicates per experiment. Horizontal lines displayed the median of the samples. Gapdh was used as an internal normalization control. b Survival analysis of H4SV cells in HAT-selecting medium in the presence or absence of Xi-reactivating compounds. Fold changes (FC, A.U) of the total number of colonies for control (DMSO) and Xi-reactivating drugs are shown. Bars indicate the standard deviation (SD). *Statistical significance, p ≤ 0.05. The average of three biological replicates are shown. Representative plates are shown below. c qRT-PCR data showing the effect of the indicated drug treatments on Mecp2, Hprt, and Xist mRNA levels; mean (± SD) are shown, *statistical significance, p ≤ 0.05 (n = 3 per compounds, the average of three biological replicates are shown)