Testing the PEST hypothesis using relevant Rett mutations in MeCP2 E1 and E2 isoforms

Abstract

Mutations in methyl-CpG binding protein 2 (MeCP2), such as the T158M, P152R, R294X, and R306C mutations, are responsible for most Rett syndrome (RTT) cases. These mutations often result in altered protein expression that appears to correlate with changes in the nuclear size; however, the molecular details of these observations are poorly understood. Using a C2C12 cellular system expressing human MeCP2-E1 isoform as well as mouse models expressing these mutations, we show that T158M and P152R result in a decrease in MeCP2 protein, whereas R306C has a milder variation, and R294X resulted in an overall 2.5 to 3 fold increase. We also explored the potential involvement of the MeCP2 PEST domains in the proteasome-mediated regulation of MeCP2. Finally, we used the R294X mutant to gain further insight into the controversial competition between MeCP2 and histone H1 in the chromatin context. Interestingly, in R294X, MeCP2 E1 and E2 isoforms were differently affected, where the E1 isoform contributes to much of the overall protein increase observed, while E2 decreases by half. The modes of MeCP2 regulation, thus, appear to be differently regulated in the two isoforms.

Keywords: MeCP2; PEST sequences; Rett; chromatin; methyl CpG binding protein.

Figure 1

Comparison of MeCP2 expression levels in C2C12 cells transiently transfected with several GFP-tagged MeCP2 mutants. (A) Cartoon depicting the primary structure of MeCP2 and its constitutive NTD, MBD, ID, TRD, and CTD domains. The position of AT-hook, NID, NLS and the RNA binding (RBD) domains are indicated. A schematic representation of the different mutants and PEST-altered constructs used in this work is shown. Notice the absence of the entire NID and the CTD in the R294X mutant. (B) A representative western blot of isolated nuclei of C2C12 cells transfected with MeCP2 mutants is shown in (A). Below is a bar plot of the western blot analysis normalized by histone H4. MeCP2 antibody that binds to the Ct- epitope (green) and the GFP antibody (red) for R294X were used. (C) Bar plots, as in (B) for the MeCP2 PEST constructs, are shown in (A). Data are mean ± SEM; n = 5 biological replicates; one-way ANOVA with post hoc Dunnett's test; ^*P ≤ 0.05, ^**P ≤ 0.01; ^***P ≤ 0.001. For quantification of MeCP2, only the upper band, MeCP2-GFP, was used.

Figure 2

Ionic strength dependent DNA binding affinity of MeCP2 constructs within chromatin. (A) A representative western blot of the MeCP2 soluble fraction at different salt concentrations. (B) MeCP2 ionic-strength dependent elution patterns for the different mutants and constructs used in this work. (C) Bar plots of the data shown in (B) at three different NaCl concentrations for better data visualization and to emphasize the main differences exhibited by the different constructs. The nomenclature of the constructs is the same as in previous figures. One-way ANOVA with post hoc Dunnett’s test; n = 5 biological replicates; ^*P ≤ 0.05, ^**P ≤ 0.01; ^***P ≤ 0.001.

Figure 3

Confocal fluorescence microscopy of transiently transfected C2C12 with different MeCP2 constructs. (A) Confocal image stacks of DAPI (red), GFP (green), and merged channels. The scatter plot depicts the overlap of red and green pixels. MTV is the empty vector, pcDNA3.1-Ct-GFP. The missense mutations T158M and P152R are within the MDB, while R306C is part of the NID. The double mutants, modified PEST (mod P2) and null PEST (null P2), contain the missense mutation T158M. The nonsense mutation R294X is missing the critical NID and the entire CTD. The 10 μm scale length applies to all images. (B) Bar plot representations of Pearson correlation coefficient (PCC) of the DAPI and GFP within the nuclei for quantitative co-localization of MeCP2 to chromocenters. (C) and (D) are the bar graphs for the chromocenter size (~50 chromocenters measured) and chromocenter number of the nuclei, respectively. Data are mean ± SEM; n = 15–10 from three biological replicates. (E) Frequency of size distribution for three representative MeCP2 constructs: WT, T158M and R306C mutants. (F) A scatter plot of the different mutants' nuclear size (average diameter) distribution is shown in (A). Each point corresponds to an individual nucleus size (average diameter in μm). Data are mean +/− SEM, n = 11–15. One-way ANOVA with post-hoc Dunnett’s test. ^*P ≤ 0.05, ^**P ≤ 0.01; ^***P ≤ 0.001.

Figure 4

FRAP analysis of transfected C2C12 comparing the mobility and binding dynamics to chromocenters of different MeCP2 mutants and constructs. (A) Fluorescence recovery of the first 60 s of the GFP at the C-terminus of transfected MeCP2 after photobleaching. The blue arrow specifies the chromocenter that is targeted by the laser. The recovery for the first 60 s at 0 s, 10 s, 30 s, and 60 s is shown next to the pre-bleached full view of the nucleus. (B) FRAP recovery curves of MeCP2 constructs at the chromocenters; n = 30–32. (C). Bar graph comparisons of half recovery time of GFP in MeCP2 constructs. One-way ANOVA with post hoc Dunnett’s test. Comparisons were made only to WT and T158M. (D) Same as (C) except comparisons are made to T158M as both double mutants, nullPEST and modified PEST, contain the missense mutation T158M. Unpaired t-test. P-values for nullPEST and mPEST are < 0.0001 and 0.0045, respectively. Data are mean ± SEM; ^**P ≤ 0.01 ^***P ≤ 0.001.

Figure 5

Comparison of MeCP2 levels using in-vivo mice model. (A) Mean values of the MeCP2/β-actin levels observed in the brains of the Jaenisch (Mecp2^tm1.1Jae/y), T158M (Mecp2^T158M/y), R306C (Mecp2^R306C/y) and R294X (Mecp2^R294X/y) MeCP2 mutant mice respectively, compared to the brains of the original Mecp2^+/y mice strains used to generate the mutant mice. (B) Percentile of cytoplasmic MeCP2 was determined as in [41]. No cytoplasmic MeCP2 was detected in R294X. (C, left) SDS-PAGE of the nuclear proteins from C2C12 cells transfected with a MeCP2 R294X plasmid and (Mecp2^R294X/y) brain compared to their wild-type counterparts. CM, chicken erythrocyte histone marker. (C-right) Western blot of the gel shown on the left side, using N-terminal MeCP2, histone H1 cocktail and histone H3 antibodies. PM, protein marker (the molecular weights in kDa of the different bands are indicated). (D) Bar plots of the MeCP2/Histone H3 and histone H1/histone H3 ratios. The error bars correspond to the standard deviation obtained with two sample replicates.

Figure 6

MeCP2-E1/-E2 ratio in mice mutants. (A) Western blot, using an antibody against the N-terminus of MeCP2, of male mice brain expressing the R294X mutation. Lane 1 Mecp2^+/y and lane 2 Mecp2^R294X/y. M, protein marker. (B) Detail of similar westerns as in (A) stained with the N-terminal antibody (αN), MeCP2-E1 (αE1) and MeCP2-E2 (αE2) specific antibodies. Lane 1: Mecp2^+/y lanes 2–3: Two samples of Mecp2^R294X/y and M, protein marker. (C) Western blot using generic MeCP2 antibody (Sigma-Aldrich) for the wild-type used to generate the Mecp2^T158M/y and Mecp2^R306C/y and an n = 3 (different brains) for each is shown. (D) Bar plot representation of the MeCP2-E1 and MeCP2-E2 relative to WT in the mouse Rett models brain of the mutants. WT (TM/RC) is the same wild-type as in (C). WT (RX) is the wild-type strain used to generate Mecp2^R294X/y.