Functional consequences of Rett syndrome mutations on human MeCP2

Abstract

The neurodevelopmental disorder known as Rett syndrome has recently been linked to the methyl-CpG-binding transcriptional repressor, MeCP2. In this report we examine the consequences of these mutations on the function of MeCP2. The ability to bind specifically to methylated DNA and the transcription repression capabilities are tested, as well as the stability of proteins in vivo. We find that all missense mutations (R106W, R133C, F155S, T158M) within the methyl-binding domain impair selectivity for methylated DNA, and that all nonsense mutations (L138X, R168X, E235X, R255X, R270X, V288X, R294X) that truncate all or some of the transcriptional repression domain (TRD) affect the ability to repress transcription and have decreased levels of stability in vivo. Two missense mutations, one in the TRD (R306C) and one in the C-terminus (E397K), had no noticeable effects on MeCP2 function. Together, these results provide evidence of how Rett syndrome mutations can affect distinct functions of MeCP2 and give insight into these mutations that may contribute to the disease.

Figure 1

Schematic representation and sequence alignment of MeCP2 with Rett syndrome mutations identified. (A) MeCP2 with its two functional domains MBD and TRD indicated. Rett syndrome associated mutations found to date are marked, with missense mutations located on the top, and nonsense and deletion/insertion mutations on the bottom. Mutations used in this study are in bold type. The NLS is marked. (B) Sequence alignment of MeCP2 from human (h), mouse (m) and Xenopus (x). Conserved residues are shaded and mutations used in this study are marked by open triangles for missense mutations and closed triangles for nonsense mutations.

Figure 2

Analysis of Rett syndrome mutations of MeCP2 on its affinity for methylated DNA. MeCP2 proteins with Rett syndrome mutations were analyzed for their binding to methylated or unmethylated oligonucleotide probes by southwestern analysis. (A) Binding to methylated GAM12 probes was lost when missense or nonsense mutations were in the MBD. Missense mutations outside the MBD did not affect binding, and binding for nonsense mutations outside the MBD occurred and increased with increase in protein size. (B) Non-specific binding to unmethylated GAC12 probes was apparent for all proteins except for two nonsense mutations that resulted in the smallest MeCP2 proteins (138X and 168X). (C) Western blotting of membranes following southwestern with anti-MeCP2 antibodies shows equal amounts of proteins.

Figure 3

Analysis of Rett syndrome mutations of MeCP2 and its ability to repress transcription. MeCP2 constructs containing Rett syndrome mutations were fused to the Gal4 DBD and analyzed for their ability to repress transcription in Xenopus oocytes. (A) Schematic representation of reporter constructs used in these analyses, one containing five Gal4 upstream activation sequences (G5-HSVtk-CAT) and another without (CMV-CAT). Oocytes were injected with RNA of the Gal4–hMeCP2 constructs, followed by a mix of both reporters. RNA from the oocytes was extracted and expression of the reporters analyzed by primer extension. (B) Primer extension using primers against CAT show changes in expression from the G5-HSVtk-CAT reporter. Oocytes containing the reporters only (‘Reporter’) show basal levels of transcription. In the presence of WT MeCP2 or the TRD+C of MeCP2, expression of G5-HSVtk-CAT is repressed. All nonsense mutations in MeCP2 failed to repress transcription. Missense mutations in the TRD (R306C) or C-terminus (E397K) repressed transcription at levels comparable to WT. Activity from the CMV-CAT promoter remained unchanged in the presence of MeCP2 and was used as a control for oocyte transcription ability and injection efficiency. RNA loading and integrity was monitored using a primer against endogenous H4. (C) The ability of R306C and E397K to repress transcription was confirmed by transient transfections into human Bosc 23 cells. Gal4–MeCP2 constructs were co-transfected with G5-HSVtk-CAT and RNA was analyzed by primer extension. (D) Western analyses of oocyte or cell extracts using anti-Gal4 DBD antibodies show relative levels of Gal4–MeCP2 proteins. Results of the primer extensions were counted and represented in histograms.

Figure 4

Effects of mutations in MeCP2 on its stability in vivo. RNA from MeCP2 constructs were injected into Xenopus oocytes and allowed to translate. Oocytes were then placed in buffer containing cyclohexamide, in order to arrest protein translation, and collected at various time points after. Extracts were prepared and protein levels monitored by western blotting. (A) Western blots at indicated time points show full-length proteins (WT, R306C and E397K) have the greatest stability at 12 h, while L138X becomes largely degraded between 5 and 8 h. R294X shows moderate degradation as does the TRD+C. (B) The experiment was repeated twice and results were counted, averaged and graphed.