Figure 2.

DNA sequence determinants of cytosine methylation-independent binding by MeCP2. (A) EMSAs using varying amounts of MeCP2[1–205] or no protein (−) with double-stranded oligonucleotide probes containing CAC, mCAC, GGTGT, GT₁, GT₂, GT₃, GT₄ or GT₅ (see Table 1) to assess the influence of GT-length on binding. Note that CAC is the complement of GTG, indicating that this trinucleotide motif does not bind MeCP2[1–205]. (B) EMSAs in which MeCP2[1–205] was incubated with oligonucleotide probes altered at either the first (XTGT), third (GTXT), or fourth (GTGX) positions in the presence of varying amounts of MeCP2[1–205] or no protein (−). Probes containing GT₁ or GT₃ were used as negative and positive controls, respectively. (C) Determination of the effect of DNA sequences flanking GTGT binding to MeCP2. EMSAs using varying amounts of MeCP2[1–205] or no protein (−) with probes containing GT₁, GT₂, GT₃ or GT₂-inv bot (GT₂-inverted indicated by yellow highlighting and inverted arrow). (D) The effect of altering the base composition of the 3′-AT-flank adjacent to GTGT was assayed by EMSAs using varying amounts of MeCP2[1–205] or no protein (−) with probes containing GT₁, GT₂, GT₃ or GT₂-A/T mut in which the TT-, AA- and AA-dinucleotides 3′-adjacent to GTGT were substituted by CG (yellow highlight).