(A-D) Each clone from bisulfite sequencing data from autism or control females was analyzed for the number of CpG sites methylated in each clone (from data in Supplementary Figure 1). Histograms represent the distribution of percentage of clones based on degree of methylation. The bimodal peaks observed at all three loci are consistent with X chromosome inactivation. Clones with no sites methylated are designated as clear active alleles (Xa), clones with >50% of maximum number of methylated sites are designated as inactive (Xi), and the intervening clones as potentially undefined XCI status (X?). A) MECP2 region I has a maximum of 14 out of 15 possible methylation sites and control females show a clear bimodal peak between Xa and Xi, although there was an inherent bias towards Xi clones (7-14 sites methylated) for this highly CpG dense promoter. B) MECP2 region II has 6 possible methylation sites and more clones within the undefined XCI (X?) category (1-3 sites methylated). C) The control X linked gene AR has a maximum of 9 out of 10 possible methylation sites. Although most clones from AR fell clearly into either Xa or Xi categories, the number of methylated sites on the inactive X (Xi) was variable, ranging from 5-9 sites and no clones showing complete methylation of all 10 sites. D) The total percentage of undefined XCI (X?) clones for autism and control females is graphed for each of the three loci. Interestingly, both MECP2 loci (but not AR) show an increase in X? clones for autism compared to control females, suggesting aberrant methylation of MECP2 in females with autism. E-F) To examine the possibility of site specificity of methylation in clones showing undefined XCI status, the frequency of methylation of each possible CpG site is graphed for Region I (E) and Region II (F) only for X? clones.

A) Schematic of MECP2 promoter regions analyzed by bisulfite sequencing. Vertical lines indicate CpG sites, designated as A-F for bisulfite region II and 1-15 for bisulfite region I. The hatched bar depicts the location of the boundary between methylated and unmethylated regions on the active X chromosome in males and females. Methylation sites are shown as filled circles and the filled circle with a “A” represents an autism-specific methylation site previously described (Nagarajan et al, 2006). B) The total percent methylation of all sites in bisulfite region II are graphed for each brain sample, designated as control males, autism males, control females and autism females. MECP2 region II shows a high degree of variable methylation independent of age, gender, or clinical status. C) The total percent methylation for MECP2 bisulfite region II is shown with each category grouped, showing an only modest nonsignificant increased methylation in females compared to males and no significant difference between autism and control. D) Male brain samples were compared for percent methylation of each CpG site within MECP2 bisulfite region II (A-F). While no significant differences were observed between autism and control males in the highly methylated sites A-D, there was an abrupt transition observed between sites D and E from a highly methylated region to a mostly unmethylated region at sites E and F. Control and autism brain samples showed significantly higher % DNA methylation of combined sites A-D compared to sites E-F by t-test (P<0.0001). Interestingly, autism males showed a trend of increased methylation (P=0.07) at site F which borders the sites in bisulfite region I previously shown to have increased methylation in autism males (A and Nagarajan et al, 2006). E) ChIP using chromatin isolated from SH-SY5Y and a specific anti-CTCF antibody or nonspecific antibody (IgG control). “Total DNA” represents non-manipulated genomic DNA from SH-SY5Y cells, while “input” is the DNA sample isolated from chromatin but without immunoprecipitation. The anti-CTCF (but not IgG control) showed specific enrichment of MECP2 (-595 to -389) over input. F) Female brain samples were compared for percent methylation of each CpG site within MECP2 bisulfite region II (A-F). Similar to male samples, no significant differences were observed between autism and control females, and sites A-D were methylated at higher than the expected 50% methylation (hatched line), demonstrating partial methylation of the active allele, as in males. Similar to male samples, sites A-D combined had significantly higher % methylation than sites E-F combined for female control and autism brain samples (P<0.0001).

X-axes of histograms are binned into 10% intervals, showing the percent cells with the small allele inactivated. No significant differences were observed between groups (Fisher’s exact test).