S with PE (Figure 4a). In addition, we further tested the correlation between the DNA methylation and gene expression data, the result showed that LEP expression and differential LEP methylation was negatively correlated, but did not reach significant 22948146 level (Figure S1).Figure 3. Schematic representation of LEP and SH3PXD2A gene. (A) Maps of the TSS and the proximal promoter region of the human LEP gene. Positions and orientation of the MassARRAY primers are indicated by black arrows. (B) Maps of SH3PXD2A locus 
presenting the CGIs region in the promoter region (CGI71) and in the gene body (CGI74, CGI18 and CGI34). Positions and orientation of the MassARRAY primers are indicated by black arrows. doi:10.1371/journal.pone.0059753.gUpregulation and Hypomethylation of Genes in PEFigure 4. Quantitative DNA methylation of LEP and SH3PXD2A in all PE (n = 16) and control (n = 16) samples. (A) The average methylation level of the CGI of LEP. Several CpG sites in the vicinity of TSS (unit 34) are all significantly hypomethylated in the preeclamptic placentas. Position with significantly differential methylation between PE and control groups is marked by asterisks. **p,0.01, *p,0.05. (B) The average methylation level of CGIs of SH3PXD2A. The TSS CGI comprised the CpG sites from unit 1 to13; The CGI74 comprised the CpG sites from unit 14 to 30; The CGI18 comprised the CpG sites from unit 31 to 39; The CGI34 comprised the CpG sites from unit 40 to 54. Position with significantly differential methylation between PE and control groups is marked by asterisks. **p,0.01, *p,0.05. doi:10.1371/journal.pone.0059753.gIn the SH3PXD2A gene, the CpG sites of the promoter region (amplicon 1) in both preeclamptic Pentagastrin placentas and normal placentas are almost unmethylated with no statistical difference, which is in line with the hypothesis that the promoter CGI generally remains unmethylated. The methylation patterns in the gene body of SH3PXD2A (from unit 14 
to 54) were shown in Figure 4b. All the CpG sites in CGI74 region (from unit 14 to 30) were also at a low methylation with no statistical difference between preeclamptic and control placentas. One CpG site (unit 34) in CGI 18 (from unit 31 to 39) were significantly hypomethylaed in pathological placentas than that in normal placentas (the average methylation = 0.093, 0.203 in PE and control respectively, p = 0.002). The remaining CpG sites in the CGI18 were also at a low methylation level. Interestingly, the CpG sites in the CGI34 region (Figure 4b, from unit 40 to 54) as a whole were at a high methylation level (the mean methylation = 0.725, 0.616 in PE and control respectively, p = 1.5761024), most CpG sites are higher methylated in preeclamptic placentas than those in controls except unit 40 (including CpG 1 and CpG2), of which, unit 41, 42, 43, 44, 45, 47, 48, 49, 50, 51, 52, 53 and 54 were statistically significant (p = 0.003, 0.006, 0.007, 0.003, 1.9061024, 1.2961024, 0.004, 0.002, 0.007, 0.002, 1.2961024, 0.003 and 4.6261025 respectively, Figure 4b). The correlation between DNA methylation levelin CGI34 region and gene expression of SH3PXD2A in normal placentas was also performed. The result showed that gene body methylation and gene expression of SH3PXD2A was LY2409021 cost positively correlated without significance (Figure S2). Clinical characteristic gestational age presented significant differences between groups, which might have influence on the statistical result. To adjust this, covariance analysis was performed. O.S with PE (Figure 4a). In addition, we further tested the correlation between the DNA methylation and gene expression data, the result showed that LEP expression and differential LEP methylation was negatively correlated, but did not reach significant 22948146 level (Figure S1).Figure 3. Schematic representation of LEP and SH3PXD2A gene. (A) Maps of the TSS and the proximal promoter region of the human LEP gene. Positions and orientation of the MassARRAY primers are indicated by black arrows. (B) Maps of SH3PXD2A locus presenting the CGIs region in the promoter region (CGI71) and in the gene body (CGI74, CGI18 and CGI34). Positions and orientation of the MassARRAY primers are indicated by black arrows. doi:10.1371/journal.pone.0059753.gUpregulation and Hypomethylation of Genes in PEFigure 4. Quantitative DNA methylation of LEP and SH3PXD2A in all PE (n = 16) and control (n = 16) samples. (A) The average methylation level of the CGI of LEP. Several CpG sites in the vicinity of TSS (unit 34) are all significantly hypomethylated in the preeclamptic placentas. Position with significantly differential methylation between PE and control groups is marked by asterisks. **p,0.01, *p,0.05. (B) The average methylation level of CGIs of SH3PXD2A. The TSS CGI comprised the CpG sites from unit 1 to13; The CGI74 comprised the CpG sites from unit 14 to 30; The CGI18 comprised the CpG sites from unit 31 to 39; The CGI34 comprised the CpG sites from unit 40 to 54. Position with significantly differential methylation between PE and control groups is marked by asterisks. **p,0.01, *p,0.05. doi:10.1371/journal.pone.0059753.gIn the SH3PXD2A gene, the CpG sites of the promoter region (amplicon 1) in both preeclamptic placentas and normal placentas are almost unmethylated with no statistical difference, which is in line with the hypothesis that the promoter CGI generally remains unmethylated. The methylation patterns in the gene body of SH3PXD2A (from unit 14 to 54) were shown in Figure 4b. All the CpG sites in CGI74 region (from unit 14 to 30) were also at a low methylation with no statistical difference between preeclamptic and control placentas. One CpG site (unit 34) in CGI 18 (from unit 31 to 39) were significantly hypomethylaed in pathological placentas than that in normal placentas (the average methylation = 0.093, 0.203 in PE and control respectively, p = 0.002). The remaining CpG sites in the CGI18 were also at a low methylation level. Interestingly, the CpG sites in the CGI34 region (Figure 4b, from unit 40 to 54) as a whole were at a high methylation level (the mean methylation = 0.725, 0.616 in PE and control respectively, p = 1.5761024), most CpG sites are higher methylated in preeclamptic placentas than those in controls except unit 40 (including CpG 1 and CpG2), of which, unit 41, 42, 43, 44, 45, 47, 48, 49, 50, 51, 52, 53 and 54 were statistically significant (p = 0.003, 0.006, 0.007, 0.003, 1.9061024, 1.2961024, 0.004, 0.002, 0.007, 0.002, 1.2961024, 0.003 and 4.6261025 respectively, Figure 4b). The correlation between DNA methylation levelin CGI34 region and gene expression of SH3PXD2A in normal placentas was also performed. The result showed that gene body methylation and gene expression of SH3PXD2A was positively correlated without significance (Figure S2). Clinical characteristic gestational age presented significant differences between groups, which might have influence on the statistical result. To adjust this, covariance analysis was performed. O.