W significant gene-specificTuran et al. BMC Medical Genomics 2012, 5:10 http://www.biomedcentral.
W significant gene-specificTuran et al. BMC Medical Genomics 2012, 5:10 http://www.biomedcentral.com/1755-8794/5/Page 18 ofvariation in methylation levels (e.g., B cells vs. CD4 T-cells vs. CD8 T-cells in Figure 1 in Rakyan et al. 2008) longitudinal measures of site-specific DNA methylation in total lymphocytes taken from the same individuals, decades apart, rarely change by more than a few percent [59-61]. We have also examined the effect of inflammatory markers (erythrocyte sedimentation rate and levels of C-reactive protein) likely to be associated with specific leukocyte subpopulations, as well as total white blood cell count in longitudinal studies of 111 individuals [61] and none of these parameters was related to any methylation differences observed [61]. Similarly, in terms of placental subpopulations, we have compared DNA methylation levels at the IGF2/H19 and IGF2R DMRs in five section of placenta both within and between individuals. Although there is some variation within a placenta, there is substantially more variation between individuals than within an individual [27]. These observations suggest that intra-individual variation in placental or cord blood DNA methylation are unlikely to change the correlations observed between candidate gene methylation and birth weight.Candidate gene interaction may identify novel regulatory networks and provide links between low birth weight and adult diseaseOf the birth weight-associated candidate gene DNA methylation differences identified in the L1 procedure(Table 3), three are of particular interest. Methylation levels of the homeobox transcriptional repressor MSX1 in cord blood are correlated with the transcript level of four of the other candidate genes (APOE, ATP6AP1, PRSS21 and RCOR1 (Table 5)). In fact, at least seven of the top 10 genes whose transcript level is correlated with methylation of CpG sites in MSX1 (Table 9) are suspected to play roles in fetal or placental growth. On the placental side, methylation levels of multiple sites in CDK2 are correlated with expression of three of the other candidates and multiple CpG sites in CDK2 are correlated with transcript levels of GRB10 (Table 5). Methylation levels of multiple sites in GRB10 are correlated with transcript levels of four of the seven candidates (CDK2, GRB10, Pemafibrate chemical information PGRMC1 and RGS14), including itself (Table 5), and two of the genes PubMed ID:https://www.ncbi.nlm.nih.gov/pubmed/28499442 in the top ten GRB10 transcript level correlations (Table 9) have been found to have an effect on growth. The mechanisms linking low birth weight to adverse long-term health outcomes are not well understood but may be related to defective placentation, restrictions in the size of stem cell populations that lead to reduced organ size and function, and/or abnormal programming of metabolic pathways including glucose utilization. In this regard, it is noteworthy that methylation levels of three CpGs in the MSX1 transcriptional repressor are correlated with transcript levels of the glucose transporter SLC2A3 (Pearson correlation coefficient 0.42).Table 9 Top ten genes whose transcript levels are correlated with methylation of CpG sites in MSX1, CDK2 and GRBTissue Blood Methylation Gene MSX1 CpG ID Expression Gene Transcript ID ILMN_1740938 ILMN_1734176 ILMN_1754576 ILMN_1721770 ILMN_1693397 ILMN_2347145 ILMN_2309615 ILMN_1693617 ILMN_1659354 ILMN_1726266 ILMN_2067890 ILMN_1749070 ILMN_1745356 ILMN_1771385 ILMN_2114568 ILMN_1678841 ILMN_1683872 ILMN_1717261 ILMN_2261416 ILMN_1681882 ILMN_1681777 Correl.