Measuring cell-type specific differential methylation in human brain tissue - PubMed (original) (raw)

Measuring cell-type specific differential methylation in human brain tissue

Carolina M Montaño et al. Genome Biol. 2013.

Abstract

The behavior of epigenetic mechanisms in the brain is obscured by tissue heterogeneity and disease-related histological changes. Not accounting for these confounders leads to biased results. We develop a statistical methodology that estimates and adjusts for celltype composition by decomposing neuronal and non-neuronal differential signal. This method provides a conceptual framework for deconvolving heterogeneous epigenetic data from postmortem brain studies. We apply it to find cell-specific differentially methylated regions between prefrontal cortex and hippocampus. We demonstrate the utility of the method on both Infinium 450k and CHARM data.

PubMed Disclaimer

Figures

Figure 1

Figure 1

The proportion of neuronal cells in a given brain region influences the identification of differentially methylated regions. (a) Whole-tissue methylation signals show false-positive brain-region differences. Panel shows a plot of smoothed methylation signals from sorted neuronal and glial cells (teal and purple lines) from DLPFC and HF (solid and dashed lines) as well as from whole-tissue DLPFC (gold line) and HF (grey line). (b) Estimated neuronal fraction of cells for whole-tissue samples differs between DLPFC and HF (mean DLPFC = 0.53 (n = 19), mean HF = 0.30 (n = 13), two-sample t-test P value 6.3 × 10-6). (c) Estimated neuronal fraction of cells for whole-tissue samples using universal DMRs vs. estimated neuronal fraction using brain region-specific DMRs from DLPFC (gold), HF (grey), and STG (blue).

Figure 2

Figure 2

Effects of direct modeling on false-positives and accuracy. (a) Explicit modeling for differences in cell type reduces false-positive rate. Boxplots of test statistics for the difference in means based on linear regression estimation from models M1 and M2. Eighty percent of regions from M1 show a statistically significant difference in overall mean (at level 0.05); 16% and 12% of regions from M2 show a statistically significant difference in neurons or glia, respectively (at level 0.05). (b) Explicit modeling of neuronal methylation differences improves estimation accuracy. Comparison of gold-standard mean difference in methylation in neuron-specific DMRs to the estimated mean difference from models M1 (left) and M2 (right), along with the linear regression fit to the data (95% CI for the slope of the regression line of M1 = (0.29, 0.44), for M2 = (0.68, 0.95).

References

    1. Feinberg AP. Phenotypic plasticity and the epigenetics of human disease. Nature. 2007;14:433–440. - PubMed
    1. Miller CA, Sweatt JD. Covalent modification of DNA regulates memory formation. Neuron. 2007;14:857–869. - PubMed
    1. Feng J, Zhou Y, Campbell SL, Le T, Li E, Sweatt JD, Silva AJ, Fan G. Dnmt1 and Dnmt3a maintain DNA methylation and regulate synaptic function in adult forebrain neurons. Nat Neurosci. 2010;14:423–430. - PMC - PubMed
    1. LaPlant Q, Vialou V, Covington HE, Dumitriu D, Feng J, Warren BL, Maze I, Dietz DM, Watts EL, Iniguez SD, Koo JW, Mouzon E, Renthal W, Hollis F, Wang H, Noonan MA, Ren Y, Eisch AJ, Bolanos CA, Kabbaj M, Xiao G, Neve RL, Hurd YL, Oosting RS, Fan G, Morrison JH, Nestrel EJ. Dnmt3a regulates emotional behavior and spine plasticity in the nucleus accumbens. Nat Neurosci. 2010;14:1137–1143. - PMC - PubMed
    1. Herb BR, Wolschin F, Hansen KD, Aryee MJ, Langmead B, Irizarry R, Amdam GV, Feinberg AP. Reversible switching between epigenetic states in honeybee behavioral subcastes. Nat Neurosci. 2012;14:1371–1373. - PMC - PubMed

Publication types

MeSH terms

Grants and funding

LinkOut - more resources