Background
Epigenetic signatures are highly cell type specific. Separation of distinct cell populations is therefore desirable for all epigenetic studies. However, to date little information is available on whether separation protocols might influence epigenetic and/or gene expression signatures and hence might be less beneficial. We investigated the influence of two frequently used protocols to isolate intestinal epithelium cells (IECs) from 6 healthy individuals. Materials and
Conclusions
Taken together, our data highlight the importance of considering the potential effect of cell separation on gene expression as well as DNA methylation signatures. The decision to separate tissue samples will therefore depend on study design and specific separation protocols.
Methods
Epithelial cells were isolated from small bowel (i.e. terminal ileum) biopsies using EDTA/DTT and enzymatic release followed by magnetic bead sorting via EPCAM labeled microbeads. Effects on gene/mRNA expression were analyzed using a real time PCR based expression array. DNA methylation was assessed by pyrosequencing of bisulfite converted DNA and methylated DNA immunoprecipitation (MeDIP).
Results
While cell purity was >95% using both cell separation approaches, gene expression analysis revealed significantly higher mRNA levels of several inflammatory genes in EDTA/DTT when compared to enzymatically released cells. In contrast, DNA methylation of selected genes was less variable and only revealed subtle differences. Comparison of DNA methylation of the epithelial cell marker EPCAM in unseparated whole biopsy samples with separated epithelium (i.e. EPCAM positive and negative fraction) demonstrated significant differences in DNA methylation between all three tissue fractions indicating cell type specific methylation patterns can be masked in unseparated tissue samples. Conclusions: Taken together, our data highlight the importance of considering the potential effect of cell separation on gene expression as well as DNA methylation signatures. The decision to separate tissue samples will therefore depend on study design and specific separation protocols.