Page 178 - scheppingen

seven
176
sion. By the addition of methyl groups to the DNA, activity of a DNA segment is typi- cally repressed, however methylation patterns are dynamic and effects are dependent on genomic context 9. Recently, it was found that gene expression of pro-inflammatory cytokines can be epigenetically modulated by DNA methylation of the promotor region 10-12. In accordance with that, we identified hypomethylation of the IL-1β promotor region in TSC tubers, which correlated with IL-1β overexpression in tubers 13. In addition, we recently determined the genome-wide DNA methylation profile of TSC tubers (n=16) compared to controls (n=13). Amongst the most differentially hypomethylated CpG sites (selected for promotor region, with a beta value cut-off 0.2 and adjusted p-value<0.01) we found that a major part of the genes were inflammation related. Amongst these were Toll-like receptor 6 (TLR6), interleukin 6 signal transducer (IL6ST), interleukin 18 (IL-18), interleukin 21 receptor (IL21R), all genes involved in pro-inflammatory signaling. Although these recent findings still need to be validated, this indicates that epigenetic regulation of gene expression may underlie the inflammatory changes we see in TSC tubers.
Histological changes in cortical tubers in grey matter compared to white matter
In TSC tubers, several different pathological mechanisms are described for the cortex (grey matter, GM), and the white matter (WM). In the GM, a severe disturbance of cor- tical layering and blurred grey-white matter border is evident, while in the WM, most notably hypomyelination is observed 14-17. Currently, we are expanding our cohort for TSC histological analysis. Not only do we want to validate our findings described in Chapter 2 in a larger set of patient material collected at different centers throughout Europe, we also aim to make comparisons between immunopathological findings observed in the cortex (GM) and in the WM. We collected in total 49 samples of cortical tubers including 9 perituberal samples, and 17 control samples, on which we performed several immunohistochemical stainings which we subsequently digitalized (for details about the methods, see Chapter 2). In each slide, we manually selected representative regions of interest (ROI) of comparable size in both the GM and WM, in which the percentage of stained area, or the count of stained cells, was semi-automatically quantified using an open source image processing software package (FIJI ImageJ2, v64) with a specifically developed macro based on Java programming language. Preliminary data shows that GFAP expression, indicating gliosis, is increased in the GM of TSC cortical tubers com- pared to perituberal brain tissue (Fig 2A), and that GFAP expression is most prominent in type B tubers (Fig 2B). In the WM, GFAP expression did not differ overall between the tuber and the perituberal tissue (Fig 2C), nor did it differ between tuber types (Fig 2D), indicating that gliosis was most evident in the GM in TSC tubers. Additionally, we found a strong decrease in the GM oligodendrocyte cell count in all tuber types and in the peritu- beral cortex. This corresponds with earlier findings reporting WM deficiency in TSC and FCD 18. It was recently reported that these deficiencies are probably due to decreased oligodendroglia content caused by delayed maturation of precursor cells, overall lead- ing to hypomyelination 19. In accordance with these findings, it was also reported that neuron-specific TSC1 knockout in a mouse model of TSC deregulated the secretion of connective tissue growth factor, which in turn inhibited oligodendrocyte development and decreased the total number of oligodendrocytes 20. This indicates the importance between mTOR signaling in neuronal cells and their paracrine effects on oligodendro-