Gen-Expressionsanalyse aus autoptischen Formalin-fixierten und Paraffin- eingebetteten Gewebeproben von Multiple Sklerose Patienten

A protocol for RNA isolation from FFPE brain tissue was introduced
and optimized in the laboratory. It was demonstrated that both, RNA
yield and the ratio of light absorption at 260 nm vs. 280 nm (OD
260/280) in FFPE tissue are comparable to frozen tissue (23). A
total of 27 archival brain specimens of 11 MS donors obtained from
different brain banks were screened for the ability to amplify the
housekeeping gene PPIA as well as miRNA 181a and miR 124. Results
were compared to amplification of the same transcripts in 9 frozen
MS tissue samples of 9 MS patients. The ability to amplify PPIA in
FFPE tissue specimens was very heterogeneously distributed and the
loss of amplifiable transcript copies ranged from 45 fold to 200
000 fold as compared to frozen tissue. In some archival samples
PPIA could not be detected at all. These specimens were considered
not suitable for further qPCR analysis. In contrast, the
amplification ofmiRNA 181a and miR 124 in FFPE tissue was
tremendously stable with an average loss of amplifiability of 1.7
fold only (23). Among several factors which possibly have an
influence on impaired transcript amplification in FFPE tissue, the
effect of length of formalin fixation was investigated in more
detail. It was shown that duration of formalin fixation had great
impact on loss of subsequent amplification of coding transcripts
(e. g. PPIA). Compared to frozen tissue, PPIA amplification was
reduced by ~15 fold in samples which were formalin-fixed for a
day-long period, which is in contrast to a reduction of PPIA
amplification by ~200 fold in specimens which had been fixed for
years (23). Here again, miRNA amplification was demonstrated to be
remarkably stable in the same FFPE tissue samples (23). Based on
the stable miRNA detection in FFPE tissue specimens, 18 FFPE tissue
specimens (MS n=13, healthy donor n=5) were included in a study
which compared the miRNA expression pattern in MS lesions to
healthy brain tissue by qPCR analysis of 365 mature miRNAs (42).
Furthermore, an experimental setup was established which allows for
precise dissection of MS lesions from surrounding normal appearing
white matter (NAWM). To this end, FFPE sections were obtained using
a microtome, were flattened in a DEPC water bath and mounted on PEN
membrane coated slides. RNA yield and amplification of PPIA were
not altered by this approach. Parallel tissue sections were stained
with Luxol Fast Blue (LFB) and served as a model to help with the
precise dissection of MS lesions. This setup was applied to 5 FFPE
tissue samples (MS lesion n=3, healthy donor n=2). RNA was isolated
from the dissected tissue specimens to analyse differential
expression of 84 extracellular matrix (ECM) related genes in MS
lesions compared to healthy tissue using TaqMan Low Density Array
qPCR technology. This was compared to a data set derived from
frozen tissue samples that had been processed in a similar way.
Detection of gene regulation (MS/healthy) in FFPE tissue was found
to be reliable and comparable to frozen tissue, provided that the
selected genes were of sufficient abundance (23). The up-regulation
of the extracellular matrix component decorin could be validated on
protein level by immuno-histochemistry in the same FFPE MS lesions.
This result was published as part of a study which investigated the
expression of several extracellular matrix related genes in MS
lesions with frozen tissue, e.g. collagens and the protein biglycan
(61). Furthermore this study showed that fibrillar collagens,
biglycan and decorin are part of the perivascular fibrosis. These
molecules are expressed inproximity to tissue invading immune
cells, therefore suggesting a possible disease modifying function
(61). In summary, this work presents a detailed protocol for the
use of autoptic FFPE tissue specimens to obtain gene expression
profiles from dissected MS lesions (23). This protocol was
implemented as part of a study which investigated alterations of
ECM in MS lesions (61) and contributed to obtain the first miRNA
profile in MS lesions (42).