3. Results and discussion
3.1. Validation and determination of the most suitable miRNA
normalizers
Applying the geNorm algorithm, the following stability rank
order was obtained: miR92 < miR191 < miR374 < miR484 <
miR423 < RNU48. Due to missing data, candidates miR26b,
RNU24, RNU44 and RNU47 were automatically excluded from
analysis (no detectable expression of these targets across the
tested sample set). After changing from SYBR1 Green to TaqMan1,
miR191 was excluded due to chemistry related differences. With
M-values of 0.545 and 0.658, respectively, candidates miR92 and
miR374 were determined to be the most stable expressed genes
and thus used as normalizers for following experiments.
3.2. Impact assessment of using validated or non-validated
normalizers on miRNA target expression
The choice of endogenous controls for normalization influenced
the relative quantity of examined markers. Regarding the bloodspecific
marker miR16, it was clearly possible to differentiate blood
(mean NRQ: 276.28) from the other body
fluids and skin (mean
NRQs < 1) using U6B as normalizer. In comparison, it was not
possible to distinguish blood cells from the other cell types when
miR92 and miR374 were applied as normalizers. All mean NRQs
scattered in a range between 0.19 and 2.52. Results for miR451 also
showed the possibility for an unambiguous identification of blood,
applying both U6B and the previously validated references. Solely,
the expression of blood compared to all other samples was much higher using U6B (mean NRQ of blood: 2384.53). Applying
miR92 and miR374, blood samples showed a mean NRQ of 32.32.
Regarding both semen markers and skin marker miR203, an
unambiguous identification of the cell type’s origin was only
possible if both validated endogenous controls were used as
normalizers. Data of semen-specific marker miR135b generally
showed higher expressions when using miR92 and miR374 to
correct the relative quantities measured (Fig. 1). The NRQs for
blood clustered around 0.005, while the values for the other cell
types showed higher expressions (highest mean NRQ of 19.95 in
semen samples). This indicates that miR135b is not specifically
expressed in semen but rather in epithelial cells. Data for skinspecific
marker miR203 revealed similar results showing high
expression values in semen, saliva and skin samples. In comparison,
NRQs of all sample types clustered in a narrow range when
data were corrected with U6B. In these cases, a positive cell type
identification was not possible. Both saliva markers (miR205 and
miR658) did not give any results probably due to technical
problems.
4. Conclusion