15283-B-Blokker

127 PM tissue biopsies obtained at MIA: an RNA-quality analysis 7 The morphological factors necrosis and autolysis had an adverse effect on GAPDH expression, with P -values ranging from <0.001 to 0.008 depending on amplicon sizes (results not shown). The GAPDH expression differed between patients who had dyslipidemia or a BMI higher than 25 and those without these symptoms (P ranges 0.001 to 0.008 and <0.001 to 0.015 depending on amplicon size), for patients with and without hypertension only the three longest GAPDH amplicons differed significantly (P <0.001, Paired Kruskal-Wallis test, significance level 0,00833), whereas hypoxia, fever and diabetes did not influence GAPDH expression. DISCUSSION AND CONCLUSION RIN values were significantly higher in fresh frozen samples than in post-mortem samples. MIA samples showed higher RIN values than CA samples. Also, GAPDH expression was significantly higher in fresh frozen samples than in post-mortem samples. There were differences in GAPDH expression between MIA and CA samples, but they were significant for only 3 out of 6 GAPDH amplicon sizes. Since GAPDH amplicon sizes of up to 530 base pairs could be detected well within the limits of the assay’s sensitivity, we conclude that the post-mortem samples still contained RNA of reasonable quality. It is generally believed that post-mortem tissues with poor RNA integrity values (RIN value <5) are not suitable for molecular techniques. 195 In this study we showed that samples with low RIN values could still be used for determining gene expression with RT-qPCR. The Affymetrix and Illumina gene array platforms respectively use 25-mer or 50- mer probes to detect gene expression. Since it is possible to amplify 530 bp GAPDH RNA fragments, this implies that detection of GAPDH with gene array technology must be possible. GAPDH expression in post-mortem tissue is 4 Cq, approximately 16-fold lower than in fresh frozen tissue (see figure 2). Since GAPDH is an abundantly expressed housekeeping gene (i.e. high copy number per cell), the transcript could still be measured in post-mortem tissue derived RNA with qPCR. Assuming post-mortem RNA degradation is a random process, all transcripts (i.e. the entire transcriptome) will be subjected to degradation to the same degree. 196 Therefore, less abundantly expressed genes (i.e. low copy number per cell) may become undetectable after this rate of degradation. Nonetheless, successful gene array analysis of post-mortem heart tissue was previously described. 182 Recently, Romero et al. published that even with low RNA quality RNA sequencing is possible, as long as the RIN values are accounted for during data analysis. 197