Tel.+49 30 838 53555 Fax.+49 30 838-53480
In the course of the reproductive cycle, ovaries are subject to considerable changes, which result in an abundant growth of blood vessels during the formation of the corpus luteum and the regression of blood vessels during luteolysis. The process of neovascularization comprises one third of the total length of the ovarian cycle and is more vigorous than in malignant tumors.
The origin of these proliferating endothelial cells has not been completely clarified yet. The enclosed doctoral thesis supports the hypothesis that endothelial progenitor cells (EPCs) act as a local resource and thus are instrumental in the process of neovascularization in the ovaries. Therefore, the objective of this doctoral thesis was to detect the quality and quantity of endothelial progenitor cells in bovine ovaries.
By using a combination of anti-KDR and anti-CD34, it was possible to display immunohistochemically co-labeled cells in paraffine sections of the bovine ovary. KDR and CD34 are the most frequent combinations of markers used to identify endothelial progenitors. The colabeled cells were detectable in the Tunica media but only rarely in the Tunica adventitia of arteries, and only sporadically in venous blood vessels. The co-labeled blood vessels were found predominantly in the corpus luteum and in close vicinity of the ovarian functional bodies. These findings indicate that a vascular wall may be the source of the endothelial progenitor cells in bovine ovaries. In order to verify these immunohistochemical findings, the measurement of mRNA amounts of the endothelial progenitor cell marker CD34 and KDR as well as the stem cell marker CD133 (Prominin-1) was carried out by qPCR. Therefore, specimens from various locations of the bovine ovary, who were taken during different luteal phases, were examined. The mRNA expression of the endothelial progenitor cell markers CD34 and KDR revealed a correlation, which supported the hypothesis that the cells colabeled with anti-KDR and anti-CD34 had characteristics of endothelial progenitor cells. CD133 showed its maximum expression during the developmental luteal stage. During the luteal stages regression and pregnancy, the expression of CD133 was notably reduced. The significant correlation of CD34 und CD133 confirmed the hypothesis that stem/progenitor cells are present in ovarian tissue. For this reason, the presence of resident endothelial progenitor cells within the ovaries is very likely. The foundation for obtaining reliable gene expression results in RT-qPCR experiments is the normalization of the expression of target genes by means of stably expressed reference genes. Thus, it is possible to distinguish variations that have been induced by experimental treatment from their actual biological variations. Bovine ovaries contain multiple heterogeneous cell populations in their different localizations. Both on a morphological and on a molecular biological level, these cell populations are under the influence of both the estrous cycle and pregnancy. At the time of the study, no validated reference genes were available that take both the morphological and the cyclic heterogeneity of the organ into account. Using various calculation algorithms (GENORM, NORMFINDER, BESTKEEPER), a total of 12 potential reference genes were therefore assessed for their suitability. In the course of the assessment, the reference genes H3F3B, RPS9, YWHAZ, RPS18, POLR2C, and UBA52 proved to be the most stable of the expressed genes. Based on this result, these genes present a pool of reference genes that can be used to normalize RT-qPCR findings for bovine ovaries. From this pool of genes, YWHAZ, H3F3B, and RPS9 can be recommended to normalize RT-qPCR findings irrespective of the location of the specimen or the luteal stage. Still, it is advisable to revalidate the reference genes for each new experiment. Each of the software programmes GENORM, NORMFINDER and BESTKEEPER has their individual advantages but also limitations when establishing the most stable expressed reference genes. Thus, it is recommended to use at least two different calculation algorithms, taking into account that variations between different groups of specimens can only be determined by applying the NORMFINDER algorithm.