The centromere, required for kinetochore assembly and sister chromatid cohesion, is crucial for proper chromosome segregation during cell division. Centromeric histone 3 (CENP-A) is a histone H3 variant that is incorporated specifically into centromeric chromatin and is essential for establishing functional centromeres. CENP-A is inherited as an epigenetic mark that is maintained by a dedicated CENP-A loading machinery from one cell cycle to the next. This inheritance cycle is disrupted in the maternal germ line of the holocentric nematode Caenorhabditis elegans, where CENP-A-containing centromeres are removed at the mitosis-to-meiosis transition and are established de novo in oocytes. The aim of this thesis was to study the unique regulation of CENP-A, and therefore the centromere, in the hermaphrodite germ line of C. elegans. The first part of this thesis describes the identification of proteins that interact with CENP-A in C. elegans. In doing so, I discovered that many proteins that associate with the centromere could be grouped according to their function. I decided to focus my attention on proteins that belong to the ubiquitin-proteasome pathway. Subsequently, I describe how a post-translational regulation dependent on the ubiquitin pathway can explain the removal of CENP-A in the maternal germ line. Finally, I uncovered how the importance of a previously reported interaction between CENP-A and KNL-2 might be context-dependent. I observed that this interaction, dependent on the CENP-A N-terminal tail, is required for centromere establishment in the maternal germ line, but dispensable for centromere maintenance in mitosis. My results suggest that C. elegans CENP-A has evolved specific features that allow its regulation in the germ line and may influence the accurate transition from mitosis to meiosis. In addition, I conclude that centromere establishment and maintenance are functionally distinct, and require different regulatory factors, in C. elegans.