Determining the function of histone modifications and transcription during centromere establishment in Caenorhabditis elegans

Abstract

The centromere is a chromosomal region specialized for directing chromosome segregation. The kinetochore assembles on the centromere, attaching microtubules to chromosomes in mitosis. Although centromeric DNA sizes and sequences vary among organisms, centromeric-specific H3 variant CENP-A/HCP-3 and kinetochore proteins are found to be conserved on most active centromeres. However, new centromeres, known as neocentromeres, can occasionally form on ectopic regions when the original centromere is inactivated or lost due to chromosomal rearrangements, and even on exogenously introduced DNA. This suggests that centromere identity is not determined by DNA sequences alone. How centromeres are maintained, inactivated and activated is still unclear. It is proposed that CENP-A, histone modifications and transcription may epigenetically specify the centromere location and maintain its function. Injecting DNA into the germline of the C. elegans leads to formation of artificial chromosomes (ACs) in embryos, which first undergo passive inheritance, but form centromeres and get the ability to autonomously segregate within a few cell cycles. Using this in vivo model, we have injected lacO repeats DNA, visualized ACs by GFP::lacI in real time, and monitored the frequency of equal AC segregation, which functionally represents centromere formation. Histone H3K9 and H4 acetylations are enriched on newly formed ACs, when compared to endogenous chromosomes. Tethering histone deacetylase HDA-1 to ACs specifically, through fusing HDA-1 with GFP::LacI, can reduce AC histone acetylations, equal AC segregation frequency, and initial deposition of kinetochore proteins, including CENP-A and NDC-80. Similarly, inhibiting transcription mediated by RNA polymerase II also delays initial CENP-A loading. These results indicate that histone acetylations facilitate efficient centromere establishment. Histone acetylations on chromatin and transcription can create an open chromatin environment, enhancing nucleosome disassembly and assembly, which can potentially contribute to centromere establishment. Using the same scheme, different histone modifiers could be tested systematically. This will give insights into the “histone code” required for de novo centromere formation.