The incorporation of CENP-A into centromeric chromatin is an essential prerequisite for kinetochore formation. Yet, the molecular mechanisms governing this process are surprisingly divergent in different organisms. While CENP-A loading mechanisms have been studied in some detail in mammals, there are still large gaps to our understanding of CENP-A/Cid loading pathways in Drosophila. Here, we report on the characterization and delineation of at least three different CENP-A preloading complexes in Drosophila. Two complexes contain the CENP-A chaperones CAL1, FACT and/or Caf1/Rbap48. Notably, we identified a novel complex consisting of the histone acetyltransferase Hat1, Caf1 and CENP-A/H4. We show that Hat1 is required for proper CENP-A loading into chromatin, since knock-down in S2 cells leads to reduced incorporation of newly synthesized CENP-A. In addition, we demonstrate that CENP-A/Cid interacts with the HAT1 complex via an N-terminal region, which is acetylated in cytoplasmic but not in nuclear CENP-A. Since Hat1 is not responsible for acetylation of CENP-A/Cid, these results suggest a histone acetyltransferase activity-independent escort function for Hat1. Thus, our results point toward intriguing analogies between the complex processing pathways of newly synthesized CENP-A and canonical histones.

The acetylation of H4 lysine 16 (H4 K16Ac) in Saccharomyces cerevisiae counteracts the binding of the heterochromatin complex SIR to chromatin and inhibits gene silencing. Contrary to other histone acetylation marks, the H4 K16Ac level is high on genes with low transcription, whereas highly transcribed genes show low H4 K16Ac. Approximately 60% of cellular H4 K16Ac in S. cerevisiae is provided by the SAS-I complex, which consists of the MYST-family acetyltransferase Sas2, Sas4 and Sas5. The absence of SAS-I causes inappropriate spreading of the SIR complex and gene silencing in subtelomeric regions. Here, we investigated the genome-wide dynamics of SAS-I dependent H4 K16Ac during DNA replication. Replication is highly disruptive to chromatin and histone marks, since histones are removed to allow progression of the replication fork, and chromatin is reformed with old and new histones after fork passage. We found that H4 K16Ac appears in chromatin immediately upon replication. Importantly, this increase depends on the presence of functional SAS-I complex. Moreover, the appearance of H4 K16Ac is delayed in genes that are strongly transcribed. This indicates that transcription counteracts SAS-I-mediated H4 K16 acetylation, thus “sculpting” histone modification marks at the time of replication. We furthermore investigated which acetyltransferase acts redundantly with SAS-I to acetylate H4 K16Ac. esa1Δ sds3Δ cells, which were also sas2Δ sir3Δ in order to maintain viability, contained no detectable H4 K16Ac, showing that Esa1 and Sas2 are redundant for cellular H4 K16 acetylation. Furthermore, esa1Δ sds3Δ sas2Δ sir3Δ showed a more pronounced growth defect compared to the already defective esa1Δ sds3Δ sir3Δ. This indicates that SAS-I has cellular functions beyond preventing the spreading of heterochromatin.