Overview of the results
FACT is essential throughout evolution, however a deletion of one subunit of FACT complex in fission yeast (pob3Δ) is viable. We used this strain to perform a genetic screen to find pathways that cooperate with FACT in heterochromatin silencing. For this, we used pob3Δ mutant strain with ura4+ reporter inserted at peri-centromeres or in a mating type locus. In the absence of heterochromatin, ura4 is expressed and leads toxicity on 5FOA. Cells with impaired heterochromatin grow slower on 5FOA which provides a readout for the screen. Next, we crossed the FACT mutant with a deletion library of S. pombe and score for double mutant strains that show differential growth in comparison to single mutant, which indicates a genetic interaction. The screens results showed a group of mutants that grow worse suggesting that FACT is synthetic with those genes. Among those genes, there were known silencing factors, histone modifying enzymes, other histone chaperones and DNA replication factors. We also identified genes that alleviated the effect of FACT mutant at mating type locus. This group included several chromatin remodelers, transcription, RNA metabolism and DNA repair factors. Finally, we identified a small group of genes that suppressed FACT mutant at both heterochromatin regions tested. This group included genes involved in ubiquitin metabolism and chromosome segregation. In summary, our synthetic genetic screen with FACT mutant identified several known and novel genes that cooperate with FACT in heterochromatin silencing.
Some of the genes that interact with FACT, were linked to H2B ubiquitination (H2Bub). H2Bub is a highly dynamic histone post-transcriptional modification linked to transcription, DNA replication, cell differentiation and it is also missregulated in cancer cells. Moreover, H2Bub was shown to cooperate with FACT during gene transcription. This prompted us to study the link between FACT and H2Bub. For this, we used a histone H2Bub mutant that cannot be ubiquitinated (H2BK119R). We found that in this mutant, gene silencing at telomeres but not at centromeres is abolished. A double mutant of FACT and H2BK119R appears to be epistatic at telomeres. Moreover, in this mutant a main mark of heterochromatin - H3K9me2 is strongly decreased at telomeres but not at centromeres, suggesting that FACT and H2Bub specifically cooperate at telomeres to maintain heterochromatin.
Next, we assessed FACT binding to chromatin in H2Bub mutant. We found, that especially binding of Pob3 subunit was strongly decreased at both heterochromatic and euchromatic regions in the absence of H2Bub. Chromatin fractionation further revealed, that a soluble fraction of Pob3 was depleted in H2BK119R strain. Thus, our results suggest that H2Bub is important for proper FACT binding to chromatin genome-wide.
We also sought to determine if FACT and H2Bub have silencing functions outside of heterochromatin. For this, we performed RNA-seq in FACT and H2Bub mutants and we found that both FACT and H2Bub silence specific genes genome-wide. Especially stress response genes, meiotic genes and noncoding RNA were among genes repressed by both H2Bub and FACT.
Finally, we found that a large group of genes are repressed by FACT independent of H2Bub. Those genes are in a close proximity to telomeric heterochromatin in so called subtelomeres (ST). This part of the genome was showed to be silenced in H3K9me independent manner and it is characterized by low levels of active histone marks and lack of H2A.Z histone variant. To investigate FACT function at ST we sought to check H2A.Z levels at those genes. We observed accumulation of H2A.Z at some of those genes in a FACT mutant but not in H2Bub mutant. These results suggest that FACT protects silent genome from illegitimate H2A.Z assembly that may lead to activation of silent genes.
In summary SilentFACT project allowed us to identify a complex network of genetic interactions that FACT is a part of and specifically to identify H2B monoubiquitnation as an important chromatin mark that cooperates with FACT not only at active genes but also at silent part of the genome.
Results dissemination and exploitation
SilentFACT results have been disseminated to the local research community at (i) departmental seminars, (ii) pombe research club, (iii) Prof. Ladurner group meetings, (iv) Dr. S. Braun group meetings, (v) 8th Munich Chromatin Day (27.07.2017)
The results were disseminated at international research community at the EMBO Workshop – “Histone Variants: Molecular functions in health and disease” (06-08.09.2017); and 4th International Symposium "Chromatin Changes in Differentiation and Malignancies" (18 – 20.09.2017 Egmond aan Zee, The Netherlands).
Finally, the SilentFACT project has allowed me to initiate and build collaborations to other research groups, including Philipp Korber (LMU, Munich) and Tom Muir (Princeton University, USA).
