A Saccharomyces cerevisiae model and screen to define the functional consequences of oncogenic histone missense mutations

Abstract

Somatic missense mutations in histone genes turn these essential proteins into oncohistones, which can drive oncogenesis. Understanding how missense mutations alter histone function is challenging in mammals as mutations occur in a single histone gene. For example, described oncohistone mutations predominantly occur in the histone H3.3 gene, despite the human genome encoding fifteen H3 genes. To understand how oncogenic histone missense mutations alter histone function, we leveraged the budding yeast model, which contains only two H3 genes, to explore the functional consequences of oncohistones H3K36M, H3G34W, H3G34L, H3G34R, and H3G34V. Analysis of cells that express each of these variants as the sole copy of H3 reveals that H3K36 mutants show different drug sensitivities compared to H3G34 mutants. This finding suggests that changes to proximal amino acids in the H3 N-terminal tail alter distinct biological pathways. We exploited the caffeine sensitive growth of H3K36 mutant cells to perform a high copy suppressor screen. This screen identified genes linked to histone function and transcriptional regulation, including Esa1, a histone H4/H2A acetyltransferase; Tos4, a forkhead-associated domain-containing gene expression regulator; Pho92, an N6-methyladenosine RNA binding protein and Sgv1/Bur1, a cyclin-dependent kinase. We show that the Esa1 lysine acetyltransferase activity is critical for suppression of the caffeine sensitive growth of H3K36R mutant cells while the previously characterized binding interactions of Tos4 and Pho92 are not required for suppression. This screen identifies pathways that could be altered by oncohistone mutations and highlights the value of yeast genetics to identify pathways altered by such mutations.

Histone H3 E50K mutation confers oncogenic activity and supports an EMT phenotype

“…Previously characterized oncohistones cause growth defects and/or altered drug sensitivity in yeast models (7,28,30,39) demonstrating that yeast models are valuable tools for exploring cellular effects of oncohistone proteins. To explore whether H3E50 changes modeled in yeast also impact cell physiology, we utilized the model organism S. cerevisiae.…”

“…In both genetic contexts (Figure 5C, D) Hht2-E50R confers a slightly stronger growth defect in the presence of bleomycin or caffeine as compared to Hht2-E50K. H3E50-mutant histones alter cell physiology via a mechanism that is distinct from H3K36M-mutant histones One advantage of the budding yeast system is that simple genetics can be employed to identify cellular pathways impacted by histone PTM and/or oncohistones (30,44,45). We recently employed such an approach to identify high copy suppressors of the caffeine-sensitive growth defect of H3K36 mutants in budding yeast (30).…”

“…H3E50-mutant histones alter cell physiology via a mechanism that is distinct from H3K36M-mutant histones One advantage of the budding yeast system is that simple genetics can be employed to identify cellular pathways impacted by histone PTM and/or oncohistones (30,44,45). We recently employed such an approach to identify high copy suppressors of the caffeine-sensitive growth defect of H3K36 mutants in budding yeast (30). This approach identified several chromatin modifying enzymes and complexes that when overexpressed can suppress the impaired growth of H3K36 mutants on media containing caffeine (30).…”

“…We recently employed such an approach to identify high copy suppressors of the caffeine-sensitive growth defect of H3K36 mutants in budding yeast (30). This approach identified several chromatin modifying enzymes and complexes that when overexpressed can suppress the impaired growth of H3K36 mutants on media containing caffeine (30). As S. cerevisiae expressing H3E50K and H3E50R are also sensitive to caffeine (Figure 5C, D), we examined whether the growth of the H3E50 mutants like H3K36R/M mutants is suppressed by overexpression of the suppressor genes, which would suggest that similar cellular pathways are impacted in these oncohistone models.…”

“…A challenge with understanding how the amino acid changes that occur in oncohistones alter histone function in the context of chromatin is that histones genes are present in multiple copies, creating a situation where a single oncohistone protein is present in the background of wildtype histones. One approach that has been utilized to understand how specific amino acid changes alter histone function in the context of chromatin has been to employ model systems (27)(28)(29)(30). This approach benefits from the fact that the histone proteins are some of the most evolutionarily conserved proteins (31) and the fact that simple model systems such as yeast have far fewer copies of histone genes than higher eukaryotes.…”