The INO80 chromatin remodeler sustains metabolic stability by promoting TOR signaling and regulating histone acetylation

Chromatin remodeling complexes are essential for gene expression programs that coordinate cell function with metabolic status. However, how these remodelers are integrated in metabolic stability pathways is not well known. Here, we report an expansive genetic screen with chromatin remodelers and metabolic regulators in Saccharomyces cerevisiae. We found that, unlike the SWR1 remodeler, the INO80 chromatin remodeling complex is composed of multiple distinct functional subunit modules. We identified a strikingly divergent genetic signature for the Ies6 subunit module that links the INO80 complex to metabolic homeostasis. In particular, mitochondrial maintenance is disrupted in ies6 mutants. INO80 is also needed to communicate TORC1-mediated signaling to chromatin, as ino80 mutants exhibit defective transcriptional profiles and altered histone acetylation of TORC1-responsive genes. Furthermore, comparative analysis reveals subunits of INO80 and mTORC1 have high co-occurrence of alterations in human cancers. Collectively, these results demonstrate that the INO80 complex is a central component of metabolic homeostasis that influences histone acetylation and may contribute to disease when disrupted.


27
chromatin remodelers and metabolic regulators in Saccharomyces cerevisiae. We found that, 28 unlike the SWR1 remodeler, the INO80 chromatin remodeling complex is composed of multiple 29 distinct functional subunit modules. We identified a strikingly divergent genetic signature for the

94
We identified genetic interactions between many chromatin and metabolic regulators, in 95 both nutrient rich media and metabolic stress conditions to reveal nutrient-specific interactions.

110
To do this, we conducted an EMAP of unstressed and metabolically challenged cells grown on 111 rich media (untreated), rapamycin or ethanol, which generated approximately a quarter million 112 interactions ( Figure 1A and Supplementary File 1). Rapamycin inhibits the TORC1 complex, a 113 master regulator of cellular growth (Loewith & Hall, 2011). Ethanol is a non-fermentable carbon 114 source that requires cells to utilize oxidative phosphorylation, whereas yeast preferentially 6 ferment glucose (Zaman, Lippman, Zhao, & Broach, 2008). We included a test library of 1536 116 alleles covering most major cellular processes, and significantly enriched for chromatin and 117 metabolic regulators (Ryan et al., 2012). We used 54 query strains that cover several chromatin 118 remodeling complexes, histone modifiers and metabolic signaling pathways ( Figure 1B).

132
Over 5000 significant interactions were identified in both the untreated and rapamycin 133 differential networks ( Figure 1E and Supplementary File 2). In the presence of rapamycin, 134 several TOR pathway genes, such as the TORC1 effector kinase SCH9 and TORC1 subunit 135 LST8 have increased number of significant interactions, indicating that the differential network is 136 broad and effective at identifying TOR dependent genetic interactions ( Figure 1F). Several 137 subunits of the INO80 chromatin remodeling complex (IES2, IES4, IES6) also have increased 138 number of interactions in the rapamycin differential network, supporting a metabolic role for 139 INO80. In contrast, the ethanol differential network yielded fewer genetic interactions and only a 140 few query strains have increased significant interactions, suggesting a less dramatic 7 reorganization of the genetic interaction landscape upon ethanol treatment than in response to 142 rapamycin ( Figure 1E, Figure 1 -figure supplement 5). Interestingly, four of the top five query 143 strains with the most significant interactions in the ethanol differential condition were subunits of 144 the INO80 complex (Supplementary File 2). As observed before, arp5D and ies6D mutants 145 have higher growth rates than expected on ethanol, presumably because these mutants have

151
We first used our EMAP data to comprehensively map the functional modules within the 152 INO80 complex by correlating the interaction profile of each query subunit across the test library 153 in untreated growth conditions (Figure 2A). Using this method, we found that INO80 subunits 154 were organized into 4 genetic modules, which were also independently identified in principal 155 component analysis (PCA) when pairwise correlations were k-means clustered ( Figure 2B).

156
Notably, the Nhp10 structural module clustered genetically and included Nhp10, Ies1, Ies3, 157 Ies5, and the Ino80 N-terminus on which the Nhp10 module assembles. Thus, the distinct in 158 vivo function of the NHP10 genetic module is organized among the subunits that are physically 159 associated. [Note, for clarity, genetic modules are denoted with all uppercase letters (e.g.

160
NHP10 module) and structural modules are denoted with an uppercase first letter only (e.g.

162
However, other subunits of the INO80 complex assemble in genetic modules that are 163 distinct from their structural modules. For example, Ies4 is structurally in the Arp8 module but 164 was slightly more genetically similar to the NHP10 genetic module (Figure 2A and B). In 165 addition, although Arp8 and Arp5 form separate structural modules, their genetic profiles are 166 similar and constitute the ARP5 genetic module, which also includes IES2. Ies2 is needed for 8 the Arp5 structural module to assemble with the INO80 complex (Yao et al., 2015), thus its in 168 vivo function is tightly connected to Arp5 and is reflected in our genetic analysis.

169
The genetic signatures of the INO80 helicase-SANT-associated (HSA) and insertion 170 domain mutants were closely associated with each other and clustered closest to many subunits 171 that assemble within those domains (Figure 2A and B). Namely, the HSA domain is required

184
In contrast to the INO80 complex, the SWR1 complex, another member of the INO80 185 chromatin remodeling subfamily (Mizuguchi et al., 2004), formed a strikingly cohesive genetic 186 module ( Figure 2D, E, and F). As before, analysis of non-unique subunits was not performed, 187 such as several subunits that assemble in the N-terminal module of SWR1 (Nguyen et al., 2013) 188 and are also found in the NuA4 acetyltransferase complex. Notably, our genetic analysis 189 highlighted Swc7 as an outlier, the genetic profile of which did not correlate with other SWR1 190 subunits and formed a distinct module in PCA analysis and k-means clustering ( Figure 2D and 191 E). Figure 2F summarizes the genetic modules for SWR1, which are arranged according to 9 previously identified structural modules (Nguyen et al., 2013). These genetic modules were 193 largely preserved in the rapamycin and ethanol EMAP (Figure 2 -figure supplement 1E-H).

194
We next broadened our analysis to compare the SWR1 and INO80 complexes together 195 to identify subunits that may facilitate cooperative or distinct function. Interestingly, the SWC7 196 genetic profile was most similar to that of the IES6 domain mutants (Figure 2 - figure   197 supplement 2), suggesting that these subunits have common function that is distinct from both

235
To determine if Ies6 is directly involved in mitochondrial inheritance we utilized the

323
We also observed that ino80D cells were much less responsive to rapamycin treatment,

365
One way in which INO80 can facilitate TORC1-dependent gene expression is by 366 regulating histone acetylation status, thus transcriptional potential. Our study finds that INO80 367 genetically interacts with the acetyltransferases Rtt109 and SAGA, and with several rapamycin-368 responsive deacetylases, including Rpd3L and Hst3 (Figure 4A and B; Figure 5C).

369
Interestingly, both Rpd3L and acetylated H3K56, the product of Rtt109 acetylation, are in the

902
Error bars represent standard error of the mean. Significance was determined using a Wilcoxon 903 rank sum test from at least 8 independent measurements compared to wild-type.          Figure 4A, in the untreated static, rapamycin static and differential conditions.

956
Significance is determined using a Wilcoxon rank sum test.