ABSTRACT
Genome-wide association studies have successfully identified thousands of common variants associated with complex traits and diseases. If only common variants are considered, a significant proportion of heritability in common diseases remains unexplained. One of the sources considered to explain this missing heritability are the rare genetic variants. Studying the consequences of rare genetic alterations offers additional opportunity for predicting molecular mediators underlying pathway deregulation. Here, we characterized the functional role of a rare variant having a significant contribution in sporulation efficiency variation. This causal variant is present in the coding sequence of TAO3, encoding a putative scaffolding protein conserved from yeast to humans and component of RAM network in yeast involved in mitosis. We observed that the role of TAO3 allele on meiosis is independent of ACE2, a transcription factor regulated by RAM network during mitosis. By expressing TAO3 causal allele conditionally during sporulation and quantitatively measuring cell cycle progression, we determined its role within the first 6 hours in meiosis, which coincides with cells entering into meiotic cell division. Time-resolved genome-wide gene expression analysis identified genes showing early and increasing expression trend during sporulation to be target genes of UME6, a crucial meiotic regulator. Genes regulating UME6, and other target genes expressed early, specifically, in the presence of causal TAO3, were chosen as candidate genes. Allele-specific functional validations identified ERT1 (regulator of switch from fermentation to respiration) and PIP2, (regulator involved in beta-oxidation of fatty acids) as the mediating genes associated with TAO3 causal allele and responsible for sporulation efficiency variation. Our study uncovered interesting link between TAO3 and regulators of metabolic cues that modulate the switch between multiple developmental phenotypes, viz. mitosis to meiosis. Although a small proportion in a population might contain rare variants, identification of novel regulators of sporulation from our study highlighted the significance of studying rare genetic variants to obtain novel insights into the phenotype and disease biology.
INTRODUCTION
The genetic architecture of complex traits is not completely understood. The ‘common disease, common variant’ rationale of genome-wide association studies (GWAS) is being challenged owing to the limited proportion of disease heritability explained by common variants, giving rise to the ‘missing heritability’ problem in quantitative genetics (Manolio et al. 2009; Zuk et al. 2014). Not considering the effect of rare variants has been suggested as one of the potential contributor of this ‘hidden’ instead of ‘missing’ heritability (Saint Pierre and Génin 2014). This view has been substantiated by the identification of multiple rare variants that confer considerable risk in diseases such as autism, schizophrenia and epilepsy (Stankiewicz and Lupski 2010). Since the initial genomic technologies were more adept at identifying common variants, sequencing methods and rare variant association studies are now being increasingly used to identify rare variants (Cirulli and Goldstein 2010; Zuk et al. 2014). Once identified, characterizing the functional role of rare variants associated with complex diseases has the potential for revealing new biology.
Common or rare, if genetic variants are in the core regulatory pathways underlying the trait, understanding the biological mechanism of phenotypic modulation is more direct. However, many genetic variants are identified in the ancillary biological processes. These processes are not part of the core regulatory pathways underlying the trait, but could impinge on the phenotype by being associated with the core pathways. For instance, gene identified by GWAS for face shape variation is a gap junction protein, which indirectly affects the mechanical load taken up by face bones (Liu et al. 2012). Ancillary pathways, hence, provide a good hub for accumulation of genetic variation. Their location in the regulatory network confers an advantage in regulating the phenotype while the core pathway genes could be too important to be varied, and hence probably under a negative selection pressure. This could be one of the reasons why crucial face forming gene of Shh has not been identified for facial shape development (Hallgrimsson et al. 2014). Therefore, characterization of variants, not part of the core biological processes associated with the phenotype is challenging.
Here, we characterized the functional role of a rare variant in the novel mitotic gene, TAO3, in yeast sporulation efficiency variation. We studied genome-wide gene expression dynamics in presence of both the alleles (common and causative rare alleles) of TAO3 (G4477C) during sporulation, and identified the core meiotic pathways and the ancillary pathways associated to sporulation showing differential expression. Our approach showed regulatory pathways involved in nutrient metabolism impinging on the core sporulation pathway, contributing to trait variation.
MATERIALS AND METHODS
Yeast strains and media
The yeast strains were grown in standard conditions at 30°C in YPD (1% Yeast extract, 2% Bacto peptone, 2% dextrose). Whole-genome resequencing of the strain YAD331 (Deutschbauer and Davis 2005) with S288c strain as the reference strain, identified two additional polymorphisms (Supporting Figure S1, Supporting Table S1). Three consecutive backcrosses were performed between the haploid obtained from this strain and the haploid reference strain, to remove the secondary polymorphisms. The sole genetic difference between the reference S288c strain and the backcrossed allele replacement strain was at TAO3(G4477C) position, which was confirmed by performing PCR-based sequencing 650bp up and downstream around the two secondary polymorphisms and the TAO3 polymorphism region. This backcrossed strain was diplodized to make it homozygous at TAO3(4477C) position and was termed as “T strain” in this study. The diploid parental strain S288c was termed as “S strain” in the study. All gene deletions in the study were made in the haploids of T and S strains, except the ones made in SK1 strain (Supporting Table S2). Deletions were performed and verified as described previously (Goldstein and McCusker 1999; Gietz and Woods 2002). All the experiments in this study were performed using the diplodized parent strains and their diploid derivatives. For replacing the endogenous TAO3 promoter (-150 to -1bp upstream start site) in T strain, with a tetracycline-responsive promoter, a tetO7-based promoter substitution cassette containing kanMX4 was amplified from the plasmid pCM225 (Bellí et al. 1998b). Diploid T strain with this tetO7-based cassette is termed PTet-TAO3(4477C) strain. The primers for sequencing, deletions and their confirmations are listed in Supporting Table S3.
Phenotyping
Sporulation efficiency estimation at 48h, progression through meiotic landmark events Meiosis I (MI) and Meiosis II (MII) and its quantitation was done as described previously (Gupta et al. 2015). For quantitation of meiotic landmarks in T strain, parametric curves assuming delayed and 1st order kinetics were fitted to the DAPI-stained meiotic progression time course data and fitting uncertainties were estimated by bootstrapping (Supporting File S1). Cell cycle progression data for S288c and SK1 strains was taken from Gupta et al. (2015) (Figure 1D-E). Conditional expression of TAO3(4477C) was performed by constructing PTet-TAO3 strain (details in Supporting File S1), which was responsive to tetracycline-analogue doxycycline (Bellí et al. 1998a; b). Addition of 2μg/ml doxycycline was utilized to decrease the activity of the gene, unless any other concentration was stated. For each strain, minimum three biological replicates were used and the experiment was carried minimum two times. Approximately 300 cells were counted per replicate. Fold difference was calculated as the ratio of mean sporulation efficiencies of the two strains A and B when the sporulation efficiency of A is greater than of B.
Statistical test for calculating sporulation efficiency
For comparing sporulation efficiency, two statistical tests were used: the pair test and the interaction test. The pair test tests the null hypothesis that two given strains (S and T) have the same sporulation efficiency.
The number yi,k of sporulated cells (4-nuclei count) among the total number of cells ni,k of strain i in replicate experiment k was modeled with a quasi-binomial generalized linear model using the logit link function and subject to a common log-odd ratio βi between replicates, i.e.: where .
The pair test tests the null hypothesis of equality of log odd-ratios for two strains i and j, i.e. .
In the case of S and T strains, the interaction test tests the null hypothesis that the effect of mutation A is independent of the effect of mutation B, taking the T strain as reference background. This test, thus, compares four strains: mutation A only, mutation B only, both A and B and neither A nor B (T strain). Here, the strain S was considered as a T strain mutated for TAO3(4477). For every interaction test, we considered the dataset of the four strains of interest and fitted a quasi-binomial generalized linear model using the logit link function and subject to: where, Ai and Bi are indicator variables of the mutations A and B in strain i respectively. The interaction test tested the null hypothesis that the odd ratio of sporulation in the double mutant equals the product of the odd ratios of each mutation, i.e. H0: βA,B = 0.
Both the pair test and the interaction test were implemented in the statistical language R with the function glm() assuming a constant variance function fitted by maximizing the quasi-likelihood and using the t-test on tested parameters (Gupta et al. 2015).
Whole genome gene-expression profiling
Sporulating yeast cell collection at 0h, 30m, 45m, 1h10m, 1h40m, 2h30m, 3h50m, 5h40m and 8h30m (logarithmic time-series), RNA isolation and cDNA preparation were performed as described (Xu et al. 2009). Samples were hybridized to S. cerevisiae yeast tiling array (Affymetrix, Cat# 520055). Arrays at each time point for both the strains were normalized together using vsn normalization method (Huber et al. 2002).
Whole genome gene-expression analysis
Within each strain, the log2 expression values obtained were smoothed using locfit at optimized bandwidth parameter h = 1.2 (Supporting Figure S2), base transformed for each transcript by subtracting the expression value at each time point from the baseline value at time point t = 0h (t0) (Supporting Table S4. This log2 fold change value with respect to t0 is described as “expression” throughout the manuscript. For identifying the genes showing temporal differential expression between T and S strains (Supporting Table S5, method implemented in EDGE software was used, which calculated statistically significant changes in expression between T and S strain over time (Storey et al. 2005). The differentially expressed genes were clustered according to their temporal expression patterns using time abstraction clustering algorithm implemented in the TimeClust software (Magni et al. 2008) (see Supporting File S1). Four major clusters were identified in each strain: Cluster I (Early trend), Cluster II (Increasing trend), Cluster III (Late trend), Cluster IV (Repressing trend) (Supporting Table S6). The transcription factors regulating a cluster of genes were extracted using the YEASTRACT database (Teixeira et al. 2014). Only those transcription factors were considered as candidate genes whose target genes were significantly enriched in the corresponding cluster (P ≤ 0.05, odds ratio ≥ 1.5). YEASTRACT database was also used to obtain the regulation matrix of yeast, for identifying target genes of regulators in this study such as UME6. Target genes for ACE2 were obtained from Nelson et al. (2003). Significantly enriched Gene Ontology terms by biological process (Bonferroni corrected P < 0.05, Table 1) were obtained from SGD Yeastmine (Balakrishnan et al. 2012).
Data availability
The array data for T strain reported has been deposited in ArrayExpress (http://www.ebi.ac.uk/arrayexpress/) with accession number E-MTAB-3889. The whole genome sequence data for T strain has been deposited in the European Nucleotide Archive (http://www.ebi.ac.uk/ena/) with the accession number PRJEB8698. The rest of the data is available as Supporting Information. Array data and whole genome sequence data for S strain were downloaded from Gupta et al. (2015).
RESULTS
Role of TAO3 causative allele in sporulation efficiency variation
The causative allele in TAO3(4477C), identified in high sporulating SK1 strain (Deutschbauer and Davis 2005), was not present in other high sporulating natural and laboratory yeast strains in the SGRP collection (Figure 1A). This indicated that TAO3(4477C) was a rare variant for high sporulation efficiency. Substitution of this causative single nucleotide polymorphism (G4477C) in 7,131bp long TAO3, in the low sporulating S288c strain, was sufficient to increase its sporulation efficiency by three-fold (P = 1.8 × 10−10, pair test in Methods, Figure 1B) compared to the S288c strain (TAO3(4477G)), reconfirming a significant role of TAO3(4477C) SNP in sporulation efficiency variation (Deutschbauer and Davis 2005). The increased sporulation efficiency in T strain saturated within 48h and did not increase beyond three-folds even when kept in sporulation medium for a week (Figure 1C). Quantitative analysis of cell cycle progression in T strain predicted that the difference between T and S strain occurred already during their entry into meiosis, since both, the time to initiate meiosis and rate of transition from G1/G0 into MI, were significantly different between the two strains (Figures 1D-E, Supporting Figure S3). Studying the progression of meiotic phases also showed that the T strain initiated meiosis within 12h (Figures 1D-E). To further resolve this 12h temporal phase when TAO3(4477C) could have an effect on the phenotype, this allele was conditionally expressed during sporulation (see Methods). The sporulation efficiency of the tetracycline-responsive T strain (PTet-TAO3(4477C) strain) in the presence of 2μg/ml doxycycline throughout the sporulation period was equivalent to the S strain (Figure 1F). Shorter periods of inactivation of TAO3 activity (in presence of doxycycline) showed that a 3h inhibition indeed reduced sporulation efficiency but this was significantly different (P = 0.02) from 48h inhibition, although, a 6h inhibition was equivalent to 48h inhibition. This showed that TAO3(4477C) allele affected sporulation efficiency within the first 6h in sporulation.
Role of TAO3 in meiosis is distinct from its role during mitosis
A part of the RAM (Regulation of Ace2p activity and cellular Morphogenesis) signaling network, Tao3 is required for activation and localization of an NDR protein kinase, Cbk1, another essential component of RAM network (Du and Novick 2002; Hergovich et al. 2006). This signaling network (consisting of Cbk1, Hym1, Kic1, Mob2, Tao3) contributes to various important aspects of mitotic cellular growth, particularly in cell separation by regulating transcription factor Ace2, in polarized growth (Nelson et al. 2003) and in cellular progression, through a Ace2-independent pathway (Bogomolnaya et al. 2006). Ace2 is a transcription factor that peaks early in mitosis and is involved in G1/S transition (Spellman et al. 1998). In S288c-background strains, the RAM network genes are essential and hence the affect of their deletion in sporulation efficiency has not been previously studied (Deutschbauer et al. 2002). Here, we tested if this upstream mitotic function of RAM network was also involved in meiotic regulation.
To determine the unique molecular effects of the causative TAO3 allele during sporulation, we compared genome-wide expression between T and S strains (see Methods). Genome-wide gene expression was measured for both these strains from 0h to 8h30m in sporulation medium (see Methods). Over one thousand gene transcripts (1,122 including non-coding SUTs, Supporting Table S5) were identified to be statistically significant in gene expression as a function of time between the two strains (FDR cut-off 10%). However, in this set, none of the RAM network genes showed significant differential expression in the presence of TAO3(4477C). A few of ACE2-regulated genes showed differential expression in the presence of TAO3(4477C) (11 genes labeled green in Figure 2A), however, deletion of ace2Δ had no affect on the sporulation efficiency of T strain (Figure 2B). Furthermore, even in the high sporulating parent SK1, ace2Δ, which showed a lesser clumping phenotype during growth in rich medium, had no effect on sporulation efficiency (Figure 2B). Since previously Ace2-independent effect of RAM network on cellular polarization have been observed (Nelson et al. 2003), therefore, RAM network could still be involved in meiosis independently of Ace2. Since by reducing TAO3(4477C) activity only during the growth phase in rich medium (in glucose) had no effect the sporulation efficiency (Figure 1F), hence we concluded that the causative allele of TAO3 might affect sporulation efficiency by interacting with a set of genes unique from the gene set involved during mitosis.
While studying the expression profiles of RAM network genes, we also observed a higher expression of TAO3 in the presence of the causative polymorphism in the T strain, compared to the S strain (P = 0.004, Figure 2C). By altering TAO3 expression, using different concentrations of doxycycline, we found that higher than native expression of TAO3 increased sporulation efficiency (no doxycycline added, P < 1 × 10−4) and concordantly, a decrease in expression (increased doxycycline concentration) reduced sporulation efficiency (Figure 2D). These results suggested an interesting possibility of how the coding polymorphism in the transcriptional activator TAO3 might be involved in sporulation efficiency variation by affecting its gene expression.
Global gene expression variation during sporulation in presence of causative TAO3 allele
While majority of regulation of sporulation occurs during both early and middle stages of meiosis (Neiman 2011), even late genes are known to affect sporulation efficiency in yeast (Deutschbauer et al. 2002). Between T and S strains, all of these classes of sporulation genes, i.e., early (IME2, HOP1, DMC1), middle and late (NDT80, SPR3, SPS1) showed differential expression (Figure 3A, Supporting Table S5). Most of the known sporulation genes (IME2, DMC1, NDT80, SMK1, SWM1, RAD6, REC8), including various regulators of meiosis (IME1, IME2, NDT80) were enriched (P = 5.5 × 10−12) in the cluster showing an increasing expression (Cluster II) during sporulation in T strain (Figure 3B). Approximately 50% of genes in this cluster (including IME1, IME2, DMC1, ECM11, NDT80) showed a similar trend in S strain also (Supporting Figure S4, Supporting Table S6). Our cell cycle progression results showed that lesser number of cells in S strain entered meiosis compared to T strain (Figure 1C). Higher expression of meiotic and meiosis-associated genes in T strain also reflected more cells entering meiosis. These results indicated that the early effects of TAO3 allele on sporulation were important for regulation of sporulation. The early expressing Cluster I genes of T strain belonged to biological processes regulating entry in sporulation, such as carbohydrate metabolic process, ion transport, mitochondrion organization and cellular respiration (Table 1). Sparse overlap (7%, Supporting Figure S4) was observed in genes of Cluster I between T and S strains, with many meiosis-associated pathway genes (carbohydrate metabolic process and mitochondrial organization) showing repression in S (Table 1, Supporting Figure S5). We next studied the genes regulating sporulation genes in Cluster II and of those getting expressed earlier than these sporulation genes, viz. early genes in Cluster I (Supporting Tables S7, S8).
Identifying candidate genes mediating the allele specific effects of TAO3 during sporulation using the temporal gene expression data
Regulators of genes upregulated either early or increasingly uniquely in T strain as time progresses in sporulation were enriched in nutrient metabolism and chromatin modification - the biological processes that are important for initiation of meiosis (Neiman 2011). UME6, a transcriptional repressor whose degradation is required for meiotic progression (Kassir et al. 2003) and is known to cause sporulation defects in both S288c and SK1 strains, was also identified. This core sporulation gene is also associated with various other important biological processes initiating meiosis (Lardenois et al. 2015) and was identified as a regulator for most of the biological processes occurring early in the presence of TAO3(4477C) (Table 2). From the list of regulators (Supporting Tables S7, S8), we identified those regulators that were upstream to UME6 also (Figure 4A, Supporting Table S9). These sporulation-associated regulators were selected as candidate causal mediating genes. They included ERT1, regulator of carbon source utilization (Turcotte et al. 2009) involved in the switch from fermentation to respiration in glucose-limiting conditions (Gasmi et al. 2014); OAF1, regulator of lipid metabolism forming a protein complex with PIP2 (Karpichev and Small 1998); and DAL81, regulator of nitrogen degradation pathway (Marzluf 1997). Interestingly, similar to UME6, OAF1 target genes were repressed (Cluster IV) in S strain (Supporting Table S10). Earlier work in S and SK1 strains has shown up regulation of ERT1, PIP2 and DAL81 in SK1 strain during sporulation (Primig et al. 2000), though their deletion in S strain did not effect sporulation efficiency (Deutschbauer et al. 2002).
A few other sporulation-associated regulators (Supporting Tables S7, S8) that were not upstream UME6 were also considered as candidate genes to be further investigated. These were: GAT3, a transcription factor involved in spore wall assembly (Lin et al. 2013); RSC2, required for the expression of mid-late sporulation genes (Bungard et al. 2004); GAT1 and PHO4, regulators of nitrogen and phosphorus metabolism, respectively. These sporulation genes are not known to have an effect on sporulation efficiency in S288c strain (Deutschbauer et al. 2002). Another putative causal mediating gene whose target genes were enriched in the Cluster II uniquely in TAO3(4477C) allele was XBP1, a transcriptional repressor induced in stress and starvation, involved in chromatin modification and G1 to S cell cycle progression. In SK1 strain, XBP1 was highly induced during meiosis and its deletion reduced sporulation efficiency (Mai and Breeden 2000), whereas, xbp1A in S strain showed no effect on its sporulation efficiency (Deutschbauer et al. 2002). Interestingly, XBP1 regulates VAC8, a vacuolar membrane protein required for efficient sporulation (Neiman 2005; Tang et al. 2006) that is also known to interact with TAO3 by yeast-two hybrid studies (Tang et al. 2006).
In order to functionally validate a few of the candidate genes identified from the above analysis, we used a genetic model described previously (Gupta et al. 2015). According to this model, if a gene has no effect on sporulation efficiency, its deletion would not affect on the phenotype in T strain. If a gene had an independent role in sporulation, reduction in sporulation efficiency by its deletion would be independent of the TAO3 background, and we would observe an additive effect irrespective of the background. Any significant deviation from this expectation would imply dependence on the genotype, such as the case of epistasis where deleting the gene in S background would not lead to decreased sporulation efficiency. While no difference was observed for GAT1 (Supporting Figure S6), single deletions of ERT1, PIP2 and GAT3 reduced the mean sporulation efficiency in the T strain significantly, by about 1.5-fold (P = 2.1 × 10−12, P = 6.1 × 10−13, P = 9.6 × 10−10 respectively, pair test in Methods, Figure 4B). Significant interaction terms (see Methods) were obtained between the genetic background (S and T) and ert1Δ and pip2Δ (P = 2.3 × 10−4, P = 0.04), but not for gat3Δ. This showed that the effect of ert1Δ and pip2Δ on sporulation efficiency was specific to TAO3(4477C). These analyses concluded that the meiotic role of allelic variant of TAO3 was complex and was effecting sporulation efficiency by interacting with regulatory pathways involved in carbon metabolism (ERT1) and membrane component proteins involved in beta-oxidation (OAF1-PIP2).
DISCUSSION
Strong effects on phenotypic variation have been observed as a consequence of rare coding variants (Cohen et al. 2004; 2005). TAO3(4477C), a rare variant specific to the high sporulating laboratory yeast strain of SK1 isolated from soil, is not present in the other yeast strains in the SGRP collection (Liti et al. 2009). Here, we showed the applicability of temporal genome-wide transcriptome profiling approach (Gupta et al. 2015) to understand more about regulation of yeast sporulation efficiency variation for the rare causal genetic variant TAO3.
Tao3, conserved from yeast to humans (Hergovich et al. 2006) has been functionally annotated for mitotic cell division (Du and Novick 2002; Nelson et al. 2003), until it got mapped for sporulation efficiency variation (Deutschbauer and Davis 2005). The SK1 allele of TAO3 has also been identified to affect growth variation in glycerol and high temperature (Wilkening et al. 2014). Tao3 localizes to polarized bud sites during mitosis (Nelson et al. 2003). Here, we observed genetic interaction of TAO3(4477C) with membrane-associated ERT1. Thus, it would be interesting to ascertain the localization of TAO3 variants during sporulation. Similar to other scaffolding proteins like Fry (Drosophila) and SAX-2 (C. elegans), Tao3 has multiple conserved Armadillo-like repeats (Hergovich et al. 2006) and the causal sporulation variant is present in one of these domains. Since overexpression of TAO3(4477G) is known to negatively affect the number of cells transitioning from G1 to S phase during mitosis (Bogomolnaya et al. 2006), a higher expression of the causative TAO3 allele specifically in meiosis raised the interesting possibility of this expression level variation to be involved in G1 to S phase transition in meiosis.
In complex diseases, combinations of molecular perturbations, which may vary in different patients, may be dysregulating similar components of a cellular system. Using a reductionist view of disease biology, many drug candidates have been successfully identified by studying the core pathways associated to a disease (Schadt et al. 2009). However, variation in the ancillary molecular players impinging on the core pathways and residing in the genetic bottleneck could also contribute variation in the phenotype (Lorenz and Cohen 2014). Genetic variation in mitosis-associated genes of MKT1 and TAO3 are examples of the non-canonical pathways associated with sporulation efficiency variation. Moreover, the novel metabolic and mitochondrial regulators identified for MKT1 (Gupta et al. 2015) and in this work for TAO3 as the mediating pathways causing variation in sporulation efficiency, are not part of the core sporulation regulatory framework. Their modulation by genetic variants provides important implications for understanding the rewiring of cellular networks underlying phenotypic variation. In this study, ERT1, regulator of carbon catabolite activation of transcription involved in the shift between fermentation to respiratory growth (Gasmi et al. 2014) and OAF1-PIP2, regulators of lipid metabolism necessary for the expression of sporulation specific genes SPS18 and SPS19 (Gurvitz et al. 2009), were identified as mediating pathways contributing to efficiency in meiosis. The molecular pathways in which these regulators were involved in, emphasized the role of pleiotropic metabolic pathways that act as a switch between multiple developmental phenotypes (Granek et al. 2011), to also have important functional roles in variation in these phenotypes. Since, Tao3 is a scaffolding protein involved in mitosis, and our findings link this protein in meiosis, this puts forth an interesting hypothesis for its dual role of in recruiting different molecular players in the two developmental phenotypes. Identification of these ancillary pathways contributing to variation in meiosis suggested that a systems view is necessary to understand the underlying complex biology of phenotypes including various diseases. This systems-based study provides support to the emerging view of disease as a complex network of interconnected pathways (Schadt et al. 2009), which could be used to refine the clinical translation of GWAS results to understand disease mechanism.
SUPPORTING INFORMATION
Supporting Files
File S1. Detailed methods
Supporting Tables
Table S1. Whole genome resequencing results for the TAO3 allele replacement strain
Table S2. Strains names
Table S3. Primer names
Table S4. Smoothed expression data, base transformed with respect to t0 for T and S strains
Table S5. Differentially expressed genes between T and S strains, with their P and Q values calculated using EDGE
Table S6. Genes in each cluster using TimeClust
Table S7. Transcription factors regulating unique early (Cluster I) genes of the T strain
Table S8. Transcription factors regulating unique increasing (Cluster II) genes of the T strain
Table S9. Differentially expressed target genes of regulators of candidate genes mediating the affect of TAO3
Table S10. Transcription factors regulating unique repressing (Cluster IV) genes of the S strain
Supporting Figures
Figure S1. Whole genome resequencing of TAO3 allele replacement strain (YAD331, (Deutschbauer and Davis 2005) in comparison to S288c reference strain
Figure S2. Smoothing of normalized temporal data using locfit
Figure S3. Mathematical modeling to identify stage of meiosis affected by TAO3 causal allele
Figure S4. Comparison of genes showing early (Cluster I) and increasing trend (Cluster II) between T and S strains
Figure S5. Genes showing early expression in T strain show expression at later time points or repressed in S strain
Figure S6. Sporulation efficiency of gat1Δ in T strain
ACKNOWLEDGEMENTS
We thank Manu Tekkedil for help with whole genome sequencing sample preparation. This research was supported by Tata Institute of Fundamental Research intramural funds and Department of Biotechnology grant BT/PR14842/BRB/10/881/2010 (H.S.); Bavarian Research Center for Molecular Biosystems and Bundesministerium für Bildung und Forschung through the Juniorverbund in der Systemmedizin “mitOmics” grant FKZ 01ZX1405A (J.G.); the National Institutes of Health, Deutsche Forschungsgemeinschaft and a European Research Council Advanced Investigator Grant (L.M.S.). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.