Kar4 is required for the normal pattern of meiotic gene expression

Abstract

Kar4p, the yeast homolog of the mammalian methyltransferase subunit METTL14, is required for efficient mRNA m⁶A methylation, which regulates meiotic entry. Kar4p is also required for a second seemingly non-catalytic function during meiosis. Overexpression of the early meiotic transcription factor, IME1, can bypass the requirement for Kar4p in meiotic entry but the additional overexpression of the translational regulator, RIM4, is required to permit sporulation in kar4Δ/Δ. Using microarray analysis and RNA sequencing, we sought to determine the impact of removing Kar4p and consequently mRNA methylation on the early meiotic transcriptome in a strain background (S288c) that is sensitive to the loss of early meiotic regulators. We found that kar4Δ/Δ mutants have a largely wild type transcriptional profile with the exception of two groups of genes that show delayed and reduced expression: (1) a set of Ime1p-dependent early genes as well as IME1, and (2) a set of late genes dependent on the mid-meiotic transcription factor, Ndt80p. The early gene expression defect is likely the result of the loss of mRNA methylation and is rescued by overexpressing IME1, but the late defect is only suppressed by overexpression of both IME1 and RIM4. The requirement for RIM4 led us to predict that the non-catalytic function of Kar4p, like methyltransferase complex orthologs in other systems, may function at the level of translation. Mass spectrometry analysis identified several genes involved in meiotic recombination with strongly reduced protein levels, but with little to no reduction in transcript levels in kar4Δ/Δ after IME1 overexpression. The low levels of these proteins were rescued by overexpression of RIM4 and IME1, but not by the overexpression of IME1 alone. These data expand our understanding of the role of Kar4p in regulating meiosis and provide key insights into a potential mechanism of Kar4p’s later meiotic function that is independent of mRNA methylation.

Author summary

Kar4p is required at two stages during meiosis. Cells lacking Kar4p have a severe loss of mRNA methylation and arrest early in the meiotic program, failing to undergo either pre-meiotic DNA synthesis or meiotic recombination. The early block is rescued by overexpression of the meiotic transcription factor, IME1. The kar4Δ/Δ cells show delayed and reduced expression of a set of Ime1p-dependent genes expressed early in meiosis as well as a set of later genes that are largely Ndt80p-dependent. Overexpression of IME1 rescues the expression defect of these early genes and expedites the meiotic program in the wild type S288c strain background. However, IME1 overexpression is not sufficient to facilitate sporulation in kar4Δ/Δ. Completion of meiosis and sporulation requires the additional overexpression of a translational regulator, RIM4. Analysis of kar4Δ/Δ’s proteome during meiosis with IME1 overexpression revealed that proteins important for meiotic recombination have reduced levels that cannot be explained by equivalent reductions in transcript abundance. IME1 overexpression alone rescues the defect associated with a catalytic mutant of Ime4p, implying that the early defect reflects the loss of mRNA methylation. The residual defects in protein levels likely reflect the loss of a non-catalytic function of Kar4p, and the methylation complex, which requires overexpression of RIM4 to suppress.

Introduction

Meiosis is a highly conserved eukaryotic cell differentiation process that begins with a single diploid cell and culminates in the production of four haploid cells. In the budding yeast, Saccharomyces cerevisiae, meiosis occurs in response to environmental stress and results in the production of four haploid spores (gametes) contained in a single ascus.

The yeast meiotic program is initiated when diploid cells are exposed to conditions lacking nitrogen and a fermentable carbon source. Nutrient signaling is coupled with ploidy sensing mediated by MATa1/α2 leading to expression of the early meiotic transcription factor, Ime1p [1]. Ime1p initiates the expression of a set of genes required for the initiation and early steps of the meiotic program, including pre-meiotic S-phase and meiotic recombination. Within Ime1p’s regulon is the meiotic protein kinase Ime2p. Ime2p functions to initiate pre-meiotic S-phase, activate the middle meiotic transcription factor, Ndt80p, and turn off IME1 expression, as well as other regulatory roles that are essential for the proper completion of meiosis [2–8]. Activation of Ndt80p induces the expression of genes required for completion of the meiotic divisions and spore maturation [1,7,9–12]. In addition to these three key meiotic regulators, the RNA-binding translational regulator Rim4p has also emerged as an important regulator of the meiotic program. Rim4p has been shown to have two functions: 1) it activates the expression of early meiotic genes, including IME2 [13,14], and 2) it blocks the translation of mid-late meiotic genes including CLB3 that are transcribed before their protein products are required [15]. The block to translation is mediated by the sequestration of the regulated mRNAs into amyloid-like aggregates [16]. The aggregates are targeted for degradation by Ime2p phosphorylation, which releases the bound mRNAs making them accessible to the translation machinery [17–19].

Methylation of mRNA is another key mechanism of regulation during meiosis. Among the most abundant mRNA modifications, m⁶A methylation is widespread among eukaryotes. Methylation in mammals is catalyzed by a trimeric complex, composed of METTL3, METTL14, and WTAP, which is highly conserved across eukaryotes [20–22]. In yeast, the complex was initially identified as containing Ime4p (the ortholog of METTL3), Mum2p (the ortholog of WTAP), and Slz1p (the ortholog of ZC3H13) [23,24]. However, work described in Park et al. (2023) and Ensinck et al. (2023) showed that Kar4p, the ortholog of METTL14, is also part of the catalytic complex and required for efficient mRNA methylation [25,26]. In yeast, mRNA methylation levels peak early in meiosis before the induced expression of NDT80 and initiation of the first meiotic division [24]. Methylation is present mainly around 3’ UTRs and the methylated transcripts are enriched on translating ribosomes. The methylated transcripts are enriched from genes involved in early stages of meiosis including DNA replication and recombination. Of particular importance, the transcripts of key regulators of meiosis have been reported to be methylated including IME1, IME2, and RME1 [27–29]. RME1 encodes the main transcriptional repressor of IME1; methylation of RME1 transcripts is associated with more rapid turnover. The reduction in RME1 transcript levels leads to reduced Rme1 protein production and increased expression of IME1; increased IME1 subsequently licenses cells to initiate the meiotic program. Remarkably, Ime4p has also been implicated in meiotic functions that are independent of mRNA methylation [30]. Taken together, transcriptional, post-transcriptional/translational, and post-translational mechanisms of regulation are all required to ensure proper completion of the meiotic program.

Kar4p is required early in meiosis before pre-meiotic S-phase and has two distinct functions in meiosis, termed Mei and Spo [25]. These two functions are distinct from its function during yeast mating, where it acts with the key mating transcription factor, Ste12p, to promote the transcription of genes required for late steps during mating [31–33]. Kar4p is required for efficient mRNA methylation during meiosis, resulting in higher Rme1p levels in kar4Δ/Δ and lower Ime1p levels, causing an early defect in meiosis. Ectopic overexpression of IME1 suppresses the early Mei^- kar4Δ/Δ defect. However, the additional overexpression of RIM4 is required to suppress the Spo^- defect and allow sporulation in kar4Δ/Δ. The Mei^- defect is associated with Kar4p’s role in mRNA methylation given that methylation acts upstream of IME1 and is similar to that of an IME4 catalytic mutant. That Rim4p overexpression is required to suppress the Spo^- defect but is not needed to suppress the IME4 catalytic mutation, suggests that Kar4p, like Ime4p, may also regulate meiosis via a mechanism that is independent of mRNA methylation.

The genetic analysis suggested that Kar4p and the methyltransferase complex plays a key role in the regulation of early meiotic transcription, as well as having less well-defined regulatory functions later in meiosis. Here, we use transcriptomic and proteomic analysis to examine how Kar4p impacts the meiotic transcriptional landscape and to determine the nature of the Spo^- meiotic defect. Microarray and RNA-seq data show that kar4Δ/Δ mutants have both an early transcriptional defect, which is rescued by overexpressing IME1, and a late transcriptional defect, which is rescued by additionally overexpressing RIM4. The late transcriptional defect largely involves genes within Ndt80p’s regulon, and there is a strong defect in NDT80 expression in kar4Δ/Δ. In addition, mass spectrometry (MS) identified a subset of proteins that are disproportionately reduced relative to their transcript levels in kar4Δ/Δ, even with overexpressed IME1. The protein levels were restored by the additional overexpression of RIM4, without correlated changes in the level of gene expression. This suggests that RIM4 overexpression likely impacts the efficiency of translation of these transcripts as opposed to increasing their overall levels. Taken together, these findings support a model in which Kar4p acts early through regulating IME1 expression to facilitate meiotic entry and has a later function that appears to be at least partially upstream of Ndt80p, which functions to positively regulate the translation of a set of transcripts required during various stages of meiosis.

Results

Meiotic transcriptional profile of kar4Δ/Δ

The requirement for Kar4p in mRNA methylation during meiosis, that mRNA methylation acts upstream of IME1, and that IME1 overexpression bypasses the Mei^- kar4Δ/Δ defect [25] all suggest that there should be differences in the meiotic transcriptional profiles between wild type and kar4Δ/Δ cells. To determine if there is a transcriptional defect in kar4Δ/Δ cells during meiosis, we used microarrays to measure mRNA abundance in wild type and kar4Δ/Δ strains over the first 16 hours of meiosis. RNA sequencing was also conducted across several time points to validate the microarray data. The data show remarkably similar expression profiles between wild type and kar4Δ/Δ. However, starting at 7 hours there is a prominent gene cluster with lower expression in kar4Δ/Δ. The genes in this cluster are implicated in mid- and late-meiotic processes based on gene ontology (GO) term analysis (Figs 1A, 1B and S1). Although this cluster is noteworthy, it is likely that they are indirect effects. First, the initial kar4Δ/Δ defect occurs before premeiotic S-phase, with DNA replication, meiotic recombination, and sporulation absent in kar4Δ/Δ cells. Second, the defect of Mei^- Kar4p mutants can be suppressed by over-expressing IME1, which acts at the initiation of meiosis, long before expression of the mid- and late-meiosis genes [25]. Thus, it is likely that the large changes in late gene expression reflect the consequences of an earlier, more subtle defect.

Fig 1. The meiotic transcriptome of kar4Δ/Δ.

(A) Heatmap of microarray data across the first 16 hours post induction of meiosis in wild type and kar4Δ/Δ. Expression was normalized to wild type in starvation conditions pre-induction of sporulation. Genes were clustered in Cluster3.0 and the heatmaps were constructed with Java TreeView. Red box highlights cluster of altered late gene expression. Source data for heatmap can be found in the supplementary file S2 Data. (B) Table of the top gene ontology terms of the cluster of impacted genes in kar4Δ/Δ. (C) Heatmap of RNA-seq data showing the fold change relative to t = 0 in expression level of key meiotic regulators in wild type and kar4Δ/Δ.

RNA-seq revealed that IME1 transcript levels as well as the transcript levels of other key early meiotic regulators, RIM4 and IME2, showed delayed and reduced expression in kar4Δ/Δ. In contrast, there was increased expression of RME1, the negative transcriptional regulator of IME1, across the time course (Fig 1C). Previous work demonstrated that mRNA methylation destabilizes RME1 transcripts and leads to increased expression of IME1 [30]. Thus, the reduction in IME1 expression may be due to the loss of negative regulation on RME1 transcripts in kar4Δ/Δ, which has greatly reduced levels of mRNA methylation [25]. However, it is surprising that a relatively small effect on IME1 levels in kar4Δ/Δ results in such a severe delay in meiotic entry [25]. One possibility is that kar4Δ/Δ has a stronger impact on a small fraction of Ime1p-dependent genes that are required for early stages of meiosis. To address this issue, we sought to identify a more specific set of Ime1p-dependent meiotic genes and examine their expression in kar4Δ/Δ.

KAR4 is required for a subset of IME1-dependent genes

The direct transcriptional targets of Ime1p are not well characterized. YEASTRACT [34], a curated repository of yeast transcription factors and target genes, contains 126 “experimentally” identified targets and 1088 putative targets based on the upstream activation sequence (UAS) “TTTTCHHCG” [35]. With nearly 20 percent of the yeast genome listed as possible genes of interest, this dataset was not useful for identifying Ime1p-dependent genes potentially impacted by Kar4p in our transcriptome data. We therefore sought to create a more specific set of IME1-dependent genes using a strain with IME1 under the control of the estradiol-inducible P_Z3EV (referred to as Pzev) promoter [36] to overexpress IME1 and identify genes that are rapidly induced under sporulation conditions. Specifically, we compared gene expression profiles of Pzev-IME1 strains after 2 hours in sporulation media with and without estradiol. Our data show IME1 expression in the estradiol sample, with no IME1 expression in the control sample.

We identified 236 genes with greater than 2-fold increased expression after IME1 induction. However, it is possible that strong induction of IME1 caused expression of off-target genes. Therefore, we compared the Pzev-IME1 induction data with the wild type IME1 meiotic transcription profile to identify the subset of Pzev-IME1 induced genes that show increased expression during normal meiosis. This criterion resulted in a list of 136 IME1-induced genes.

To assess the quality of our Ime1p-dependent gene list, we performed GO-term analysis on our list and the YEASTRACT IME1-regulated gene list. Analysis of our 136 IME1-induced genes returned 65 GO terms; “meiotic nuclear division” and “meiotic cell cycle” are the top two results (P-value 1.1 x 10⁻³⁰ and 2.2 x 10⁻²⁹ respectively) (Fig 2A). In contrast, analysis of the YEASTRACT IME1 regulated genes resulted in 17 GO-terms that were not as strongly enriched for meiotic processes (Fig 2A). In addition, our gene list includes 50% of the 42 genes comprising the core meiotically induced regulon of Ume6p [37]. The enrichment of genes involved in meiotic processes and substantial overlap with the Ume6p regulon support these data as a more accurate set of genes that are directly regulated by Ime1p than are annotated as such in YEASTRACT.

Fig 2. Kar4p is required for the expression of Ime1p dependent genes.

(A) Gene Ontology results of the YEASTTRACT list of Ime1p dependent genes (Left) and of the list of Ime1p dependent genes found in this study (Right). (B) Heatmap of microarray data showing the expression profile of Ime1p dependent genes in wild type and kar4Δ/Δ. The black box highlights the reduced and delayed expression of a set of these genes in kar4Δ/Δ. (C) List of Ime1p- and Kar4p-dependent genes (Left) and the top gene ontology terms of those genes (Right). Source data for heatmap can be found in the supplementary file S4 Data.

Using this list, we examined whether kar4Δ/Δ altered the expression of the Ime1p-induced genes. To visualize the expression dynamics of these Ime1p-dependent genes, we ordered them by their average expression between 6 and 13 hours and found that in wild type cells there were two waves of IME1-dependent gene expression. In the first wave, between 0 and 3.5 hours, 9 genes were strongly expressed. The second wave occurred between 6 and 13 hours and 38 genes were strongly expressed. The early expressing group of genes showed no differential expression between wild type and kar4Δ/Δ. However, a large group of genes expressed between 6 and 13 hours appeared to show severely delayed and reduced expression in kar4Δ/Δ cells compared to wild type (Fig 2B). We defined KAR4-dependence as genes whose average expression from 1 to 3.5 (early I) or 6 to 13 hours (early II) is 2-fold or greater in wild type. Using those criteria, we identified 25 genes with a 2-fold or greater defect in expression in kar4Δ/Δ compared to wild type, although there were many more that showed smaller effects (Fig 2C).

Previous work has shown that there is an early burst of IME1 expression that is independent of mRNA methylation, but the sustained increase in IME1 expression that occurs as the cells continue through meiosis requires mRNA methylation [30]. The lack of an effect on the expression of the early set of Ime1p-dependent genes in kar4Δ/Δ is consistent with this early burst of methylation-independent IME1 expression driving the expression of those genes. The delayed and reduced expression of the later set of genes in kar4Δ/Δ is consistent with the requirement for mRNA methylation in the normal expression of IME1 and genes in its regulon as the cells continue to move through meiosis. That these genes are eventually expressed in kar4Δ/Δ may be because mRNA methylation is not totally lost and/or that enough Ime1p is eventually made to drive their expression. Taken together, these data support a role for Kar4p in regulating the progression of cells into meiosis in part via regulation of IME1 and genes in its regulon.

IME1 overexpression suppresses the kar4Δ/Δ early transcript abundance defect

IME1 overexpression partially rescued the kar4Δ/Δ meiotic defect allowing pre-meiotic S-phase and meiotic recombination [25]. To determine whether IME1 overexpression rescues the reduction in Ime1p-dependent gene expression we examined gene expression in wild type and kar4Δ/Δ strains containing Pzev-IME1. As expected, estradiol induction resulted in rapid induction of the IME1-induced genes (Fig 3A), earlier than in cells with the wild type IME1 promoter. In kar4Δ/Δ, overexpression of Ime1p was sufficient to rescue the early transcript abundance defect of the Ime1p-dependent genes (Fig 3A), supporting the hypothesis that IME1 overexpression bypasses the requirement for Kar4p in establishing the early meiotic transcriptional profile. However, overexpression of Ime1p did not suppress the defect in gene expression observed in the late gene cluster described above. In addition, IME1 overexpression revealed two other late gene clusters that are reduced in kar4Δ/Δ (Fig 3B). These clusters are not enriched for genes involved in meiotic process, but do contain the polo-like kinase, CDC5, and the 14-3-3 protein encoding gene, BMH1. Both Cdc5p and Bmh1p have been shown to be important for a host of meiotic functions including recombination and meiotic commitment [38]. Given that impacted genes are largely involved in later meiotic processes and processes leading to meiotic commitment, we asked if expression of the mid-meiotic transcription factor, NDT80, was reduced in kar4Δ/Δ. The transcript abundance of NDT80 in kar4Δ/Δ was reduced 4-fold compared to wild type at 12 hours, when IME1 was overexpressed (Fig 3C and S6 Data). This suggests that the second block in meiosis in kar4Δ/Δ is at least partially upstream of NDT80 expression.

Fig 3. IME1 and RIM4 overexpression rescue the majority of the kar4Δ/Δ transcript level defect.

(A) Heatmap of microarray data of Ime1p dependent genes after IME1 overexpression in wild type and kar4Δ/Δ. Black box highlights the ability of IME1 overexpression to rescue the defect in expression of these genes seen in kar4Δ/Δ without IME1 overexpression. Source data for heatmap can be found in the supplementary file S4 Data. (B) Heatmap of microarray data in wild type and kar4Δ/Δ with IME1 overexpressed (Left) and IME1 and RIM4 overexpressed (Right). Red boxes highlight the three gene clusters that show a defect in expression in kar4Δ/Δ after IME1 overexpression but are rescued to some extent by additionally overexpressing RIM4. Source data for heatmap can be found in the supplementary file S2 Data. (C) NDT80 RNA-seq normalized counts from wild type and kar4Δ/Δ with IME1 overexpressed and with IME1 and RIM4 overexpressed. Counts were normalized using the standard normalization method in DESeq2. Error bars represent standard deviation between two biological replicates, which is equal to the range divided by the square root of 2.

RIM4 and IME1 co-overexpression suppresses the late kar4Δ/Δ transcript abundance defects

Co-overexpression of RIM4 and IME1 is necessary to fully suppress the kar4Δ/Δ meiotic defects. To understand the basis for suppression, we performed gene expression profiling on wild type and kar4Δ/Δ strains with both Pzev-IME1 and Pzev-RIM4. First, the expression data showed that the dual induction system does overexpress both IME1 and RIM4. Second, with both suppressor genes overexpressed, late gene expression was largely restored in the kar4Δ/Δ strain. Third, in wild type when Ime1p and Rim4p are co-overexpressed, the gene expression profile showed faster progression through the early meiotic transcriptional regime, relative to Ime1p expression alone (S1 Fig). However, in kar4Δ/Δ, an increase in the speed of progression through the early meiotic transcriptional regime is not as pronounced (S1 Fig) and there remained a moderate delay or reduction in the expression of NDT80 and the NDT80-regulon in kar4Δ/Δ (Fig 3B and 3C), relative to wild-type in which IME1 and RIM4 are overexpressed. This delay most likely explains why kar4Δ/Δ strains do not sporulate as well as wild type when both IME1 and RIM4 are co-overexpressed. Nevertheless, the co-overexpression of both suppressor genes does restore sporulation to kar4Δ/Δ to a level comparable to what is observed in wild type S288c cells [25].

Decreased Ime2p expression is not responsible for the defects in meiotic progression

Rim4p was first identified as a positive regulator of early meiotic gene expression including the expression of the meiotic kinase, IME2 [13,14]. Previous work showed that IME2 transcripts are methylated during meiosis [27,28]. Accordingly, we asked if Kar4p also plays a role in regulating IME2 expression. Ime2p levels were measured during meiosis using an epitope tagged Ime2p. In both wild type and kar4Δ/Δ cells, the level of Ime2p rose throughout the time course. Interestingly, we found no difference in the level of IME2 transcript or Ime2p between wild type and kar4Δ/Δ across a time course of meiosis (S2 Fig). This suggests that Ime2p is not limiting the progression of kar4Δ/Δ cells through pre-meiotic DNA synthesis and recombination, consistent with the prior defect in Ime1p expression. However, it is possible that differences in Ime2p levels might arise later in meiosis as wild type continues to progress through the meiotic program and kar4Δ/Δ does not.

Because IME1 overexpression allows progression past the initial meiotic block, we checked for defects in Ime2p expression that appear later in meiosis. Accordingly, we examined Ime2p levels in kar4Δ/Δ and wild type after overexpression of IME1, in the absence of RIM4 overexpression. IME2 transcript and protein levels were reduced only 2-fold at 12 hours post IME1 induction in kar4Δ/Δ compared to wild type (Fig 4A and 4B). However, expression of IME2 transcript and protein was similar to wild type at earlier time points and it was only the late burst of expression that was absent in kar4Δ/Δ (Fig 4A and 4B). The deficit in IME2 expression at 12 hours between wild type and kar4Δ/Δ could be due to reduced NDT80 expression in kar4Δ/Δ (Fig 3C) since mutants defective for NDT80 expression also lose this late burst of Ime2p expression [10]. Taken together, this suggests that defects in Ime2p expression are not solely responsible for the continued loss of sporulation in kar4Δ/Δ after IME1 overexpression.

Fig 4. Defects in Ime2p expression are not responsible for block in meiotic progression.

(A) Western blots of Ime2p-13MYC across a time course of meiosis in wild type and kar4Δ/Δ with IME1 overexpressed. Kar2p is used as a loading control. (B) IME2 RNA-seq normalized counts from wild type and kar4Δ/Δ with IME1 overexpressed as well as kar4Δ/Δ with IME1 and RIM4 overexpressed. Counts were normalized using the standard normalization method in DESeq2. Error bars represent standard deviation between two biological replicates, which is equal to the range divided by the square root of 2. (C) Western blots of Ime2p-13MYC across a time course of meiosis in kar4Δ/Δ with either IME1 overexpressed or IME1 and RIM4 overexpressed. Kar2p is used as a loading control.

To determine the impact of RIM4 overexpression on Ime2p levels, we assayed Ime2p in kar4Δ/Δ with both IME1 and RIM4 overexpressed. In this strain, IME2 transcript levels peaked earlier than when only IME1 is overexpressed (Fig 4B). Therefore, we looked at Ime2p at earlier time points in kar4Δ/Δ with either IME1 overexpressed or IME1 and RIM4 co-overexpressed. Ime2p levels peaked at 4 hours in the double overexpression strain and then begin to go down by 6 hours whereas Ime2p levels continue to increase in kar4Δ/Δ at these times with just IME1 overexpressed (Fig 4C). Thus, the additional overexpression of RIM4 sped up, but did not cause higher levels of the expression of Ime2p in kar4Δ/Δ.

Kar4p is required for the level of multiple meiotic proteins

Given that Ime2p levels did not appear to be limiting the progression of kar4Δ/Δ after IME1 overexpression, but that the second block is suppressed by a known translational regulator, we hypothesized that there may be critical regulatory proteins whose expression is impacted in kar4Δ/Δ. To identify candidate proteins, mass spectrometry was used to identify proteins whose levels are strongly dependent on Kar4p, but whose transcript levels are not. The P_Z3EV-IME1 strains were used for three reasons: first, they show greater meiotic synchrony; second, IME1 overexpression suppresses the early transcript abundance defects in kar4Δ/Δ; and third, IME1 overexpression alone can suppress the defect of a catalytic mutant of Ime4p, suggesting that defects that persist after IME1 overexpression do not involve mRNA methylation. Because meiotic defects appear at a later stage in IME1-overexpressed cells, we examined proteins 8- and 12-hours post-IME1 induction. Total protein samples were digested with trypsin, fractionated using ion exchange, and analyzed by liquid chromatography—mass spectrometry/mass spectrometry (LC-MS/MS).

Using LC-MS/MS, we were able to identify 4068 proteins expressed during meiosis. Spectral count data from the mass spectrometry was used to approximate relative protein levels. We used a cutoff of proteins reduced more than 2-fold in kar4∆/∆ cells relative to wild type. At 8 hours, 432 proteins were present at less than 50% the levels in kar4Δ/Δ compared to wild type. GO term analysis of these low proteins at 8 hours returned “meiotic cell cycle” as the sixth term with a P value of 7.74x10^-5. At 12 hours, 318 proteins were present at less than 50% the levels in kar4Δ/Δ relative to wild type. GO term analysis of these low proteins at 12 hours returned “meiotic cell cycle” as the second term with a P value less than 0.001. Many proteins were low at both 8 and 12 hours in kar4∆/∆ including Sps2p, Gas4p, Gmc2p, Mei5p, and Sae3p. There were also proteins including Ecm11p, Hed1p, Spo11p, and Rec8p that were expressed at or above wild type levels at 8 hours and then went down to lower than wild type levels at 12 hours.

To validate some of the candidate proteins identified by mass-spectrometry, we epitope-tagged two proteins of interest (Gas4p and Sps2p) and assayed their protein levels using western blotting and transcript levels with qPCR. We were not able to detect either Gas4p or Sps2p at 8 hours post IME1 induction in either wild type or kar4∆/∆, which supports the mass spec data (Fig 5A). At 12 hours post IME1 induction in wild type, both Gas4p and Sps2p were detectable, and this was accompanied by increased transcript levels (Fig 5A and 5B). However, in kar4∆/∆ at 12 hours there was still no detectable Gas4p or Sps2p. Based on the quantification of Gas4p and Sps2p at 8 hours when both IME1 and RIM4 are overexpressed, we infer that Gas4p and Sps2p are reduced at least 15- and 12-fold, respectively, in kar4∆/∆ relative to wild-type at 12 hours post IME1 induction (S1 Data). Thus, the reductions in protein abundance were much larger than the 3-fold reduction observed at the transcript level (Fig 5A and 5B). It is important to note that both GAS4 and SPS2 are in part under the control of Ndt80p, which is reduced 4-fold in kar4∆/∆ with IME1 overexpression (Fig 3C).

Fig 5. Kar4p is required for wild type levels of several meiotic proteins.

(A) Western blots of Sps2p-3HA and Gas4p-3HA in wild type and kar4Δ/Δ with IME1 overexpressed and in kar4Δ/Δ with IME1 and RIM4 overexpressed. Blots in the same row were exposed for the same length of time. Cropping of the Gas4p-3HA blot was to match the order of the samples to the Sps2p-3HA blot. Kar2p serves as a loading control. (B) qPCR measurements of the change in expression for SPS2 and GAS4 between wild type and kar4Δ/Δ with IME1 overexpressed. Fold changes were calculated using the ΔΔCt method and PGK1 was used as the normalizing gene. (C) SPS2 (Left) and GAS4 (Right) RNA-seq normalized counts from kar4Δ/Δ with either IME1 overexpressed or IME1 and RIM4 overexpressed. Counts were normalized using the standard normalization method in DESeq2. Error bars represent standard deviation between two biological replicates, which is equal to the range divided by the square root of 2.

We next wanted to determine how the additional overexpression of RIM4 can suppress the later kar4Δ/Δ defect. Remarkably, upon IME1 and RIM4 overexpression, Gas4p and Sps2p become detectable at 8 hours in kar4Δ/Δ, but there was no change in transcript abundance relative to kar4Δ/Δ with only IME1 overexpressed (Fig 5A and 5C). At 12 hours, both proteins were strongly expressed, although not at quite as high levels as in wild type with only IME1 overexpressed. The increase was accompanied by only a 2-fold increase in transcript abundance. Thus, the relatively small changes in transcript levels were accompanied by greater than 5- and 11-fold increases in the levels of Gas4p and Sps2p, respectively, suggesting that the overexpression of RIM4 is functioning to enhance translation of these two proteins (S1 Data). This finding points to a potential mechanism of the early positive acting meiotic function of Rim4p.

Defects in the expression of recombination proteins persist after IME1 overexpression in kar4Δ/Δ

Although the experiments described above validate the mass spectrometry and microarray/RNA-seq data, it is likely that the second block in meiosis is caused by reduced levels of proteins that act earlier than Sps2p and Gas4p. From the mass spectrometry data, key recombination proteins were found to be mis-regulated in kar4∆/∆ including Mei5p, Sae3p, Gmc2p, and Ecm11p. All four proteins are highly reduced at 12 hours in kar4∆/∆, but we see little to no impact on their transcript levels (Fig 6A–6D) suggesting that Kar4p could be positively regulating the translation of these proteins as opposed to their transcript abundance.

Fig 6. Kar4p is required for the expression of key recombination proteins.

Protein levels measured by mass spectrometry (left) and RNA-seq normalized counts (right) of (A) MEI5, (B) SAE3, (C) ECM11, and (D) GMC2. Error bars represent standard deviation of two biological replicates, which is equal to the range divided by the square root of 2. Normalized counts were calculated using DESeq2. Source data for the Mass-spec experiment can be found in supplementary data file S5 Data. (E) Screen for recombination described in [25]. Wild type and kar4Δ/Δ both carry IME1 on a high-copy number plasmid also carrying the URA3 gene. Growth on SC-Ura (left) was used to assay growth and maintenance of the plasmid. Growth on SC-Ura -His +can (right) was used to select for recombination events. Spots are 10-fold serial dilutions of a starting concentration of 1 OD unit of cells for each strain at each time point (0-, 8-, 12-, and 24-hours post movement into sporulation media).

Interestingly, we also saw lower levels of some of these proteins (Mei5p, Gmc2p, and Sae3p) at the 8-hour time point (Fig 6A, 6B and 6D), which would suggest that defects in recombination should be present in kar4∆/∆ even at this relatively early time point. To address this, we examined the timing of meiotic recombination in wild type and kar4∆/∆ cells carrying a high-copy number plasmid containing IME1. In wild type, cells that have undergone recombination begin to appear at 8 hours after transfer to sporulation conditions, increasing greater than 10-fold over the next 4 hours (Fig 6E). As predicted, in cells lacking Kar4p, cells that had undergone recombination were not observed even after 24 hours (Fig 6E). Together, these data suggest that Kar4p positively regulates the translation of proteins important for meiotic recombination including Mei5p, Sae3p, Gmc2p, and Ecm11p and loss of this regulation impairs the efficiency of meiotic recombination. The persistence of the expression defects, despite the ability of IME1 overexpression to suppress a catalytically dead mutant of Ime4p [25], implies that they reflect a loss of a function of Kar4p separate from mRNA methylation.

Removal of Mek1p does not permit sporulation in kar4Δ/Δ After IME1 overexpression

Given the persistent defect in the expression of recombination proteins, one possibility is that activation of the recombination checkpoint mediated by Mek1p [39] is preventing the kar4Δ/Δ mutants from progressing into the meiotic divisions. To address this, we created mutants lacking Mek1p, which abolishes the cells’ ability to activate this checkpoint and allows them to progress into the divisions even if recombination is not complete [39]. Deletion of MEK1 in a kar4Δ/Δ strain overexpressing IME1 still did not permit sporulation (Fig 7A). Moreover, high copy NDT80, the target of Mek1p regulation, also did not permit sporulation (S1 Data).

Fig 7. Removal of Mek1p does not permit sporulation in kar4Δ/Δ.

(A) Sporulation levels of wild type, kar4Δ/Δ, mek1Δ/Δ, and kar4Δ/Δ rme1Δ/Δ with IME1 overexpressed across a time course of meiosis. All dyads, triads, and tetrads were counted. Error bars represent standard deviation of three biological replicates. (B) Western blot of Ndt80p across a time course of meiosis in wild type, kar4Δ/Δ, and kar4rme1Δ/Δ with IME1 overexpressed and kar4Δ/Δ with IME1 and RIM4 overexpressed. Kar2p is used as a loading control. “*” indicates a non-specific band.

One possibility is that Kar4p is also required downstream of Ndt80p; removing Mek1p could have licensed Ndt80p expression, but defects remained that prevent sporulation. To address this, we used an antibody to Ndt80p [40] to determine if it was expressed in kar4Δ/Δ mek1Δ/Δ after IME1 overexpression. Ndt80p is strongly expressed at 12 hours in wild type, but was not detectable in either kar4Δ/Δ or kar4Δ/Δ mek1Δ/Δ (Fig 7B). Thus, Ndt80p expression is impacted much more strongly than NDT80 mRNA levels (Fig 3C), similar to Gas4p, Sps2p, Mei1p and Sae3p, among others (Figs 5 and 6). This suggests that the requirement of Kar4p for the expression of Ndt80p goes beyond facilitating timely recombination and may involve more direct regulation of Ndt80p expression.

Given these findings, we asked if the overexpression of both IME1 and RIM4 facilitates Ndt80p expression in kar4Δ/Δ. Ndt80p was faintly detectable at 12 hours in kar4Δ/Δ with both genes overexpressed but reduced compared to wild type with just IME1 overexpressed, consistent with the RNA-seq data (Figs 3C and 7B). The persistent defect in Ndt80p expression explains in part why sporulation is still delayed and reduced in these strains compared to wild type [25].

Discussion

Through analyzing both the transcriptome and proteome of kar4Δ/Δ mutants during meiosis we now have a better understanding of the molecular underpinnings underlying the kar4Δ/Δ meiotic defects, and how overexpression of IME1 and RIM4 suppresses those defects. Establishing a more defined set of Ime1p-dependent genes demonstrated the impact of loss of Kar4p on the IME1 regulon. Overexpression of IME1 revealed that a later block in meiosis is at least partially upstream of NDT80 expression. RIM4 overexpression appears to rescue this later defect by impacting the translation of transcripts as opposed to facilitating increased gene expression.

Microarray and RNA-seq data showed that kar4Δ/Δ mutants have a wild type transcriptional profile with two exceptions: first, they have an early defect in the transcript level of IME1, as well as a subset of Ime1p-dependent genes that are not immediately activated by the early burst of IME1 expression that is independent of mRNA methylation. The early defect is suppressed by the overexpression of IME1. Second, a late defect that is not suppressed by IME1 overexpression alone but is suppressed by the additional overexpression of RIM4. However, the late transcriptional defect is most likely an indirect effect of a prior arrest point of kar4Δ/Δ mutants suppressed by IME1 overexpression. The late genes that are impacted are downstream of Ndt80p, but the block in kar4Δ/Δ after IME1 overexpression is upstream of Ndt80p expression. Thus, the low levels of the late genes are most likely due to the loss of expression of Ndt80p. The early defect in IME1 expression and the expression of Ime1p dependent genes support findings that showed lower levels of Ime1p and IME1 transcript in kar4Δ/Δ during meiosis [25]. Therefore, the overexpression of IME1 bypasses the requirement of Kar4p in regulating the expression of IME1 and its targets. The later block in gene expression after IME1 overexpression suggests that Kar4p is required at another step that impacts the expression of Ndt80p.

Given that the overexpression of the translational regulator RIM4 is required to suppress the later kar4Δ/Δ meiotic defect, we hypothesize that Kar4p may also have a role in translational regulation during meiosis. Rim4p’s role in promoting the expression of Ime2p made it a potential candidate for Kar4p translational regulation. With IME1 under its own promoter, we saw no difference in the expression of IME2 transcript or Ime2p between wild type and kar4Δ/Δ. Thus, it is unlikely that Ime2p is limiting the progression of kar4Δ/Δ cells early in meiosis. However, there is a defect in both IME2 transcript and Ime2p levels that is revealed when IME1 is overexpressed. Ime2p is required for the activation of the mid-meiotic transcription factor Ndt80p, which initiates the expression of genes required for the completion of meiotic recombination and prophase I exit as well as the meiotic divisions and spore maturation. Low Ime2p levels were rescued in kar4Δ/Δ when both IME1 and RIM4 were overexpressed. In addition, the overexpression of both IME1 and RIM4 together significantly sped up the time to peak expression of Ime2p in kar4Δ/Δ.

We initially screened Ime2p levels because Rim4p is required for Ime2p expression, but an unbiased mass spectrometry approach identified many proteins that are reduced in kar4Δ/Δ at both 8- and 12-hours post induction of IME1 expression. Many of these proteins were important for meiotic recombination including Mei5p, Gmc2p, Sae3p, and Ecm11p. These proteins showed little or no difference in transcript levels between wild type and kar4Δ/Δ at 8 and 12 hours, implying that the reduced protein levels at 8 and 12 hours in kar4Δ/Δ is likely due to defects in translation as opposed to transcript abundance. Consistent with the fact that many of the impacted proteins are involved in recombination, we found that recombination was delayed in kar4Δ/Δ, even with IME1 overexpressed. Defects in recombination would activate the meiotic recombination checkpoint mediated by Mek1p [39], which acts antagonistically to Ime2p to block the activation of Ndt80p. Activation of the MEK1 checkpoint could explain the defects in NDT80 expression in kar4Δ/Δ, as well as in Ime2p expression; Ndt80p also induces IME2 expression, resulting in Ime2p and Ndt80p levels peaking at similar times later in meiosis [10]. However, removal of this checkpoint still did not permit sporulation nor the expression of Ndt80p in these strains. This suggests that Kar4p’s role in facilitating entry into the meiotic divisions goes beyond promoting meiotic recombination and may involve more direct activation of Ndt80p expression. Because overexpression of IME1 suppresses the meiotic defect of a catalytic mutant of Ime4p, we hypothesize that the defect in protein expression involves the proposed non-catalytic function of the methyltransferase complex. Consistent with this, recent work has shown that the ortholog of Ime4p in mammals, METTL3, positively regulates the translation of transcripts in an m⁶A-independent manner by interacting with PABP and cap-binding factors [41,42]. Given that METTL3 and METTL14 work together to bind mRNA, it is likely that METTL14, Kar4p’s ortholog, is also involved in this function. However, due to the nature of the arrest after IME1 overexpression, we cannot determine if mRNA methylation is also important for events downstream of Ndt80p. Future work will determine if Kar4p and other members of the complex are regulating the translation of transcripts in yeast in a similar manner and if mRNA methylation plays a role in regulating later steps of meiosis.

The co-overexpression of IME1 and RIM4 partially rescued the protein level defects. Rim4p is best known as a repressor of translation that functions as an aggregate to sequester mRNAs from the translational machinery [16]. However, Rim4p was first identified as a positive regulator of IME2 [14]. No further work was done to explore the positive regulatory role of Rim4p. The fact that overexpression of RIM4 rescues the kar4Δ/Δ translational defect suggests that Rim4p acts directly as an enhancer of translation or that the overexpression facilitates Rim4p aggregates to expand their regulon and sequester a negative regulator that is not normally bound by the aggregates. Work on an analogous protein to Rim4p in mammals, “Deleted in Azoospermia Like” (DAZL), has shown that DAZL exists as both a monomer and an aggregate. The aggregate acts to block translation and the monomeric form promotes translation through interactions with PABP [16,43,44]. Rim4p monomers are seen early in meiosis suggesting that it is the monomeric form of Rim4p that is important for its role in meiotic entry and these monomers may act to positively regulate translation in a similar manner to DAZL monomers. An interaction between Kar4p and Rim4p has not been detected, which suggests that the suppression bypasses the requirement for Kar4p.

Taken together, these data point to Kar4p being a key player in the post-transcriptional/translational regulation of meiosis. In support of Kar4p’s role in mRNA methylation, we see that expression of both IME1, and Ime1p-dependent genes are lower in kar4Δ/Δ during meiosis. These results indicate that Kar4p is required upstream of IME1 expression. That IME1 overexpression bypasses the requirement for Kar4p in meiotic entry, but cells remain blocked before the induction of NDT80 expression, suggests that Kar4p is required at another step during meiosis. That the additional overexpression of RIM4 allows kar4Δ/Δ cells to complete sporulation points to Kar4p playing a role in translational regulation that is important upstream (and possibly downstream) of NDT80 expression. Interestingly, deletions of other methylation complex members (Ime4p and Mum2p) can also be made to sporulate after overexpression of both IME1 and RIM4 [25]. Future work will determine whether the regulation of these protein levels is similar to the non-catalytic functions described in other eukaryotes. Understanding the non-catalytic function of Kar4p may also provide insights into how Rim4p functions positively in meiosis. These findings position Kar4p as a key regulator of meiosis at multiple levels and further experimentation will seek to determine how exactly this intrepid protein is carrying out that regulation.

Materials and methods

Sporulation

Cultures were grown overnight at 30°C in YPD (yeast nitrogen base (1% w/v), peptone (2% w/v), and 2% glucose), back diluted into YPA (yeast nitrogen base (1% w/v), peptone (2% w/v), and potassium acetate (1% w/v), and allowed to grow for 16–18 hours before being transferred into 1% (w/v) potassium acetate sporulation media supplemented with histidine, uracil, and leucine at a concentration of 0.5 OD₆₀₀ unit/ml. Sporulating cells were cultured at 26°C for various amounts of time depending on the nature of the experiment. For experiments involving overexpression, 1 μM of β-estradiol was added to cultures once they were moved into the 1% potassium acetate media.

RNA preparation

Cells from sporulation cultures were collected by vacuum filtration on nitrocellulose filters and flash frozen with liquid nitrogen. Samples were stored at -80°C until extracted for RNA. RNA was extracted by acid-phenol method. Lysis buffer was added and then vigorously vortexed. Following lysis, phenol saturated with 0.1 M citrate (Sigma-Aldrich, P4682) was added. Lysates were incubated for 30 minutes at 65°C, vortexing every 5 minutes. Lysates were chilled on ice and then spun. Supernatant was added to a heavy phase lock tube along with chloroform. After light mixing, the samples were centrifuged. The aqueous layer was moved to a new tube containing sodium acetate. The samples were washed with ice-cold 100% ethanol and left to incubate at -20°C for 30 minutes to 16 hours. Following the incubation, ethanol-precipitated samples were spun at full speed for 5 minutes to pellet the RNA. The pellets were washed with ice cold 70% ethanol and pulse spun. Remaining alcohol was aspirated, and the RNA samples were resuspended in 100 μl of water. RNA samples were cleaned up using the Qiagen RNeasey kit (Qiagen 74106) and then quantified using a Nanodrop.

Gene expression microarrays

Microarray analysis was performed as described by [45] with slight modifications: RNA samples were handled in an ozone-free environment during the labeling process and labeling was performed using the Quick Amp labeling kit (5190–0447)) according to a modified labeling protocol. Reference RNA was labeled with Cy3-CTP (NEL580) and experimental samples were labeled with Cy5-CTP (NEL581). Labeled cRNA samples were cleaned up using the Qiagen RNeasy Cleanup kit protocol with an additional wash step, then quantified using a Nanodrop. Labeled cRNA was fragmented and allowed to hybridize to Agilent microarray slides (8x15k, AMADID:017566) for 17 hours at 65°C and 20 RPM. After hybridization, slides were washed successively with wash buffer 1 for 1 minute, washer buffer 2 for 1 minute, and acetonitrile for 30 seconds. Slides were scanned using the Agilent High-Resolution Microarray Scanner. After scanning, Feature Extraction software was used to map spots to the specific genes. Resulting microarray intensity data were stored at the PUMA Database (http://puma.princeton.edu) during analysis. Source data for the microarray experiments were deposited at the Gene Expression Omnibus, NCBI, and can be accessed using GEO number GSE220125.

Microarray analysis

Sample and reference channel intensities were first floored to a value of 350. Once log₂ ratios were computed between samples and reference, the data were time-zero transformed. Data were hierarchically clustered in the Cluster 3.0 software package with average linkage using the Pearson correlation distance as the metric of similarity between genes. Gene Ontology terms were determined using YEASTTRACT.

RNA-seq

Cells were induced to sporulate as described above and samples were taken at the indicated time points. Cells were lysed using bead beating and the lysis buffer included in the Qiagen RNeasy Kit. After lysis, samples were cleared by centrifugation and RNA was purified using the Qiagen RNeasy kit with on-column DNase treatment. RNA samples were then sent to Novogene Corporation for library prep and mRNA sequencing using an Illumina based platform (PE150). Resulting data was analyzed using the open access Galaxy platform [46]. Reads were first mapped to the yeast genome using BWA-MEM and then counted using htseq-count. Differential expression analysis was conducted using DESeq2 and heat maps were made using Cluster 3.0 and Java Tree View.

qPCR

Cells were induced to sporulate as described above and samples were taken at the indicated time points. RNA was harvested as described in the section on RNA-seq. cDNA libraries were constructed using the High-Capacity cDNA Reverse Transcription kit (Applied Biosystems) with 10 μl of the total RNA sample. The concentration of the resulting cDNA was measured using a nanodrop. qPCR reactions were set up using Power SYBR Green PCR Master Mix (Applied Biosystems) with 50 ng of total RNA. The reactions were run on a CFX96 Real-Time System (BioRad) with reaction settings exactly as described in the master mix instructions with the only change being the addition of a melt curve at the end of the program. Results were analyzed using CFX Maestro. Primer sequences were as follows: PGK1 Forward 5’-CTCACTCTTCTATGGTCGCTTTC-3‘, PGK1 Reverse 5’-AATGGTCTGGTTGGGTTCTC-3’, GAS4 Forward 5’-GACCTGGAAGGAGAAGAAGAACAAG-3’, GAS4 Reverse 5’-ACAATGGGCCGGAAATAGAG-3’, SPS2 Forward 5’-GCCGGTCGTTCGATCATAA-3’, SPS2 Reverse 5’-CATTGTCAGTTTCCTGCTTTCC-3’.

Protein extraction by alkaline lysis

Optical density was measured and a total of 6 OD₆₀₀ units was collected for each time point. Cells were pelleted and stored at -80°C. Cell lysates were prepared by adding 150 μl lysis buffer (1.85 M NaOH, 1/100 B-ME, 1/50 protease inhibitors) followed by a 10-minute incubation on ice. After incubation, 150 μl of 50% TCA (Sigma-Aldrich T9159) was added. After a 10-minute incubation at 4°C, samples were spun for 15 seconds at 15000 RPM and supernatant was aspirated. The remaining pellet was washed with one ml acetone, briefly spun, and acetone aspirated. 100 μl of 2x sample buffer (ThermoFisher NP0007) with 10% B-ME were added to the protein pellets, mixed well, and boiled for 5 minutes.

Immunoblotting

For each lane, 10 μl of protein extracts were added to 8% Bis-Tris acrylamide gels. Protein ladder (Precision Plus Protein Standard from Bio Rad, 1610374) was used for determining band sizes. Electrophoresis was run at 60 volts for 30 minutes through the stacking gel and at 150 volts until the samples moved through the resolving gel. Gels were transferred to PVDF membranes using a semi-dry transfer apparatus (TransBlot SD BioRad) and a standard Tris-Glycine transfer buffer without methanol at 16 volts for 36 minutes. Membranes were blocked with 10% milk for 30 minutes, followed by primary antibody (anti-MYC 1:1000 (9E10), anti-HA 1:1000 (12CA5), anti-Ndt80p (gift of Michael Lichten) 1:10,000, anti-Kar2p [47] 1:5000) in 0.1% TBST for 1 hour with rocking at room temperature. Membranes were washed three times for 10 minutes with TBST. Membranes were incubated with secondary antibody (Donkey anti-mouse (Jackson ImmunoResearch) IgG 1:10,000) in 1% milk with rocking for 30 minutes. Membranes were then washed three times for 10 minutes with TBST. Immobilon Western HRP substrate (Millipore) was added and incubated for 5 minutes before being imaged using the G-Box from SynGene. Densitometry was conducted using ImageJ. All westerns were run at least twice with each run being a unique biological replicate.

Trypsin digest and 8-step fractionation mass spectrometry

For each time point, the optical density was measured and a total of 30 OD₆₀₀ was aliquoted. Samples were pelleted, washed with water, and flash frozen by liquid nitrogen. Frozen yeast pellets were resuspended in lysis buffer (6M guanidium hydrochloride, 10mM TCEP, 40mM CAA, 100mM Tris pH 8.5). Cells were lysed by sonication using 5x 30s pulses with 1 min rest in ice between pulses. Samples were then heated to 95°C for 15 min, and allowed to cool in the dark for 30 min. Samples were then centrifuged, and lysate removed to a fresh tube. Lysate was then diluted 1:3 with digestion buffer (10% CAN, 25mM Tris pH 8.5) containing LysC (1:50) and incubated at 37°C for 3 hours. Samples were then further diluted to 1:10 with digestion buffer containing Trypsin (1:100) and incubated O/N at 37°C. TFA was added to 1% final. Samples were then centrifuged, and digested lysate removed to a new tube. Samples were desalted on C18 cartridges (Oasis, Waters) as per manufacturer protocol. Dried down peptide samples were then fractioned using High pH Reversed-Phase peptide fraction kit (Pierce) into 8 fractions using manufacturer’s instructions. Fractions were dried completely in a Speedvac and resuspended with 20μl of 0.1% formic acid pH 3. 5ul was injected per run using an Easy-nLC 1000 UPLC system. Samples were loaded directly onto a 45cm long 75μm inner diameter nano capillary column packed with 1.9μm C18-AQ (Dr. Maisch, Germany) mated to metal emitter in-line with a Q-Exactive (Thermo Scientific, USA). The mass spectrometer was operated in data dependent mode with the 700,00 resolution MS1 scan (400–1800 m/z), AGC target of 1e6 and max fill time of 60ms. The top 15 most intense ions were selected (2.0 m/z isolation window) for fragmentation (28 NCE) with a resolution of 17,500, AGC target 2e4 and max fill time of 60ms. Dynamic exclusion list was evoked to exclude previously sequenced peptides for 120s if sequenced within the last 10s.

Raw files were searched with MaxQuant (ver 1.5.3.28) [48], using default settings for LFQ data. Carbamidomethylation of cysteine was used as fixed modification, oxidation of methionine, and acetylation of protein N-termini were specified as dynamic modifications. Trypsin digestion with a maximum of 2 missed cleavages were allowed. Files were searched against the yeast SGD database download 13 Jan 2015 and supplemented with common contaminants. Results were imported into the Perseus [49] workflow for data trimming and imputation. Final data were exported as a table. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE [50] partner repository with the dataset identifier PXD043798.

Recombination assay

Diploid wild type and kar4Δ/Δ strains carrying a P_MFA1-HIS3 cassette integrated in place of the CAN1 gene in the MATα parent were transformed with a high-copy number plasmid containing IME1 and the URA3 gene as a selectable marker. Strains were induced to sporulate as described above and samples were taken across a time course of meiosis (0-, 8-, 12-, and 24-hours post movement into sporulation media). At each time point, 1 OD unit of cells was removed and serially diluted 10-fold three times. Four microliters of each serial dilution (10°, 10⁻¹, 10⁻², and 10⁻³) were plated on either SC-Ura (growth plate) or SC-Ura -His +can (recombination selection plate). Plates were allowed to grow for 2–3 days at 30°C before pictures were taken.

Supporting information

S1 Fig. RNA-seq derived meiotic transcriptome of kar4Δ/Δ across overexpression conditions.

Heatmap of RNA-seq data across a time course of meiosis (0, 2, 4, 6, 8, and 12 hours) in wild type and kar4Δ/Δ with either pIME1/pRIM4, Pzev-IME1/pRIM4, or Pzev-IME1/Pzev-RIM4. Expression was normalized to wild type pre-induction of sporulation (t = 0). Genes were clustered in Cluster3.0 and the heatmaps were constructed with Java TreeView. Note that genes are clustered differently from Fig 1. Source data for heatmap can be found in the supplementary file S3 Data.

(TIF)

S2 Fig. Ime2p levels are not impacted in kar4Δ/Δ under endogenous expression conditions.

(A) Western blots of Ime2p-13MYC across a meiotic time course in wild type and kar4Δ/Δ. Kar2p is used as a loading control. (B) IME2 RNA-seq normalized counts from wild type and kar4Δ/Δ. Counts were normalized using the standard normalization method in DESeq2. Error bars represent standard deviation between two biological replicates, which is equal to the range divided by the square root of 2. (C) Quantification of western blots in A. Error bars represent the standard deviation between two biological replicates, which is equal to the range divided by the square root of 2. (D) Quantification of western blots in Fig 4A. Error bars represent the standard deviation between two biological replicates, which is equal to the range divided by the square root of 2. (E) Quantification of western blots in Fig 4C. Error bars represent the standard error between two biological replicates, which is equal to one half the range.

(TIF)

S1 Table. Strains used for this study.

All auxotrophic markers are standard BY alleles.

(DOCX)

S2 Table. Plasmids used in this paper.

(DOCX)

S1 Data. Data underlying all graphs presented in figures in the paper and quantification used to determine changes in Gas4p and Sps2p levels.

(XLSX)

S2 Data. Data from microarray experiments used to construct the heat-maps in Figs 1A and 3B.

The data is presented as log₂ fold-changes from a starvation condition control sample.

(XLSX)

S3 Data. Data from RNA-seq experiments used to construct the heat-map in S1 Fig.

The data is presented as log₂ fold-changes from a starvation condition control sample.

(XLSX)

S4 Data. Data from microarray experiments used to construct the heat-maps in Figs 2B and 3A.

These heat-maps display the expression of the 136 Ime1p dependent genes determined in this study and their identity and expression data are contained in this data file. Genes highlighted in green are those that are conserved between the Ime1p dependent genes and the core meiotic Ume6p regulon.

(XLSX)

S5 Data. Imputed intensities from the mass spectrometry experiment analyzing the meiotic proteome in wild type and kar4Δ/Δ.

(XLSX)

S6 Data. DESeq2 log₂ fold-changes between wild type and kar4Δ/Δ at each individual time point from the RNA-seq experiments.

(XLSX)

Data Availability

Source data for the microarray and RNA-seq experiments can be found using GEO accession numbers GSE220125 and GSE221451, respectively. Mass spectrometry data has been uploaded to ProteomeXchange: PXD043798.

Funding Statement

This work was supported by NIH grants GM037739 and GM126998 to MDR. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.