Rhn1, a Nuclear Protein, Is Required for Suppression of Meiotic mRNAs in Mitotically Dividing Fission Yeast

In the fission yeast Schizosaccharomyces pombe, many meiotic mRNAs are transcribed during mitosis and meiosis and selectively eliminated in mitotic cells. However, this pathway for mRNA decay, called the determinant of selective removal (DSR)-Mmi1 system, targets only some of the numerous meiotic mRNAs that are transcribed in mitotic cells. Here we describe Rhn1, a nuclear protein involved in meiotic mRNA suppression in vegetative fission yeast. Rhn1 is homologous to budding yeast Rtt103 and localizes to one or a few discrete nuclear dots in growing vegetative cells. Rhn1 colocalizes with a pre-mRNA 3′-end processing factor, Pcf11, and with the 5′–3′ exoribonuclease, Dhp1; moreover, Rhn1 coimmunoprecipitates with Pcf11. Loss of rhn1 results in elevated sensitivity to high temperature, to thiabendazole (TBZ), and to UV. Interestingly, meiotic mRNAs—including moa1+, mcp5+, and mug96+—accumulate in mitotic rhn1Δ cells. Accumulation of meiotic mRNAs also occurs in strains lacking Lsk1, a kinase that phosphorylates serine 2 (Ser-2) in the C-terminal domain (CTD) of RNA polymerase II (Pol II), and in strains lacking Sen1, an ATP-dependent 5′–3′ RNA/DNA helicase: notably, both Lsk1 and Sen1 have been implicated in termination of Pol II-dependent transcription. Furthermore, RNAi knockdown of cids-2, a Caenorhabditis elegans ortholog of rhn1 +, leads to elevated expression of a germline-specific gene, pgl-1, in somatic cells. These results indicate that Rhn1 contributes to the suppression of meiotic mRNAs in vegetative fission yeast and that the mechanism by which Rhn1 downregulates germline-specific transcripts may be conserved in unicellular and multicellular organisms.


Introduction
Many kinds of compartments, collectively designated nuclear bodies, are present in nuclei of eukaryotic cells [1,2]. Nuclear bodies comprise proteins and RNAs and are believed to have important roles in various nuclear functions, including RNA processing, snRNP assembly, and transcription. For example, nuclear speckles are thought to function in the assembly and storage of pre-mRNA splicing factors [3]. Similarly, paraspeckles participate in the regulation of gene expression by retaining mRNAs that carry inverted repeats and are extensively A-to-I edited; moreover, these paraspeckles may be useful as a structural marker in discriminating between pluripotent and differentiated cells [4]. Apparently, many nuclear proteins involved in specific tasks are assembled into discrete nuclear compartments to perform their respective functions efficiently. However, the mechanisms of nuclear body assembly and the exact function of each type of these structures are not fully understood [5].
RNA metabolism is an important activity in both prokaryotes and eukaryotes. RNA degradation pathways are basically classified as follows: (1) RNA processing of primary transcripts to produce mature mRNAs; (2) regulated turnover of mRNAs and non-coding RNAs to modulate gene expression; and (3) quality control systems that remove defective RNAs and RNA-protein complexes from cells [6,7]. In eukaryotic cells, these tasks require many exoribonuclease, endoribonucleases, RNA helicases, poly(A) polymerases, chaperones, and small RNAs [6,7].
Recent studies have highlighted the fact that several mRNA degradation factors play key roles in other cellular mechanisms, including termination of RNA polymerase II (Pol II) transcription [7]. In the budding yeast Saccharomyces cerevisiae, serine 2 (Ser-2) phosphorylation of heptamer repeat sequences in the C-terminal domain (CTD) of Pol II facilitates in vitro interactions between this CTD and Rtt103, a protein with a CTD-interacting domain (CID), and between this CID and Pcf11, a pre-mRNA 39-end processing factor [8,9]. Rtt103 also associates with Rat1, a 59-39 exoribonuclease, and with Rai1, a Rat1-interacting protein [8]. Rat1 degrades transcripts with uncapped 59-ends generated by pre-mRNA cleavage, and it eventually promotes the dissociation of Pol II from template DNAs [10]. Thus Rat1, Rai1, Rtt103, and polyadenylation factors contribute to, at least in part, termination of Pol II-dependent transcription. This termination mechanism has been designated the torpedo model and is apparently conserved in human because RNAi knockdown of Xrn2, a Rat1 homolog in higher eukaryotes, results in a decrease in transcriptional termination efficiency [10,11].
In the fission yeast Schizosaccharomyces pombe, selective mRNA decay, as well as transcriptional control, has emerged as a fundamental regulatory mechanism in the control of meiotic gene expression. Many genes that are upregulated during meiosis are transcribed even in vegetative cells, but these meiotic transcripts are subjected to mRNA degradation in the vegetative cells [12][13][14]. This mRNA degradation pathway requires specific sequences (designated determinant of selective removal (DSR)), an RNAbinding protein (Mmi1), a 39-end processing factor (Rna15), the canonical poly(A) polymerase (Pla1), a nuclear poly(A)-binding protein (Pab2), a zinc-finger protein (Red1), and the nuclear RNase complex exosome; taken together, these requirements indicated that the DSR system is a poly(A)-dependent mRNA degradation pathway [12,[14][15][16][17][18]. In addition, DSR-mediated mRNA elimination may be inactivated during meiosis, and inactivation of the DSR pathway may allow meiotic mRNAs to be translated [12]. Thus, in fission yeast, the untimely expression of meiosis-specific mRNAs is believed to be prevented via this DSR-mediated mRNA decay. However, it is still unclear whether transcriptional control and DSR-mediated selective mRNA elimination are sufficient for regulating all of the genes that are upregulated during meiosis, and there may be additional mechanism(s) necessary for the control of meiotic gene expression.
We used protein localization methods to identify several factors that localize to nuclear or cytoplasmic foci in fission yeast. We previously reported that one such protein, Red1, localizes to cleavage bodies and promotes selective elimination of meiosisspecific mRNAs in vegetative fission yeast [15]. Here, we report that Rhn1, another factor that localizes to nuclear dots, contributes to the suppression of meiotic genes in vegetative fission yeast. Characterization of Rhn1 should help to elucidate the complex mechanisms that suppress untimely expression of numerous meiotic mRNAs in vegetative fission yeast, as well as in somatic cells of higher eukaryotes, including the nematode Caenorhabditis elegans.

Rhn1 localizes to the nucleus and forms nuclear dots
To identify novel factors that regulate RNA metabolism in fission yeast, we utilized a subcellular localization database [19] and selected candidate proteins that localize to the nucleus. We then constructed fission yeast strains that each expressed one candidate protein as a fusion protein that had either a carboxyterminal tdTomato tag, an mCherry tag, or a GFP tag [20,21]. Among these candidate proteins, the previously uncharacterized protein encoded by SPBC337.03 formed one or a few distinct dots under vegetative growth conditions ( Figure 1A). We also found that the signal from the SPBC337.03-GFP fusion protein was distributed evenly throughout nuclei of vegetative cells ( Figure 1A). SPBC337.03 was predicted to encode a 387 amino acid protein that had homology to budding yeast Rtt103 ( Figure 1B); we named this gene rhn1 + (Rtt103 homolog that localizes to the nucleus 1). S. cerevisiae Rtt103 has a CID (CTD-interacting domain) that associates with the CTD of Pol II, and was originally identified as a regulator of Ty1 transposition [22]. Rhn1, like Rtt103, contained a CID ( Figure 1C); this finding indicated that Rhn1 may interact with Pol II. rhn1D cells are sensitive to high temperature, TBZ, and UV To characterize the function of Rhn1, we constructed an rhn1 deletion strain (rhn1D) and analyzed the phenotypes of rhn1D cells. We observed neither severe growth defects nor growth retardation of rhn1D cells at 18 or 30uC; however, rhn1D cells did not at all grow at 37uC ( Figure 2A); this finding indicated that Rhn1 was necessary for growth at high temperature. Analyses of serial dilutions of cell cultures also revealed that rhn1D cells were more sensitive to thiabendazole (TBZ), a microtubule-destabilizing drug, and UV than were wild-type cells ( Figure 2B); therefore, Rhn1 may function to maintain genomic stability.

rhn1D cells are defective in mating
Fission yeast cells that have Ser-2 to alanine substitutions in the Pol II CTD are sterile [23][24][25], and Rtt103 binds to phosphorylated Ser-2 of CTD [8]; therefore, we hypothesized that Rhn1 had a role in mating. To test this hypothesis, we compared sporulation of wild-type cells and rhn1D cells and found that iodine staining, which selectively labels sporulated cells, was less intense in rhn1D cells than in wild-type cells ( Figure 2C). Consistent with  [20,21] in YEA cultures were observed using fluorescent microscopy. Representative images are shown. (B) Amino acid sequence alignments of Rhn1 and its budding yeast homolog Rtt103. ClustalW2 (http://www.ebi.ac.uk/Tools/clustalw2/index.html) was used to align these sequences. Black and gray shadings represent identical and similar amino acids, respectively. (C) Schematic representation of Rhn1. CID, CTD-interacting domain. doi:10.1371/journal.pone.0042962.g001 this result, mating efficiency was lower in rhn1D cells than in wildtype cells (70.4% vs 52.0%) ( Figure 2C). The reduction in iodine staining may have been due to impaired mating in rhn1D cells, and it indicated that Rhn1 was required for efficient mating.

Rhn1 contributes to the suppression of meiotic mRNAs during mitosis
We next performed transcriptional expression profiling of vegetative rhn1D cells to investigate the role(s) of Rhn1 in controlling steady-state levels of mRNAs. Using a microarray technique, we identified 51 mRNAs with steady-state levels that were increased more than 1.6-fold in rhn1D cells relative to wildtype cells ( Figure 4A and Table S1). The majority of these mRNAs, 35 out of 51 (68.6%), were previously reported to be meiotic mRNAs that were upregulated during meiosis [28,29]. Among the 51 mRNAs, 10 (19.6%) were non-meiotic mRNAs, and 6 (11.8%) have not yet been as meiosis-related ( Figure 4A and Table S1). We also identified 17 mRNAs with steady-state levels that were lower (,0.5-fold reduction) in rhn1D than in wild-type cells ( Figure 4C and Table S2). Many (14 out of 17 mRNAs) of these downregulated genes were located in subtelomeric regions, and 8 out of 17 mRNAs were classified as meiotic genes. Some of these array results were confirmed in conventional RT-PCR experiments ( Figure 4B and D). These analyses indicated that Rhn1 functions in vegetative cells predominantly to downregulate meiosis-related genes and to contribute to the maintenance of expression of several subtelomeric genes.  Total cell lysates prepared from strains expressing Rhn1-PK, Rhn1-PK and Dhp1myc, Rhn1-PK and Din1-myc, or Rhn1-PK and Pcf11-myc were subjected to immunoprecipitation using anti-myc; immnoprecipitates and inputs were analyzed on Western blots. The asterisk indicates the degradation product of Dhp1-myc protein. Note that, because of its low abundance, full-length Dhp1-myc was only visible when it was concentrated by immunoprecipitation. doi:10.1371/journal.pone.0042962.g003 Rhn1 is dispensable for DSR-mediated mRNA elimination Meiotic mRNAs such as mei4 + and rec8 + are transcribed, but are then subjected to DSR-dependent mRNA degradation, in mitotically dividing cells. This degradation system requires the DSR degradation sequences (UUAAAC/UCAAAC) and several protein factors including Mmi1, Red1, Pab2, and Dhp1 [12][13][14][15][16][17][18]. Considering that some of the Red1/Mmi1-target mRNAs (e.g. moa1 + , mug96 + , and mcp5 + ) accumulated in rhn1D ( Figure 4A and Table S1) and that Rhn1 colocalized with Dhp1 ( Figure 3A), Rhn1 might be involved in DSR-dependent mRNA decay. To assess this possibility, we first compared Rhn1-regulated genes (those genes that were upregulated or downregulated in rhn1D cells relative to wild-type cells) with Red1-regulated genes [15]; we found that the set of Rhn1-regulated genes overlapped with that of Red1regulated genes. Specifically, 31 of 51 mRNAs (61.0%) upregulated in rhn1D cells were also upregulated in red1D cells (p,1.4610 238 ) ( Figure 5A), while 7 of 17 mRNAs (41.2%) downregulated in rhn1D cells were also downregulated in red1D cells (p,2.5610 212 ) ( Figure 5B). These results indicated that Rhn1 and Red1 suppress or promote the expression of common target mRNAs in vegetative cells and that the two proteins work in a parallel or in overlapping pathways.
Although Rhn1 and Red1 share target mRNAs, the mRNA profiles of rhn1D cells indicated that Rhn1 did not target four representative DSR-containing mRNAs, mei4 + , rec8 + , ssm4 + , and spo5 + (Table S1). We then examined whether Rhn1 was essential for DSR-dependent RNA elimination. We introduced the ura4 + -DSR marker into wild-type and rhn1D cells ( Figure 5C); this marker comprises the thiamine-repressive nmt81 promoter, the ura4 + open reading frame, and the DSR region of mei4 + [12]. Wild-type and rhn1D cells carrying this marker did not grow on uracil-deficient (PMG-ura) plates; in contrast, red1D cells with the ura4 + -DSR grew well on PMG-ura plates ( Figure 5D), indicating that ura4 + -DSR was stable in red1D, but not in wild-type or rhn1D cells. This conclusion was confirmed by RT-PCR analyses; ura4 + -DSR mRNA accumulated only in red1D cells, not in wild-type or rhn1D cells ( Figure 5E). Moreover, vegetative rhn1D cells did not accumulate four DSR-containing meiotic mRNAs, mei4 + , spo5 + , ssm4 + , and rec8 + (Figure 5F), indicating that Rhn1 was not essential for DSR-dependent meiotic mRNA degradation.
Lsk1, a subunit of Pol II CTD Ser-2 kinase, also suppresses meiotic mRNAs and genetically interacts with Rhn1 S. cerevisiae Rtt103 binds to the Pol II CTD, which is phosphorylated at Ser-2 in vitro [8]. Lsk1 is a subunit of Pol II CTD Ser-2 kinase that plays a crucial role in proper meiotic processes [23,24,30]. Therefore, we hypothesized that Lsk1, like Rhn1, might play a role in the suppression of meiotic mRNAs during mitosis. RT-PCR analyses showed that mitotic lsk1D cells accumulated elevated levels of meiotic mRNAs, including mug96 + and mug37 + ( Figure 6A). We also tested whether Sen1, an ATPdependent 59-39 RNA/DNA helicase [31], contributed to the suppression of meiotic mRNAs. The levels of mug96 + transcripts, but not SPBPB2B2.03c or mug37 + , transcripts significantly increased in mitotic sen1D cells ( Figure 6A). Our results also indicated that some of meiotic mRNAs that accumulated in mitotic rhn1D cells did not accumulate markedly in mitotic lsk1D cells or mitotic sen1D cells ( Figure 6A and data not shown). We next tested for a genetic interaction between Rhn1 and Lsk1 to examine the relationship between the two proteins. We observed a synthetic growth defect at high temperatures (34 to 36uC) in rhn1Dlsk1D cells ( Figure 6B). Taken together, these analyses indicated that Rhn1, Lsk1, and Sen1 mediated downregulation of several meiotic mRNAs in overlapping or parallel pathways during mitosis.
About 2000 genes are upregulated during nitrogen starvation and meiotic processes [28]. Mmi1, Red1, and Pab2 apparently target only a fraction of these meiosis-related genes for selective elimination in mitosis [12,14,15,17]. Therefore, we believe that there are at least four mechanisms that control proper meiotic mRNA expression and that these mechanisms are as follows: (1) transcriptional activation by Ste11 during meiosis [32]; (2) mRNA degradation by Red1, Mmi1, and other players in vegetative cells [33]; (3) transcriptional repression by Fkh2-mediated antisense transcription in vegetative cells [34]; and (4) a novel mechanism regulated by factors required for Pol II transcription termination pathways in vegetative cells (this study). The diversity of these mechanisms illustrates the complexity of meiotic mRNA regulation in fission yeast.

Rhn1 promotes ste11 + expression in response to nitrogen starvation
Phosphorylation of Ser-2 in the Pol II CTD, as well as Lsg1, contributes to Ste11 expression in response to meiotic induction [23][24][25]. Interestingly, we found that Lsk1, which physically binds to Lsg1 and regulates Lsg1 localization [30], suppressed three Rhn1-regulated meiotic genes in vegetative cells. We then investigate whether Rhn1 was required for ste11 + induction upon nitrogen starvation. RT-PCR analyses demonstrated that the level of ste11 + transcript induced in rhn1D cells was relatively lower than that induced in wild-type cells ( Figure 7A). Moreover, the expression of exogenous ste11 + enhanced the mating efficiency of homothallic rhn1D cells ( Figure 7B). Together, these findings indicated that, upon the initiation of meiotic processes, Rhn1 was involved in proper ste11 + expression and that the reduced mating efficiency observed in rhn1D results from the incomplete induction of ste11 + .
Expression of ste11 + in response to nitrogen starvation depends on phosphorylation of Ser-2 in the Pol II CTD, and this  rhn1D and red1D strains (B). We examined the expression of 4977 genes for these comparisons. We caluculated p-values to assess the statistical significance of the overlap between these groups; these p-values are shown under the respective diagrams. (C) Schematic representation of the ura4 + -DSR construct. The DSR region derived from mei4 + was fused to the ura4 + gene, and transcription of the ura4 + -DSR mRNA was driven by the thiamine-repressive nmt81 promoter. This construct was integrated into the lys1 + locus. (D) The effect of ura4 + -DSR mRNA elimination in rhn1D cells. Serial dilutions of wild-type (WT), rhn1D, and red1D cells carrying the ura4 + -DSR were spotted onto minimal media plates without thiamine (PMG) and onto minimal media plates lacking uracil and thiamine (PMG-ura); the plates were then incubated at 30uC for 2 to 3 days. (E) No marked accumulation of ura4 + -DSR transcripts was evident in rhn1D cells. Total RNA samples were prepared from wild-type (WT), rhn1D, or red1D cells that carries both ura4 + -DSR (DSR) and the ura4 + minigene (ura4DS/E (DS/E)) grown in PMG media without thiamine. The levels of ura4 + -DSR and ura4DS/E mRNAs were determined by RT-PCR. The ura4DS/E transcript was used as an internal control. RT (-), no reverse transcription. (F) The steady-state levels of mei4 + , spo5 + , ssm4 + , and rec8 + mRNAs were not higher in mitotic rhn1D cells than in wild-type cells. The total RNA samples used in (E) were also subjected to RT-PCR analyses. The 28S rRNA (28S) was used as a control. RT (-), no reverse transcription. doi:10.1371/journal.pone.0042962.g005 phosphorylation is mediated by the Lsk1-Lsg1 CTD kinase complex, Wis1 (a MAP kinase kinase), and Sty1 (a MAP kinase) [23,24,30]. Since Rhn1 also contributes to ste11 + expression during meiosis, we hypothesize that Rhn1 promotes ste11 + transcription or maintains CTD Ser-2 phosphorylation via binding to the Ser-2 modification. Further analyses are required to test these hypotheses.
Knockdown of cids-2 by RNAi results in elevated expression of the germline-specific gene pgl-1 in C. elegans Meiotic mRNAs are suppressed in vegetative fission yeast, and aberrant inactivation of the meiotic mRNA elimination system, ectopic activation of positive regulators of meiosis, or both result in ectopic of meiosis-specific mRNAs in mitotic cells [33]. We previously suggested that meiotic mRNA elimination may be evolutionarily conserved, in particular in the nematode C. elegans [15], because a similar phenomenon, called ''soma-to-germline transformation'' has been reported [35][36][37]. A BLAST search showed that C. elegans cids-1 and cids-2 were most closely related to rhn1 + . Both cids-1 and cids-2 encode Rtt103-related proteins and may play roles in 39-end cleavage of pre-mRNAs [38]. To examine whether these putative 39-end cleavage factors might have roles in suppression of germline-specific transcripts, we performed cids-2 knockdown by RNAi feeding (see Materials and Methods). We chose Cids-2 because Rhn1 was more similar to Cids-2 than to Cids-1. We monitored gfp::pgl-1 expression from the pie-1 promoter, which is predominantly active in germline cells, in cids-2 RNAi animals. We found that cids-2 RNAi resulted in higher expression of gfp::pgl-1 in hypodermal somatic cells, including the head hypodermis of adult worms, than did control RNAi ( Figure 8A). We quantified GFP signals in the head regions, and the intensity of GFP in cids-2 RNAi worms was significantly higher than that in control RNAi worms ( Figure 8B). The increased GFP signal was specific to cids-2 RNAi; two independent RNAi constructs targeting different cids-2 regions resulted in the same phenotype ( Figure 8B and data not shown). The increase in gfp::pgl-1 expression resulting from cids-2 knockdown was similar to that resulting from knockdown of lin-35, which is essential for preventing soma-to-germ transformation [37]. To further confirm the role of Cids-2 in the suppression of germ-specific transcripts, we performed cids-2 RNAi in glp-4(bn2) mutant worms ( Figure 8C, left); this mutation has been widely used to generate germlinedeficient animals and to assess germline versus somatic gene expression [39]. RT-PCR analysis indicated that the endogenous pgl-1 transcripts accumulated to abnormally high levels in germline-deficient worms in which cids-2 RNAi or lin-35 RNAi was performed ( Figure 8C, right panels). These results indicated  that meiotic mRNA suppression by Rhn1-related proteins might be conserved in fission yeast and worms.
Our data indicate that 39-end cleavage factors also contribute to the suppression of soma-to-germ transformation and that the meiotic mRNA suppression system found in fission yeast might be conserved in multicellular organisms. It would be of great interest to investigate whether selective suppression of certain kinds of mRNAs prevents differentiation in mammals.

Materials and Methods
Fission yeast strains, media, plasmids, and imaging The general genetic methods used in this study were as previously described [40]. Complete medium (YEA), minimal medium (PMG), PMG without uracil (PMG-ura), PMG lacking leucine (PMG-leu), and nitrogen-free PMG (PMG-N) were used to culture the cells [40,41].
A strain carrying the ura4 + -DSR marker in a wild-type background was provided by Dr. M. Yamamoto (University of Tokyo), and the parental wild-type strain was obtained from the National BioResource Project of Japan (NBRP) at Osaka City University. Other strains expressing a fusion protein with an mCherry, tdTomato [20], GFP [21], PK, or myc epitope-tag and the deletion strains were constructed using PCR-based methods as described previously [42]. All the strains used in this study are listed in Table S3.
Two plasmids, pREP1 and pREP1-ste11 + , were obtained from Dr. M. Yamamoto and NBRP, respectively. The mating efficiency of each strain was calculated as described [15]. For ste11 + expression in rhn1D, we did not use media that lacked thiamine because the thiamine-repressive nmt1 promoter of pREP1 was strong; therefore, the basal level of ste11 + expression under repressed condition was enough to rescue the mating efficiency. We also found that ste11 + induction caused by depleting thiamine was toxic to rhn1D cells.
An Axio Imager M1 microscope (Carl Zeiss MicroImaging) was used for fluorescence microscopy. The raw images were processed using AxioVision software (Carl Zeiss MicroImaging).

RNA analyses of fission yeast
Total RNA samples were prepared from exponentially growing yeast cells in complete or minimal media using the mirVana TM miRNA isolation kit (Life Technologies). DNA contamination in the RNA samples was removed using a TURBO DNA-free TM kit (Life Technologies). The DNase-treated total RNA samples were then subjected to RT-PCR using the PrimeScriptH II 1st strand cDNA synthesis kit (TaKaRa Bio Inc.) and TaKaRa ExTaqH (TaKaRa Bio Inc.). Quantification of RT-PCR results was performed using Quantity OneH software (Bio-Rad Laboratories, Inc.) For microarray analyses, the S. pombe expression 4x72K array (090408 Spom exp X4, Roche-NimbleGen) was used. The microarray analyses were independently preformed twice according to the manufacturer's protocol. The data from wild-type cells were compared with those from rhn1D cells; the average in the fold changes from two independent experiments were calculated, and the transcripts that showed more than a 1.6-fold increase or a 0.5fold decrease were selected as ''increased'' or ''decreased'' mRNAs, respectively. The lists of increased and decreased genes are shown in Tables S1 and S2, respectively. The statistical significance of overlap of two groups was determined using a calculation program (http://nemates.org/MA/progs/overlap_stats.html).
For the RT-PCR analyses shown in Figure 8C, the glp-4(bn2) worms, which have a temperature-sensitive sterile phenotype, were maintained at 15uC, but were shifted to 25uC as synchronized L1s, which is when feeding RNAi was initiated. Total RNA was extracted from RNAi-induced synchronized glp-4(bn2) L4 worms using the RNAiso Plus reagent (TaKaRa Bio Inc.). Reverse transcription and the subsequent PCR were performed using ReverTra AceH with gDNA remover (TOYOBO Co., Ltd.) and a Thunderbird TM SYBRH qPCR kit (TOYOBO Co., Ltd.), respectively. The PCR primers used to amplify pgl-1 and act-1 were previously described [46,47]. Table S1 The genes upregulated (.1.6) in rhn1D.