A role for the transcription factor Mca1 in activating the meiosis-specific copper transporter Mfc1

When reproduction in fungi takes place by sexual means, meiosis enables the formation of haploid spores from diploid precursor cells. Copper is required for completion of meiosis in Schizosaccharomyces pombe. During the meiotic program, genes encoding copper transporters exhibit distinct temporal expression profiles. In the case of the major facilitator copper transporter 1 (Mfc1), its maximal expression is induced during middle-phase meiosis and requires the presence of the Zn6Cys2 binuclear cluster-type transcription factor Mca1. In this study, we further characterize the mechanism by which Mca1 affects the copper-starvation-induced expression of mfc1+. Using a chromatin immunoprecipitation (ChIP) approach, results showed that a functional Mca1-TAP occupies the mfc1+ promoter irrespective of whether this gene is transcriptionally active. Under conditions of copper starvation, results showed that the presence of Mca1 promotes RNA polymerase II (Pol II) occupancy along the mfc1+ transcribed region. In contrast, Pol II did not significantly occupy the mfc1+ locus in meiotic cells that were incubated in the presence of copper. Further analysis by ChIP assays revealed that binding of Pol II to chromatin at the chromosomal locus of mfc1+ is exclusively detected during meiosis and absent in cells proliferating in mitosis. Protein function analysis of a series of internal mutants compared to the full-length Mca1 identified a minimal form of Mca1 consisting of its DNA-binding domain (residues 1 to 150) fused to the amino acids 299 to 600. This shorter form is sufficient to enhance Pol II occupancy at the mfc1+ locus under low copper conditions. Taken together, these results revealed novel characteristics of Mca1 and identified an internal region of Mca1 that is required to promote Pol II-dependent mfc1+ transcription during meiosis.


Introduction
Meiosis is a specialized type of cell division through which sexually reproducing organisms produce haploid gametes from diploid precursor cells [1][2][3]. In the model organism Schizosaccharomyces pombe, a diploid precursor cell is formed by conjugation of two haploid cells of opposite mating types. Following conjugation, the two haploid nuclei fuse and that subsequently induces replication of chromosomal DNA, generating pairs of homologous chromosomes. After meiotic recombination between homologous chromosomes that increases genetic diversity, two successive meiotic divisions occur in which homologous chromosomes (meiosis I) and then sister chromatids (meiosis II) are segregated without any intervening DNA replication phase. These two divisions are followed by a differentiation program that includes forespore membrane biogenesis and production of four mature haploid spores (gametes) that are enclosed in an ascus [4,5]. A number of studies have shown that transition metals such as copper, iron, and zinc are essential for meiosis [6][7][8][9][10]. In murine oocytes, zinc insufficiency blocks the exit from meiosis I [10]. In the case of female porcine gametes, zinc-insufficient oocytes fail to segregate homologous chromosomes, exhibiting an arrest at metaphase I [11]. Similarly, studies with the yeast S. pombe have shown that cells undergo a meiotic block at metaphase I when copper or iron concentrations are insufficient [6,7]. In order to identify proteins involved in the mechanisms by which transition metal insufficiency perturbs the meiotic program, studies have been undertaken using different model systems. Among them, S. pombe is of special interest as growth conditions and temperature-sensitive mutants have been developed that allow synchronization of zygotic cells prior their entry into the meiotic program [12]. For example, a strain harboring a pat1-114 mutation produces a temperature-sensitive Pat1 114 kinase. Under low ambient temperature, Pat1 114 is active and inhibits cells from entering meiosis. In pat1-114 cells undergoing a transition from low to elevated temperature, Pat1 114 is readily inactivated and that triggers a cell cycle switch from mitosis to meiosis in a highly efficient and synchronous manner [13,14].
Investigation of copper homeostasis proteins during meiosis in S. pombe has revealed that the copper transporters Ctr4 and Ctr5 co-localize at the plasma membrane of zygotic cells shortly after induction of meiosis under low copper conditions [15]. In the case of Ctr6, which is the third member of the copper transporter (Ctr) family in fission yeast, the protein localizes to vacuolar membranes in early meiosis and then undergoes redistribution in a time-dependent manner to reach forespore membranes where it persists until sporulation [15]. In cells proliferating in mitosis as well as cells entering and progressing throughout the meiotic cell cycle, the copper starvation-dependent transcriptional induction of ctr4 + , ctr5 + , and ctr6 + genes is under the control of the copper-dependent transcription factor Cuf1 [6,15]. In the case of ctr6 + , its meiosis-specific expression also requires the participation of the meiotic regulator Mei4 [15]. DNA microarrays using probes derived from RNA isolated from copperstarved versus copper-replete meiotic cells allowed the identification of mfc1 + [6]. The mfc1 + transcript is the most highly expressed of all of the mRNAs detected in meiotic cells under copper-limiting conditions [6]. Under the control of its own promoter, mfc1 + expression was exclusively detected during meiosis in copper-starved cells. Mfc1 belongs to the major facilitator superfamily (MFS) of transporters and lacks overall sequence homology with Ctr-like proteins. At the protein level, Mfc1 is produced and observed throughout the meiotic divisions and spore maturation process. Mfc1 localizes in the surroundings of the forespore membrane and its presence is required for copper accumulation into forespores and production of fully active copper amine oxidases [6]. Investigation of the full-scale meiotic copper-dependent transcriptional program has revealed that mfc1 + exhibits a distinct temporal expression profile when compared to that of ctr4 + and ctr5 + . Whereas ctr4 + and ctr5 + are primarily expressed within the first 3 h of meiosis, followed by their repression after metaphase I, mfc1 + is transcriptionally induced after the 3-h meiotic time point, with its transcript levels remaining sustained throughout the meiotic program [6,15]. A second important distinction is that, the deletion of cuf1Δ impairs the transcriptional induction of ctr4 + and ctr5 + but mfc1 + remains unaffected, suggesting the existence of a distinct transcriptional regulator for induction of mfc1 + in response to Cu starvation [6,15].
Functional analysis of regions in the promoter of mfc1 + has demonstrated that two TCGGCG cis-acting elements harboring CGG triplets are necessary for transcriptional activation of mfc1 + under copper-limiting conditions [16]. Consistent with the fact that several regulatory elements containing CGG triplets are known to serve as binding sites for members of the Zn 2 Cys 6 cluster transcriptional regulator family, Mca1 (a Zn 2 Cys 6 transcription factor) is required for maximum copper-starvation-induced expression of mfc1 + [16]. Strikingly, developing zygotic mca1Δ/mca1Δ cells incubated in the presence of the copper chelator TTM undergo a meiotic block at metaphase I [16]. With respect to primary features of Mca1, its Nterminal region contains one zinc-finger unit that includes six Cys residues in the pattern CX 2 CX 6 CX 6 CX 2 CX 6 C (residues  which are known to coordinate two zinc ions [17][18][19]. The zinc finger unit of Mca1 is followed by a linker region (residues 52-116). Although sequence alignments show an absence of similarities between linkers in various Zn 2 Cys 6 cluster transcription factors, the linker region generally determines DNA half-site spacing and orientation preference of cis elements containing CGG triplets [17,20]. In the case of Mca1, the fact that its linker region is unusually large may justify its capability to bind the two CCG triplets that are found in the direct palindromic TCGGCGN 13 TCGGCG sequence within the mfc1 + promoter. Within this palindrome (TCGGCGN 13 TCGGCG), the distance between the two CGG triplets is 16 bp, which is an atypical long stretch of nucleotides compared to canonical TCGGCG elements containing CGG trinucleotides. Analysis of the amino acid sequence of Mca1 has identified a region that is composed of three heptad repeats of Leu residues (positions 117 to 138). This region is predicted to form an α-helix and may constitute a dimerforming region of Mca1 that is involved in an intermolecular Mca1-Mca1 interaction. Alternatively, this region may be involved in the formation of a heterodimeric complex with Mca1 and another member of the Zn 2 Cys 6 cluster protein family or, a partner from another transcription factor family [17,20]. Based upon this observation and as reported in the case of several Zn 2 Cys 6 cluster transcription factors, the zinc finger unit, the linker and the dimerization region would form the DNA-binding domain of Mca1. In vitro binding studies have consistently shown that the N-terminal 150 amino acids of Mca1 specifically associates with the palindromic TCGGCGN 13 TCGGCG sequence of the mfc1 + promoter region [16]. A large region that includes three sub-regions which are termed ID1, ID2, and ID3 (inhibitory domains) is located C-terminal to the DNA binding domain of Zn 2 Cys 6 cluster transcription factors [21,22]. The ID1/2/3 sub-regions are also called the middle homology region (MHR) and are thought to participate in regulating the transcriptional activity of zinc binuclear cluster proteins [17,[20][21][22]. Deletions and mutations within the MHR frequently render Zn 2 Cys 6 transcriptional regulators constitutively active, suggesting a role for MHR in preventing transactivation in the absence of specific stimuli [17,20]. With respect to transcriptional activation function of zinc binuclear cluster transcription factors, several of them contain an acidic amino acid region at their C termini that acts as transactivation domain.
In this report, we sought to gain insight into the mechanism by which Mca1 operates to activate mfc1 + gene expression. Using a chromatin immunoprecipitation (ChIP) approach, we showed that Mca1 is constitutively bound to mfc1 + promoter. Under conditions of copper starvation, results of ChIP experiments showed that Mca1 operates through a transcriptional mechanism to promote RNA polymerase II (Pol II) occupancy along the mfc1 + transcribed region. In marked contrast, association of Pol II with mfc1 + locus drastically decreases under copper-replete conditions. Further ChIP analysis showed that, although Mca1 occupies the mfc1 + promoter in vegetative cells, Pol II occupancy along the mfc1 + transcribed region is exclusively detected during meiosis. Deletion analysis of the Mca1 protein revealed that a central portion (residues 299 to 600) fused to the DNA-binding domain (residues 1 to 150) of Mca1 was required to enhance Pol II occupancy at the mfc1 + locus in meiotic cells under low copper conditions. Taken together, these findings reveal novel features of Mca1, a key regulator required for promoting Pol II-dependent mfc1 + transcription.

Plasmids
The mca1 + promoter containing 500 bp of the 5' noncoding region and the first 150 codons (450 bp) of the mca1 + gene was isolated by PCR, using primers designed to generate ApaI and XmaI restriction sites at the termini of the PCR product. The PCR product obtained was digested with ApaI and XmaI and cloned into the corresponding sites of pBPade6 [25]. The resulting plasmid was designated pBPade6prom-mca1DBD. To create a plasmid containing full-length Mca1, an XmaI-SacII PCR-amplified fragment containing the last 547 codons of the mca1 + ORF was isolated. The XmaI-SacII DNA fragment was inserted in-frame into the corresponding sites of pBPade6prom-mca1DBD, generating a plasmid that was denoted pBPa-de6prom-mca1 + . To construct a TAP epitope-tagged Mca1, the TAP coding sequence derived from Cuf2-TAP [26] was isolated by PCR using primers designed to generate SacII and SacI sites at the 5' and 3' termini of the TAP ORF. The resulting DNA fragment was used to clone the TAP coding sequence into pBPade6prom-mca1 + plasmid to which SacII and SacI restriction sites had previously been introduced by PCR and placed immediately before the mca1 + stop codon. For this particular construct, named pBPade6prom-mca1 + -TAP, the SacII-SacI TAP-encoded fragment was inserted in-frame with the C-terminal region of Mca1.
To create the pSKcuf1 + plasmid, an ApaI-XmaI PCR-amplified DNA segment containing the cuf1 + locus starting at -977 bp from the translational start codon up to the stop codon was inserted into the corresponding sites of pBluescript SK (Stratagene). Subsequently, the myc 12 epitope coding region was isolated from pctr4 + -X-myc 12 plasmid [27] using XmaI and XbaI and then inserted into the corresponding sites of pSKcuf1 + plasmid, resulting in an in-frame fusion of the cuf1 + gene with twelve copies of the Myc epitope. Once created, the cuf1 + -myc 12 fusion gene and its promoter region were isolated using ApaI and SacI and inserted into the corresponding sites of pBPade6. The resulting integrative plasmid was designated pBPade6-cuf1 + -myc 12 . In the case of the untagged cuf1 + allele, the integrative plasmid pBPade6cuf1 + was constructed by subcloning the ApaI-XmaI fragment from pSKcuf1 + into the corresponding sites of pBPade6.

RNA isolation and analysis
Total RNA was extracted using a hot phenol method as described previously [28]. Gene expression profiles were analyzed using RNase protection assays as described previously [29]. Plasmids pSKmfc1 + [6], pSKmca1 + [16], pSKctr4 + [30] and pSKact1 + [31] were used to produce antisense RNA probes that served to determine mfc1 + , mca1 + , ctr4 + , and act1 + steadystate mRNA levels, respectively. 32 P-labeled antisense RNA probes were produced using the above-mentioned BamHI-linearized plasmids and with the use of [α-32 P]UTP and T7 RNA polymerase. The act1 + riboprobe was used to detect act1 + transcript as an internal control for normalization during quantification of the RNase protection products.

Protein extraction and analysis
To determine the steady-state protein levels of Mca1-TAP or its mutant derivatives and α-tubulin, whole cell extracts were prepared using a trichloroacetic acid (TCA) extraction method [32]. Equal amounts of each sample preparation were resuspended in sodium dodecyl sulfate loading buffer and proteins were resolved by electrophoresis on 8% sodium dodecyl sulfatepolyacrylamide gels. Proteins were then electroblotted onto nitrocellulose membranes for 1 h at 4˚C. Membranes were blocked by treatment with 5% powdered skim milk (Difco) in TBS (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1% bovine serum albumin) containing 0.1% Tween 20 (TBST). Following washings with TBST, membranes were incubated with primary antibodies dissolved in 1% powdered skim milk in TBST for 16 h at 4˚C. The following antibodies were used for immunodetection of Mca1-TAP (or its mutant derivatives) and α-tubulin: polyclonal anti-mouse IgG antibody (ICN Biomedicals) and monoclonal anti-α-tubulin antibody (clone B-5-1-2; Sigma-Aldrich), respectively. Following incubation, the membranes were washed and incubated with the appropriate horseradish peroxidase-conjugated secondary antibodies (Amersham Biosciences), developed with enhanced chemiluminescence (ECL) reagents (Amersham Biosciences), and visualized by chemiluminescence using an ImageQuant LAS 4000 instrument (GE Healthcare) equipped with a Fujifilm High Sensitivity F0.85 43 mm camera.

Mca1 binds to the mfc1 + promoter in vivo in a copper-independent manner
We have previously reported that the transcription factor Mca1 was required for maximal induction of mfc1 + mRNA levels in response to low concentrations of copper (Fig 1). Despite its transcriptional regulatory role, mfc1 + promoter occupancy in vivo by Mca1 remained uncharacterized. To understand the extent by which Mca1 occupies the promoter of mfc1 + in vivo under copper starvation compared to its occupancy in the presence of copper, we used a pat1-114/pat1-114 mca1Δ/mca1Δ mutant strain that had been transformed with a control integrative plasmid or a plasmid harboring an untagged or a TAP-tagged version of Mca1 under the control of its own promoter. pat1-114/pat1-114 mca1 + /mca1 + (control) and pat1-114/pat1-114 mca1Δ/mca1Δ diploid cells expressing the indicated integrative plasmid (empty, mca1 + or mca1 + -TAP) were induced to undergo synchronous meiosis in the presence of the copper chelator TTM (150 μM) or CuSO 4 (25 μM). Subsequently, total RNA was isolated from culture aliquots taken at different time points. In the case of mca1Δ/mca1Δ cells expressing Mca1 or Mca1-TAP, mfc1 + transcripts were initially detected after 3 h of meiotic induction under low copper. These observations contrasted with the absence of significant induction of Mca1 is required for maximal copper-starvation-induced expression of mfc1 + . A, pat1-114/pat1-114 mca1Δ/mca1Δ cells containing an empty integrative vector (vector alone) and re-integrated untagged mca1 + /mca1 + and mca1 + -TAP/mca1 + -TAP alleles were synchronously induced to undergo meiosis under copper-starved (TTM, 150 μM) and copper-replete (Cu, 25 μM) conditions. Total RNA was isolated from culture aliquots taken at the indicated time points after meiotic induction. Following RNA isolation, mfc1 + and act1 + steady-state mRNA levels were analyzed by RNase protection assays. pat1-114/pat1-114 mca1 + /mca1 + is the isogenic parental strain that was used as a control. B, Graphic representation of the quantification of the results of three independent RNase protection assays, including assays shown in panel A. Values are represented as averages ± S.D. The asterisks correspond to p<0.0001 ( ÃÃÃÃ ) (paired Student's t-test). transcription of mfc1 + mRNA in cells harboring an empty plasmid at the 3-h time point under similar conditions (Fig 1). Results showed that, under copper-starved conditions, mfc1 + mRNA levels reached high levels after 7 h of meiotic induction in mca1Δ/mca1Δ cells expressing an untagged or a TAP-tagged version of Mca1 (7.2-fold) in comparison with cells transformed with the vector alone (Fig 1). The expression profile of mfc1 + mRNA in mca1Δ/mca1Δ cells expressing Mca1 or Mca1-TAP was similar to that observed in the case of mca1 + /mca1 + (control) cells, except that the magnitude of induction of mfc1 + mRNA was modestly higher (9%, 3%, and 23% at 3-, 5-and 7-h time points, respectively) (Fig 1). Under copper-replete conditions, transcript levels of mfc1 + were expressed to near background levels in mca1 + / mca1 + or mca1Δ/mca1Δ cells whether the cells contained an empty vector or an integrative vector expressing Mca1 or Mca1-TAP (Fig 1).
Results showed that the in-frame TAP insertion did not interfere with Mca1 function. For example, when Mca1-TAP was expressed in mca1Δ/mca1Δ cells, it conferred copper starvation-dependent regulation of mfc1 + expression in a fashion similar to that of the untagged Mca1 protein. This observation was taken as evidence that Mca1-TAP retained wild-type function. Using mca1Δ/mca1Δ cells expressing Mca1-TAP, we first determined the steady-state protein levels of Mca1-TAP over a time period of 1 to 7 h after meiotic induction. Results showed that Mca1-TAP levels were constitutively present irrespective of the cellular copper status (Fig 2). We took advantage of the constitutive expression of Mca1-TAP to test whether Mca1-TAP could be detected at the mfc1 + and cuf2 + promoters in vivo using a ChIP approach. pat1-114/pat1-114 mca1Δ/mca1Δ diploid cells expressing either an untagged or a TAP-tagged Mca1 were synchronized to initiate and proceed through meiosis under copper-deficient or copper-replete conditions. After 1, 3, 5, and 7 h of meiotic induction, results showed that Mca1-TAP constitutively occupied the mfc1 + promoter (Fig 2). Although results showed that Mca1-TAP chromatin occupancy of mfc1 + promoter was slightly higher in copper-starved cells compared to copper-replete cells, Mca1-TAP clearly occupied the mfc1 + promoter under both low and high copper conditions (Fig 2). At the 1-h time point, Mca1-TAP level occupancy was lower in comparison to level observed after 3, 5, and 7 h of meiotic induction. This reduced association of Mca1-TAP with the mfc1 + promoter may contribute to the markedly reduced transcription of mfc1 + in cells at this meiotic time point under low and high copper concentrations (Fig 1). Promoter occupancy by Mca1-TAP was detected using primers amplifying a DNA region located between positions -183 and -32 relative to the initiator codon of mfc1 + . This amplified promoter region is known to contain two functional TCGGCG elements that are bound by Mca1 in vitro [16]. Results from ChIP analysis showed that Mca1-TAP did not associate with the cuf2 + promoter (used as a negative control). Only very weak levels of cuf2 + promoter fragments were immunoprecipitated (Fig 2). These low levels of immunoprecipitated chromatin were similar to the background signals observed when ChIP assays were performed in mca1Δ/mca1Δ cells expressing untagged mca1 + , which had been re-integrated (Fig 2). Taken together, these results showed that Mca1 is recruited to the promoter of mfc1 + in a constitutive manner, with a slight increased ability to bind mfc1 + promoter under conditions of low levels of copper.

The RNA polymerase II is recruited to mfc1 + locus in vivo in response to copper starvation
The RNA polymerase II (Pol II) is a fundamental multi-subunit enzyme that transcribes numerous genes into mRNA in eukaryotic cells [36]. General, signal-and stress-responsive transcription factors control either positively or negatively gene expression through multiple mechanisms that affect Pol II occupancy at specific coding regions of target genes [37]. Taking into account the result that Mca1 constitutively occupied the mfc1 + promoter, we used a Pol II ChIP assay to determine whether the presence of Mca1 was required for Pol II chromatin occupancy and whether Pol II was primarily recruited in response to low copper availability. Analysis of Pol II chromatin occupancy was performed using an anti-Rpb1 antibody that recognizes the heptapeptide sequence YSPTSPS, which is found in multiple copies at the C-terminal domain (CTD) of Pol II [38]. pat1-114/pat1-114 mca1Δ/mca1Δ cells harboring an empty integrative plasmid or expressing untagged mca1 + were synchronously induced into meiosis. Immediately before their passage to meiosis, cell cultures were treated with TTM (150 μM) or CuSO 4 (25 μM). Aliquots of cell cultures were taken 1, 5, and 7 h after meiotic induction and cells were fixed by formaldehyde treatment for ChIP assays. Primers that hybridized immediately upstream (positions -183 to -32) and at the 3' end of mfc1 + (positions +875 to +960) coding region were used for qPCR analysis (Fig 3A). Results of ChIP analysis showed that in cells expressing Mca1, association of Pol II with the mfc1 + locus was highly enriched in response to copper starvation conditions and was negligible or absent under copper-replete conditions ( Fig 3B). In copper-starved cells expressing Mca1, Pol II exhibited 14.0-, 40.1-, and 45.0-fold increases in its binding to mfc1 + (positions -183 to -32) relative to a control 18S ribosomal DNA coding region after 1, 5, and 7 h of meiotic induction (Fig 3B). Under the same conditions, levels of Pol II enrichment were higher (17.6-, 92.4-, and 100.0-fold increases) in its binding at the 3' end of mfc1 + (positions +875 to +960) coding region relative to a 18S ribosomal DNA region reference (Fig 3B). Taken together, these results revealed that association of Pol II with an mfc1 + transcribed region in vivo requires the presence of the transcription factor Mca1 in response to low concentrations of copper.

The mfc1 + promoter is constitutively bound by Mca1-TAP during mitosis but there is absence of expression of mfc1 + regardless of changes in copper concentrations
Previous studies have shown that the copper transporter ctr4 + mRNA levels are induced in a Cuf1-dependent manner in cells proliferating in mitosis under conditions of copper starvation [30,39,40]. After treatment with copper, Cuf1 is inactivated [39,41,42] and then ctr4 + mRNA is not detected because its expression is dependent of Cuf1 [40]. Here, we used ctr4 + as a control gene known to be induced under copper-limiting conditions ( Fig 4A). Furthermore, we performed ChIP assays using Cuf1-Myc 12 as a control transcription factor known to interact with the ctr4 + promoter in vivo under low copper conditions (Fig 4B and 4C) [43]. In the case of the mfc1 + transcription profile, there was no detection of the mfc1 + transcript in mitotically growing cells in the presence or absence of TTM or CuSO 4 (Fig 4A) [6]. However, the ability of Mca1 to associate with mfc1 + promoter in vivo in cells proliferating in mitosis remains unknown. To examine whether there was absence or presence of Mca1 in the promoter region of mfc1 + in mitotic cells, ChIP assays were performed in vegetatively growing mca1Δ cells expressing a functional mca1 + or mac1 + -TAP allele under copper-starved or copper-replete position relative to the A of the initiator codon of each gene (mfc1 + or cuf2 + ). Numbers in parentheses indicate promoter regions that were used for qPCR analysis in ChIP assays. Black boxes depict wild-type TCGGCG sequences known to serve as a binding site for Mca1. C-D, pat1-114/pat1-114 mca1Δ/mca1Δ cells expressing Mca1-TAP or untagged Mca1 were synchronously induced to undergo meiosis. Following induction of meiosis, chromatin was immunoprecipitated using Sepharose-bound anti-mouse IgG antibodies at the indicated time points. Specific regions of mfc1 + and cuf2 + promoters were analyzed by qPCR to determine Mca1-TAP occupancy in the presence of TTM or copper. Binding of Mca1-TAP to promoters was calculated as the enrichment of specific mfc1 + and cuf2 + promoter regions relative to a 18S ribosomal DNA coding region. The cuf2 + promoter served as a negative control. ChIP data were calculated as values of the largest amount of chromatin measured (fold enrichment). Results are shown as the averages ± S.D. of a minimum of three independent experiments. conditions. ChIP analysis showed that in the case of cells treated with TTM for 1 and 3 h, Mca1-TAP immunoprecipitated 3.8-and 4.3-fold more chromatin corresponding to mfc1 + promoter compared to an 18S ribosomal region reference (Fig 4B and 4C). In the case of cells exposed to copper for the same times of treatments, chromatin enrichment levels by Mca1--TAP were 3.2-and 3.6-fold, respectively ( Fig 4C). These results were consistent with the interpretation that in mca1Δ cells expressing Mca1-TAP, the mfc1 + promoter was occupied by Mca1-TAP in the presence of TTM in a manner similar to that observed for mca1Δ cells expressing Mca1-TAP in the presence of copper (Fig 4C; Figure A in S1 File). Results showed that untagged Mca1 (expressed in mca1Δ cells) immunoprecipitated only background levels of mfc1 + promoter region ( Fig 4C). As control experiments, results from ChIP analysis showed that in the case of cuf1Δ cells expressing Cuf1-Myc 12 that were treated with TTM for 1 and 3 h, Cuf1-Myc 12 occupied the ctr4 + promoter with 3.7-and 3.8-fold enrichments, respectively, relative to a control 18S ribosomal DNA region reference (Fig 4C). In these cells, steady-state levels of ctr4 + mRNA consistently increased 11-and 9-fold above the levels observed in copperreplete cells (Fig 4A). In contrast, when cells expressing Cuf1-Myc 12 were incubated in the presence of copper for 1 and 3 h, only very weak levels of ctr4 + promoter fragments were immunoprecipitated (Fig 4C). These low levels of immunoprecipitated chromatin were similar Copper starvation-dependent induction of Mfc1 to the background signals observed when ChIP assays were performed in a cuf1Δ strain expressing an untagged cuf1 + allele, which had been re-integrated ( Fig 4C). The association of Cuf1-Myc 12 with the ctr4 + promoter was detected using primers amplifying a DNA region located between positions -737 and -584 relative to the initiator codon of ctr4 + (Fig 4B). This amplified promoter region is known to contain functional CuSE elements that are specifically recognized by Cuf1 [30].
Despite a constitutive mfc1 + promoter occupancy by Mca1-TAP, mitotic cells grown under low and high concentrations of copper revealed that the target gene mfc1 + was silent (Fig 4A). To gain insight into this observation, we analyzed Pol II chromatin occupancy using a pair of primers located between positions +875 and +960 relative to the initiator codon of mfc1 + . Results showed no enrichment of immunoprecipitated chromatin corresponding to the mfc1 + coding region relative to an 18S ribosomal region reference. The absence of association of Pol II with the mfc1 + locus was observed under low and high concentrations of copper (Fig 4D  and 4E). In the case of ctr4 + , ChIP experiments showed that Pol II occupied the ctr4 + coding region upstream of the initiator codon of ctr4 + (positions +718 and +817) under copper-limiting conditions. In contrast, in cells undergoing a transition from low to high copper, Pol II was absent from the ctr4 + locus (Fig 4D and 4E). Taken together, these results suggested that cells proliferating in mitosis lacked a factor (perhaps meiosis-specific) or possessed a mechanism that prevented Pol II-dependent transcriptional activation of mfc1 + despite the fact that Mca1 constitutively occupied the mfc1 + promoter in vivo.

In vivo mapping of a middle region of Mca1 that is required for induction of mfc1 + expression
In cells undergoing synchronous meiosis, Mca1-TAP was constitutively bound to the mfc1 + promoter but recruitment of Pol II and, therefore, mfc1 + expression occurred only under conditions of copper starvation. We thus sought to investigate whether a domain of Mca1 was required for copper starvation-dependent induction of mfc1 + . Previous studies had shown that the N-terminal 150-residue segment of Mca1 constituted its DNA-binding domain (DBD) [16]. Because this domain binds DNA constitutively, we kept it intact and generated two internal in-frame Mca1 deletions, starting from Ile 151 and extending to residues Leu 299 and Gly 326 , respectively. The rationale for investigating these 2 mutant forms of Mca1 was the fact that the first deleted region (residues 151-298) contains 2 potential metal-binding motifs (Cys 163 -X-Cys-X-His 165 and Met 234 -X 3 -Met 238 ). The resulting C-terminal portions of Mca1 (residues 299-697 and 326-697) were fused in-frame to the DNA binding domain of Mca1 (residues 1-150) (Fig 5A). Because a Mca1-TAP protein triggers activation of mfc1 + gene expression in a  1 and 3 h. In the cases of cuf1Δ cells expressing an integrated untagged or a Myc-tagged cuf1 + allele and mca1Δ cells expressing an integrated untagged or a TAP-tagged mca1 + allele, aliquots of these cultures were fixed by formaldehyde treatment for ChIP assays and at the same time collected to isolate RNA for assessment of transcript levels. Shown are representative RNase protection assays of ctr4 + , mfc1 + , and act1 + mRNA steady-state levels at the indicated time points. B, Schematic representation of ctr4 + and mfc1 + promoter regions. Nucleotide numbers refer to the position relative to the A of the initiator codon of each gene. Numbers in parentheses indicate promoter regions that were used for qPCR analysis in ChIP assays. White and black boxes depict wild-type CuSE and TCGGCG sequences known to serve as binding sites for Cuf1 and Mca1, respectively. C, Specific regions of ctr4 + (left) and mfc1 + (right) promoters were analyzed by qPCR to determine chromatin occupancy by Cuf1-Myc 12 and Mca1-TAP, respectively. Association of Cuf1-Myc 12 and Mca1-TAP to promoters was calculated as the enrichment of specific ctr4 + and mfc1 + promoter regions relative to a 18S ribosomal DNA coding region. ChIP data were calculated as values of the largest amount of chromatin measured (fold enrichment). Results are shown as the averages ± S.D. of a minimum of three independent experiments. D, Schematic representation of ctr4 + and mfc1 + coding regions. Numbers in parentheses indicate coding regions that were used for qPCR analysis in ChIP assays. E, ChIP analysis was performed to determine RNA Pol II chromatin occupancy at ctr4 + and mfc1 + genes. Chromatin was immunoprecipitated using RNA Pol II CTD-specific antibodies. Data were calculated as values of the largest amount of chromatin measured (fold enrichment). Results are shown as the averages ± S.D. of a minimum of three independent experiments. The asterisks correspond to p<0.001 ( ÃÃ ) and p<0.0001 ( ÃÃÃÃ ) (paired Student's t-test). https://doi.org/10.1371/journal.pone.0201861.g004 Copper starvation-dependent induction of Mfc1 manner similar to that of wild-type untagged protein, each mutant derivative was tagged with TAP. The mca1 + -TAP allele and its mutant derivatives were expressed in pat1-114/pat1-114 mca1Δ/mca1Δ cells that underwent synchronous meiosis under copper-starved or copperreplete conditions. Maximal activation of mfc1 + expression was observed in copper-starved cells expressing mca1 + -TAP and mca1DBD-299-697-TAP alleles after 7 h of meiotic induction (Fig 5B). When the middle region of Mca1 was further deleted to residue 326 to generate the mca1DBD-326-697-TAP allele, TTM-dependent expression of mfc1 + was strikingly decreased by 85% to a minimal level of activation that was similar to that observed in the case of cells harboring an empty vector (TTM, 7-h time point) (Fig 5B). In the case of all three alleles (mca1 + -TAP, mca1DBD-299-697-TAP, and mca1DBD-326-697-TAP) expressed in mca1Δ/mca1Δ cells, elevated copper concentrations resulted in a strong repression of mfc1 + mRNA expression (Fig 5B).
To further verify that Mca1-TAP and its two mutant derivatives were associated with the  (Fig 5C and 5E). The association of Mca1-TAP, Mca1DBD-299-697-TAP, and Mca1DBD-326-697-TAP with the mfc1 + promoter was detected using primers amplifying a DNA region located between positions -183 and -32 relative to the initiator codon of mfc1 + (Fig 5F). To assess that Mca1-TAP and its mutant derivatives were produced, total protein extracts were analyzed by immunoblotting at the indicated meiotic time points. Results showed that detectable levels of Mca1--TAP, Mca1DBD-299-697-TAP, and Mca1DBD-326-697-TAP proteins were present in mca1Δ/mca1Δ cells (Fig 5G and 5H).
We next sought to examine whether Mca1, Mca1DBD-299-697, and Mca1DBD-326-697 enhanced Pol II occupancy at the mfc1 + locus. To carry out these experiments, we used untagged Mca1 proteins to avoid any interference of the TAP tag (which can bind to the Fc region of IgG) with the monoclonal anti-RNA Pol II antibody (isotype IgG) that was used for immunoprecipitation of Pol II bound to chromatin. mca1Δ/mca1Δ cells expressing untagged  (Figs 5 and 6). mfc1 + mRNA levels were induced within 5 h, exhibiting high levels of expression after 7 h of meiotic induction (Fig 6A). In the case of cells expressing an untagged Mca1DBD-326-697 protein, TTM-dependent transcription of mfc1 + was strongly reduced by 95% to a minimal level of expression that was similar to that observed in the case of cells containing an empty vector (TTM, 7-h time point) (Fig 6A). In the case of all three untagged alleles (mca1 + , mca1DBD-299-697, and mca1DBD-326-697) expressed in mca1Δ/mca1Δ cells, mRNA levels of mfc1 + remained very low under copper-replete conditions (Fig 6A). Given the fact that Mca1 and its mutant derivatives were untagged, we assessed the mRNA levels of the mca1 + , mca1DBD-299-697, and mca1DBD-326-697 alleles using aliquots of cultures that served to analyze the expression profile of mfc1 + . Results showed that mca1 + , mca1DBD-299-697, and mca1DBD-326-697 genes were constitutively expressed as their corresponding transcripts were detected throughout the meiotic program under copper-starved and copper-replete conditions (Fig 6B).
In copper-starved mca1Δ/mca1Δ cells expressing Mca1 and Mca1DBD-299-697, association of Pol II with the mfc1 + locus was enriched 43.0-and 38.2-fold, respectively, relative to a 18S ribosomal region reference (Fig 6C-6E). However, these levels of Pol II enrichment due to the presence of Mca1 and Mca1DBD-299-697 were 33.2-and 29.5-fold when we took into account that Pol II chromatin occupancy was slightly enriched in TTM-treated cells lacking Mca1 (containing an empty vector) after 7 h of meiotic induction (9.8-and 8.7-fold) (Fig 6D and  6E). In the case of copper-starved mca1Δ/mca1Δ cells expressing Mca1DBD-326-697, results showed that Pol II enrichment levels (9.3-fold) were similar to those in mca1Δ/mca1Δ cells containing an empty vector (8.9-fold) (Fig 6F). We therefore concluded that Mca1DBD-326-697 interacted with the mfc1 + promoter in vivo (Fig 5E). However, this mutant was unable to foster a significant recruitment of Pol II at the mfc1 + locus, resulting in low levels of mfc1 + mRNA expression even under low concentrations of copper (Fig 6A and 6F). Under copperreplete conditions, results showed that Pol II immunoprecipitated very low or background levels of mfc1 + DNA coding region when mca1Δ/mca1Δ cells expressed mca1 + , mca1DBD-299-697 or mca1DBD-326-697 alleles (Fig 6D-6F).

In vivo mapping of a C-terminal region of Mca1 that is required for induction of mfc1 + expression
To further delineate a region of Mca1 that was required for transcriptional activation of mfc1 + expression under low copper conditions, truncations were created from the C-terminal end of a functional Mca1-TAP protein (Fig 7A). We started with two small deletions due to the fact that the C-terminal end of Zn 2 Cys 6 cluster transcription factors are known to be required for transcriptional activity of this group of regulators. The wild type mca1 + -TAP and mutant mca1 1-600-TAP and mca1 1-570-TAP alleles were returned to pat1-114/pat1-114 mca1Δ/mca1Δ cells. Subsequently, these cells were synchronously induced into meiosis and treated with TTM (150 μM) or CuSO 4 (25 μM). At the indicated times after meiotic induction, total RNA was isolated from aliquots of cultures and assayed for copper-dependent regulation of the mfc1 + copper-responsive transcript to determine whether the C-terminal end was important for Mca1 function in vivo. In the case of mca1Δ/mca1Δ cells expressing Mca1 1-600-TAP under low copper conditions, mfc1 + mRNA levels were first detected at 3-and 5-h time points.  (Fig 7B). As a control, a similar expression profile of mRNA levels of mfc1 + was observed in mca1Δ/mca1Δ cells expressing Mca1-TAP (Fig 7B). In contrast, mca1Δ/mca1Δ cells expressing the Mca1 1-570-TAP mutant were unable to mediate a robust activation of mfc1 + expression in response to copper starvation. The magnitude of the copper starvationdependent induction was decreased by 86% and 92% at 5-and 7-h time points, respectively, compared to cells expressing wild-type Mca1-TAP (Fig 7B). In fact, copper-starved mca1Δ/ mca1Δ cells expressing Mca1 1-570-TAP exhibited a minimal level of mfc1 + expression that was similar to that observed in the case of cells containing an empty vector (Fig 7B).
To test whether Mca1 1-600-TAP and Mca1 1-570-TAP occupied the mfc1 + promoter region, pat1-114/pat1-114 mca1Δ/mca1Δ cells expressing wild-type Mca1-TAP and these mutant proteins were synchronized to initiate and proceed through the meiotic program. Immediately prior to their entry into meiosis, the cells were either exposed to TTM (150 μM) or CuSO 4 (25 μM). Aliquots of cultures were retrieved after 1 and 7 h of meiotic induction and ChIP assays were performed to assess the levels of  (Fig 7C and 7D). In the case of Mca1 1-570-TAP, its enrichment levels were lower than enrichment levels of Mca1-TAP and Mca1 1-600-TAP ( Fig  7E). Notably, Mca1 1-570-TAP constitutively associated with the mfc1 + promoter, exhibiting enrichment levels of 1.9-and 2.2-fold after 1 and 7 h in the presence of TTM, and 1.9-and 2.0-fold after 1 and 7 h in the presence of CuSO 4 (Fig 7C-7E). The association of Mca1-TAP, Mca1 1-600-TAP, and Mca1 1-570-TAP with the mfc1 + promoter was detected using primers amplifying a DNA region located between positions -183 and -32 relative to the initiator codon of mfc1 + (Fig 7F). To verify that Mca1-TAP, Mca1 1-600-TAP, and Mca1 1-570-TAP proteins were expressed in pat1-114/pat1-114 mca1Δ/mca1Δ cells during the meiotic program, total cell extracts were analyzed by immunoblotting using anti-IgG and anti-α-tubulin antibodies. Results showed that these proteins were present after 1 and 7 h of meiotic induction in the presence of TTM or CuSO 4 (Fig 7G and 7H).
As mentioned previously untagged version of Mca1, Mca1 1-600, and Mca1 1-570 were created to avoid any interference of the TAP tag with the anti-RNA Pol II antibody (IgG-type) that was used to selectively pull down chromatin bound by Pol II. We examined the individual contribution of each version of the transcription factor (Mca1, Mca1 1-600, and Mca1 1-570) in the regulation of mfc1 + in the presence of TTM (150 μM) or CuSO 4 (25 μM). Steady-state transcript levels of mfc1 + were analyzed 1, 3, 5, and 7 h after meiotic induction. In the presence of TTM, cells expressing mca1 + and mca1 1-600 alleles exhibited maximal expression of mfc1 + transcripts that was induced 23-and 25-fold, respectively, compared to their levels of expression in copper-treated cells (7-h time point) (Fig 8A). In the case of cells expressing the mca1 1-570 allele, the steady-state levels of mfc1 + mRNA were dramatically decreased by 95% under copper-starved (TTM) conditions compared to transcript levels of wild-type Mca1 under the same experimental conditions (7-h time point). Furthermore, the weak levels of mfc1 + at the mfc1 + gene. Chromatin was immunoprecipitated using RNA Pol II CTD-specific antibodies. ChIP results are presented as enrichments of mfc1 + (+875/+960) coding region relative to a 18S ribosomal DNA coding region. Data were calculated as values of the largest amount of chromatin measured (fold enrichment). Results are shown as the averages ± S.D. of a minimum of three independent experiments. The asterisks correspond to p<0.001 ( ÃÃ ) and p<0.0001 ( ÃÃÃÃ ) (paired Student's t-test). https://doi.org/10.1371/journal.pone.0201861.g006 Copper starvation-dependent induction of Mfc1 transcript in this mutant (Mca1 1-570) were similar to those observed in the case of cells transformed with plasmid alone (Fig 8A).
We assayed the levels of mRNA expression of mca1 + , mca1 1-600, and mca1 1-570 alleles by RNase protection assays to obtain further indication that untagged Mca1, Mca1 1-600, and Mca1 1-570 proteins were expressed in mca1Δ/mca1Δ cells. Results showed that wild-type and mutant mca1 constructs were clearly expressed, with transcripts constitutively detected for each indicated allele regardless of cellular copper status (Fig 8B).
To confirm that Mca1 and Mca1 1-600 affected transcription per se, we analyzed Pol II chromatin occupancy at the mfc1 + locus (positions +875 to +960) in mca1Δ/mca1Δ cells expressing mca1 + and mca1 1-600 alleles. Consistent with the levels of mfc1 + mRNA that were high under low copper conditions at the 7-h time point, levels of Pol II enrichment were higher in cells that had been exposed to TTM after 7 h. Data showed 38.6-(for mca1 + ) and 40.0-fold (for mca1 1-600) enrichment at the mfc1 + locus relative to a 18S ribosomal DNA sequence reference (Fig 8C-8E). In the case of cells expressing the Mca1 1-570 mutant, Pol II enrichment levels were strikingly lower (by 4.5-fold) compared to those observed in mca1 + and mca1 1-600 cells under the same low copper conditions after 7 h (Fig 8D-8F). In fact, this decrease in the association of Pol II with the mfc1 + locus paralleled a weak expression of mfc1 + (Fig 8A). When meiotic cells were incubated in the presence of copper, there was a lack of Pol II chromatin occupancy at the mfc1 + locus regardless of the expressed allele (mca1 + , mca1 1-600, and mca1 1-570) and meiotic time points (Fig 8D-8F). Taken together, these results indicated that a region of Mca1 encompassing amino acid residues 570 to 600 is required for recruitment of Pol II to the locus of mfc1 + . In contrast, the last 97 residues of Mca1 (600-697) are dispensable for Pol II recruitment to chromatin.

Fusion of amino acid fragments 1-150 and 299-600 of Mca1 are sufficient to trigger proper and timely control of mfc1 + expression
Given the results of mfc1 + expression under the control of different Mca1 mutants, we hypothesized that residues 299 to 600 of Mca1 were required for transcriptional activation of mfc1 + expression. Accordingly, disruption of this region would result in poor expression of mfc1 + under copper starvation conditions. To test this hypothesis, we deleted the amino acid region 299 to 600 of Mca1-TAP (Fig 9A). The resulting allele was named mca1 1-298+601-697-TAP. As a complementary approach, a chimeric protein that consisted in a fusion between amino acid fragments 1 to 150 (DBD) and 299 to 600-TAP of Mca1 (denoted mca1DBD-299-600-TAP) was created to test whether it mediated proper and timely control of mfc1 + gene expression as a function of copper availability. The experiments were set by expressing mca1 + -TAP, mca1 1-298+601-697-TAP, and mca1DBD-299-600-TAP alleles into pat1-114/pat1-114  mca1 1-600/mca1 1-600, and mca1 1-570/mca1 1-570 alleles or an empty vector (vector alone) were synchronously induced to undergo meiosis under copper-depleted (TTM, 150 μM) and copper-replete (Cu, 25 μM) conditions. Total RNA was isolated from culture aliquots taken at the indicated time points. Steady-state levels of mfc1 + and act1 + mRNAs were analyzed by RNase protection assays. Results shown are representative of three independent experiments. C-E, ChIP assays were performed from aliquots of cultures used in panel B after 1 and 7 h of meiotic induction. Binding of Mca1-TAP, Mca1 1-600-TAP or Mca1 1-570-TAP to the mfc1 + promoter was calculated as the enrichment of the specific mfc1 + proximal regulatory region (positions -183 to -32) relative to a 18S ribosomal DNA coding region. Data were calculated as values of the largest amount of chromatin measured (fold enrichment). Results are shown as the averages ± S.D. of a minimum of three independent experiments using separate chromatin preparations. F, Schematic representation of the mfc1 + gene in which ChIP1 region indicates the location of the primers that were used for qPCR analysis. Black boxes depict wild-type TCGGCG sequences. G-H, Whole-cell extracts were prepared from aliquots of cultures used in panel B after 1 and 7 h of meiotic induction. Samples were analyzed by immunoblot assays using anti-IgG and anti-αtubulin antibodies to detect TAP-tagged proteins and α-tubulin steady-state protein levels, respectively. https://doi.org/10.1371/journal.pone.0201861.g007 Copper starvation-dependent induction of Mfc1  mca1 1-600/mca1 1-600 or mca1 1-570/mca1 1-570 alleles were induced to initiate and proceed through meiosis in the presence of TTM (150 μM) or copper (Cu, 25 μM). At the indicated times after meiotic induction, total RNA was extracted from culture aliquots and steady-state mRNA levels of mfc1 + (panel A) and mca1 + (panel B) were analyzed by RNase protection assays using actin (act1 + ) as an internal control. The results shown are representative of three independent experiments. C, Schematic representation of the mfc1 + gene in which the ChIP2 region indicates positions of the primers that were used for qPCR analysis. D-F, After 1 and 7 h of meiotic induction, aliquots of the cultures described for panels A and B were taken and chromatin was prepared and immunoprecipitated using antibodies against RNA Pol II CTD. ChIP results are presented as enrichments of mfc1 + (+875/+960) coding mca1Δ/mca1Δ cells that were synchronously induced into meiosis in the presence of TTM (150 μM) or CuSO 4 (25 μM). Aliquots of cell cultures were taken 1, 3, 5, and 7 h after meiotic induction, and steady-state levels of mfc1 + mRNA were analyzed by RNase protections assays. Under low copper conditions (TTM), maximal induction of mfc1 + was primarily observed in cells expressing mca1 + -TAP and mca1DBD-299-600-TAP alleles after 7 h of meiotic induction (Fig 9B). In contrast, mRNA levels of mfc1 + remained low (96% reduction) in TTM-treated cells expressing mca1 1-298+601-697-TAP or an empty vector (Fig 9B). In the case of all the three indicated alleles expressed in mca1Δ/mca1Δ cells, there was a lack of significant induction of mfc1 + mRNA under copper-replete conditions (Fig 9B).
Despite We also noticed that the association of Mca1 1-298+601-697-TAP with the mfc1 + promoter was weaker than Mca1-TAP or Mca1DBD-299-600-TAP under both copper-limiting and copper-replete conditions. Immunoprecipitated chromatin by Mca1-TAP and its mutant derivatives was detected using primers amplifying a DNA region located between positions -183 and -32 relative to the initiator codon of mfc1 + (Fig 9F). To ensure that Mca1-TAP, Mca1 1-298 +601-697-TAP, and Mca1DBD-299-600-TAP proteins were synthesized when expressed in pat1-114/pat1-114 mca1Δ/mca1Δ cells during the meiotic program, whole cell extracts were analyzed by immunoblot assays using anti-IgG and anti-α-tubulin antibodies. Results showed that these proteins were present after 1 and 7 h of meiotic induction in the presence of TTM or CuSO 4 (Fig 9G and 9H).
Untagged mca1 + , mca1 1-298+601-697, and mca1DBD-299-600 alleles were expressed in pat1-114/pat1-114 mca1Δ/mca1Δ cells to allow analysis of Pol II chromatin occupancy at the mfc1 + locus. This strategy was used to avoid any interference from the TAP tag with respect to the enrichment of chromatin-bound Pol II to Protein G Sepharose beads. Before entry into meiosis, cells were exposed to TTM (150 μM) or CuSO 4 (25 μM). Aliquots of cultures were retrieved at distinct time points after meiotic induction, and steady-state levels of mfc1 + , mca1 + , and act1 + mRNAs analyzed by RNase protection assays. In the case of cells expressing untagged mca1 + and mca1DBD-299-600 alleles, mfc1 + transcripts showed an expression profile similar to that observed in the case of cells transformed with mca1 + -TAP and mca1DBD-299-region relative to a 18S ribosomal DNA coding region. Data were calculated as values of the largest amount of chromatin measured (fold enrichment). Results are shown as the averages ± S.D. of a minimum of three independent experiments. The asterisks correspond to p<0.001 ( ÃÃ ) and p<0.0001 ( ÃÃÃÃ ) (paired Student's t-test). https://doi.org/10.1371/journal.pone.0201861.g008 Copper starvation-dependent induction of Mfc1 600-TAP (Figs 9B and 10A). Under copper-limiting conditions, mfc1 + mRNA levels were strongly induced after 5 and 7 h of meiotic induction (Fig 10A). Conversely, copper supplementation resulted in strongly reduced mfc1 + mRNA levels in comparison to those observed with cells grown under low copper conditions (Fig 10A). In cells expressing Mca1 1-298+601-697 and an empty vector, transcript levels of mfc1 + remained low, irrespective of cellular copper status (Fig 10A). We next analyzed mca1 + mRNA levels of cells expressing the indicated mutant alleles as a function of copper availability. Results showed that untagged mca1 + , mca1 1-298+601-697, and mca1DBD-299-600 alleles were expressed and their steady-state mRNA levels were constitutive and unresponsive to cellular copper status (Fig 10B).

Discussion
Four proteins have been identified to ensure copper transport in meiotic S. pombe cells. These proteins are Ctr4, Ctr5, Ctr6, and Mfc1 [6,15]. The heteromeric cell-surface Ctr4-Ctr5 complex is predicted to play a critical role for the acquisition of the primary pool of copper from the environment shortly after induction of meiosis [15]. In the case of Ctr6, the protein localizes to vacuolar membranes in early meiosis and then undergoes redistribution in a timedependent manner to reach forespore membranes [15]. Ctr6 serves to mobilize intravacuolar stores of copper and participates in the delivery of copper to the cytosolic copper-dependent enzyme Sod1 [15,44]. Ctr4/5 and Ctr6 are expressed in mitosis and at precise stages of meiosis, whereas Mfc1 is exclusively expressed during meiosis [6,15]. Mfc1 is transcriptionally induced throughout meiotic divisions and spore maturation process under low copper conditions. After meiotic divisions and FSM closure, Mfc1 resides at the FSM where it mediates copper The amino acid sequence of the indicated protein is numbered relative to its initiator codon. B, pat1-114/pat1-114 mca1Δ/mca1Δ cells expressing TAP-tagged mca1 + /mca1 + , mca1 1-298+601-697/ mca1 1-298+601-697, and mca1DBD-299-600/ mca1DBD-299-600 alleles or an empty vector (vector alone) were synchronously induced to undergo meiosis under copper-depleted (TTM, 150 μM) and copper-replete (Cu, 25 μM) conditions. Total RNA was isolated from culture aliquots and steady-state levels of mfc1 + and act1 + mRNAs were analyzed at different time points after meiotic induction. Results shown are representative of three independent RNase protection experiments. C-E, ChIP assays were performed from aliquots of cultures used in panel B after 1 and 7 h of meiotic induction. Binding of Mca1-TAP, Mca1 1-298+601-697-TAP or Mca1DBD-299-600-TAP to the mfc1 + promoter was calculated as the enrichment of the specific mfc1 + proximal regulatory region (positions -183 to -32) relative to a 18S ribosomal DNA coding region. Data were calculated as values of the largest amount of chromatin measured (fold enrichment). Results are shown as the averages ± S.D. of a minimum of three independent experiments using separate chromatin preparations. F, Schematic representation of the mfc1 + gene in which ChIP1 region indicates positions of the primers that were used for qPCR analysis. Black boxes depict wild-type TCGGCG sequences. G-H, Whole-cell extracts were prepared from aliquots of cultures used in panel B after 1 and 7 h of meiotic induction. Samples were analyzed by immunoblot assays using anti-IgG and anti-α-tubulin antibodies to detect TAP-tagged proteins and α-tubulin steady-state protein levels, respectively. https://doi.org/10.1371/journal.pone.0201861.g009 Copper starvation-dependent induction of Mfc1 At the indicated times after meiotic induction, total RNA was extracted from culture aliquots and steady-state mRNA levels of mfc1 + (panel A) and mca1 + (panel B) were analyzed by RNase protection assays using actin (act1 + ) as an internal control. The results shown are representative of three independent experiments. C, Schematic representation of the mfc1 + gene in which the ChIP2 region indicates positions of the primers that were used for qPCR analysis. D-F, At the 1 and 7-h meiotic time points, aliquots of the cultures described for panels A and B were taken and chromatin was prepared and immunoprecipitated using antibodies against RNA Pol II CTD. ChIP results are presented as accumulation into prespores [6]. As opposed to the copper starvation-dependent expression profiles of ctr4 + , ctr5 + , and ctr6 + transcripts that are regulated by Cuf1, the induction of mfc1 + gene expression is primarily dependent of Mca1 [16]. When the mfc1 + promoter was analyzed by ChIP assays, results showed that maximal transcriptional induction of mfc1 + was observed in response to copper deficiency and that Mca1 was constitutively bound to the mfc1 + promoter irrespective of changes in copper concentrations. Additional mfc1 + ChIP analysis showed that optimal induction of mfc1 + transcription coincided with maximal chromatin occupancy by RNA Pol II under low copper conditions. In contrast, there was a lack of RNA Pol II recruitment at the mfc1 + locus in response to excess copper.
The fact that Mca1 was constitutively bound to its target DNA sites upstream of mfc1 + was reminiscent of the situation observed for other members of the Zn 2 Cys 6 binuclear cluster family of transcription factors. In S. cerevisiae, the zinc cluster transcriptional activator Put3 is constitutively bound to its DNA-recognition sites [45,46]. In the presence of proline as the sole source of nitrogen, there is a conversion of DNA-tethered Put3 from a transcriptionally inert form to a transcriptionally active form [46,47]. Accordingly, the presence of proline is predicted to induce a conformational change in the DNA-bound form of Put3, thereby unmasking its activation domain and enabling the recruitment of the RNA Pol II transcriptional machinery [46]. The S. cerevisiae Gal4 protein is another example of zinc cluster transcription factor that is constitutively bound to cis-acting regulatory elements upstream of the GAL genes when galactose is present as the sole source of carbon [48]. The DNA-binding activity of Gal4 does not play a major role in controlling transcriptional activation of the GAL genes by Gal4. In the absence of galactose, the activity of DNA-bound Gal4 is inhibited by its association with the transcriptional repressor Gal80 [46,49]. Conversely, in cells undergoing a transition from a glucose to galactose source, the galactose sensor Gal3 becomes active and interacts with the transcriptional repressor Gal80. The net result of Gal3-Gal80 association is that Gal4 becomes active and induces the expression of the GAL genes [46,50]. In the case of chromatin-bound Mca1, its ability to induce transcription of mfc1 + in response to copper deficiency occurred only during meiosis but not in cells undergoing mitotic growth. This observation raises the possibility that a meiosis-specific transcriptional inducer may trigger an active state of DNAbound Mca1. An alternative possibility is that a meiosis-specific factor may inactivate an inhibitory partner of Mca1. In both cases, this would result in the freeing of Mca1 from an inactive state, resulting in transcriptional activation of mfc1 + to occur during specific stages of the meiotic differentiation program. This mechanism would require the meiosis-specific transcriptional inducer to respond to changing environmental copper concentrations because its function is required to induce a switch in Mca1 transcriptional activity to regulate mfc1 + expression as a consequence of copper starvation. Another potential mechanism could be that copper-dependent alterations in Mca1 activity might arise from the formation of a heterodimer complex between Mca1 and a second related Zn 2 Cys 6 cluster transcription factor.
We have identified an internal region (residues 299 to 600) of Mca1 that was required for RNA Pol II-dependent timely activation of the meiotic mfc1 + gene in response to low copper concentrations. Interestingly, this internal region of Mca1 could be aligned with a region encompassing two inhibitory domains, called ID2 and ID3, that have been identified in the S. cerevisiae Gal4 protein [21,22]. Although the role of Gal4 ID2 and ID3 domains is not well understood, it is thought that they may participate in two functions of Gal4. First, ID2 and ID3 enrichments of mfc1 + (+875/+960) coding region relative to a 18S ribosomal DNA coding region. Data were calculated as values of the largest amount of chromatin measured (fold enrichment). Results are shown as the averages ± S.D. of a minimum of three independent experiments using separate chromatin preparations. The asterisks correspond to p<0.001 ( ÃÃ ) and p<0.0001 ( ÃÃÃÃ ) (paired Student's t-test). https://doi.org/10.1371/journal.pone.0201861.g010 Copper starvation-dependent induction of Mfc1 may play a role in restricting transactivation in the absence of activating cues. Second, after relieving DNA-bound Gal4 inhibition, the internal region of Gal4 (containing ID2 and ID3) may serve as a spacer to augment transcription and/or it may be involved in protein-protein interactions with transcriptional coactivators to promote target gene activation. In the case of S. cerevisiae Gal4, the transcriptionally active form of DNA-bound Gal4 interacts directly with the SAGA co-activator-histone-modifying complex, specifically with the Tra1 subunit [51,52]. Other subunits of SAGA such as Spt3 and Spt8 associate directly with components of the RNA Pol II Mediator complex, which contains a number of highly conserved subunits that can interact directly with RNA Pol II, leading to the formation of a RNA Pol II holoenzyme [53]. Based on these results in S. cerevisiae, one can envision that activation of the transcriptional properties of DNA-bound Mca1 may also involve successive series of interactions with components of SAGA, the fission yeast Mediator complex, and the RNA Pol II, leading to the formation of a competent RNA Pol II holoenzyme for regulated expression of mfc1 + .
Adjacent to the internal ID2-and ID3-like domains of Mca1, an ID1-like region encompassing amino acid residues 151 to 298 was removed (Mca1DBD-299-697 mutant) to test whether this deletion would affect the function of Mca1. Results showed that the Mca1DBD-299-697 mutant protein was active and yielded appropriate timely meiotic induction of mfc1 + gene expression under low copper conditions. As observed in the case of cells expressing the wild-type protein and mutant cells expressing the mca1DBD-299-697 allele in the presence of copper, results showed lack of inducing mfc1 + transcription. These results suggested that the ID1-like region of Mca1 was not required for maintaining Mca1 transcriptionally inactive under copper-replete conditions. The bulk of these observations led us to conclude that ID2and ID3-like domains but not the ID1-like region, could be considered as a middle homology region (MHR) with a role in regulating the transcriptional activity of Mca1.
As of now, it is not known how limiting copper concentrations can trigger a signal for activation of mfc1 + gene expression in a Mca1-dependent manner. In the case of the copper-sensing transcription factor Cuf1, a highly conserved C-terminal motif containing the Cys 328 -X-Cys 330 -X 3 -Cys 334 -X-Cys 336 -X 2 -Cys 339 -X 2 -His 342 sequence (termed C-rich) constitutes its copper-sensing region [39]. A current model for activation by Cuf1 posits that, the copper-free form of Cuf1 binds to its target DNA sequences (denoted copper-signaling elements) within the promoter of the regulated genes to induce transcription. In contrast, when copper is in excess, Cuf1 is sequestered in the cytoplasm through a process that requires the presence of its C-terminal C-rich sequence, leading to cytosolic accumulation of Cuf1 [41]. Furthermore, excess copper promotes an interaction between the Cuf1 N terminus and its C terminal region encompassing the C-rich sequence. This intramolecular interaction fosters the shutoff of Cuf1 in two ways. First, Cuf1 is inhibited from entering the nucleus. Second, the pre-existing nuclear pool of Cuf1 is exported from the nucleus to the cytoplasm, resulting in down regulation of Ctr4/5/6 gene expression [41,42]. In the case of Mca1, the protein does not contain a C-rich-like motif, which suggests that Mca1 regulates mfc1 + gene expression through a mechanism different from that of Cuf1. Within the internal region of Mca1 that encompasses amino acid residues 299 to 600, two potential metal-binding motifs (Cys 426 -X 3 -Met-His-X 3 -His 478 and Met 476 -X-Met 478 ) are present. We have examined the potential role of these motifs in copper starvation-dependent regulation of mfc1 + by generating mutant alleles of mca1 + in which all of the Cys, Met, and His residues had been mutated to alanines. Each Mca1 mutant protein containing either the Ala 426 -X 3 -Ala-Ala-X 3 -Ala 478 or Ala 476 -X-Ala 478 mutated motif had no effect on the expression of the mfc1 + gene as compared to that of TTM-mediated activation of mfc1 + transcription in cells expressing a wild-type mca1 + allele ( Figure B in S1 File).
Due to the fact that Mca1 was constitutively bound to its target promoter, modulation of the DNA-binding activity of Mca1 does not appear to play a major role in controlling the activity of the genetic switch in response to induction of mfc1 + by copper starvation. Mechanisms that modulate transcriptional activities of members of the zinc cluster transcription factor family include interaction with small-molecule metabolites, modification of phosphorylation state, and recruitment of co-regulators and partner nutrient sensors [17,20]. Because an internal region (residues 299 to 600) of Mca1 contributes to the activation of mfc1 + in response to copper starvation, it is likely that an unidentified meiosis-specific factor or timely posttranslational modification mediates a conformational change in the Mca1 protein required to convert Mca1 from a transcriptionally inert form to a transcriptionally active form, leading to induction of mfc1 + expression in response to copper deficiency.