Yap7 is a Transcriptional Repressor of Nitric Oxide Oxidase in Yeasts, which arose from Neofunctionalization after Whole Genome Duplication

Yap7 is a transcriptional repressor of nitric oxide oxidase inyeasts, which arose from neofunctionalization after wholegenome duplication

Jawad Merhej,1,2 Thierry Delaveau,1,2

Juliette Guitard,3,4,5,6 Benoit Palancade,7

Christophe Hennequin,3,4,5,6 Mathilde Garcia,1,2

Gaëlle Lelandais7 and Frédéric Devaux1,2*1Sorbonne Universités, UPMC Univ. Paris 06, Institut deBiologie Paris Seine UMR 7238, Laboratoire de biologiecomputationnelle et quantitative, F-75006, Paris, France.2CNRS, UMR 7238, Laboratoire de biologiecomputationnelle et quantitative, F-75006, Paris, France.3Sorbonne Universités, UPMC Univ Paris 06, CR7,Centre d’Immunologie et des Maladies Infectieuses(CIMI-Paris), 91 Bd de l’hôpital, F-75013, Paris, France.4Inserm, U1135, CIMI-Paris, 91 Bd de l’hôpital, F-75013,Paris, France.5Assistance Publique–Hôpitaux de Paris, Hôpital StAntoine, Service de Parasitologie-Mycologie, F-75012,Paris, France.6CNRS, ERL 8255, CIMI-Paris, 91 Bd de l’hôpital,F-75013, Paris, France.7Institut Jacques Monod, CNRS, UMR 7592, Univ ParisDiderot, Sorbonne Paris Cité, F-75205 Paris, France.

Summary

Flavohemoglobins are the main detoxifiers of nitricoxide (NO) in bacteria and fungi and are induced inresponse to nitrosative stress. In fungi, the flavohe-moglobin encoding gene YHB1 is positively regu-lated by transcription factors which are activatedupon NO exposure. In this study, we show that inthe model yeast Saccharomyces cerevisiae and inthe human pathogen Candida glabrata, the transcrip-tion factor Yap7 constitutively represses YHB1 bybinding its promoter. Consequently, YAP7 deletionconferred high NO resistance to the cells. Co-immunoprecipitation experiments and mutant analy-ses indicated that Yap7 represses YHB1 by recruitingthe transcriptional repressor Tup1. In S. cerevisiae,YHB1 repression also involves interaction of Yap7

with the Hap2/3/5 complex through a conserved Hap4-like-bZIP domain, but this interaction has been lost inC. glabrata. The evolutionary origin of this regulationwas investigated by functional analyses of Yap7 and ofits paralogue Yap5 in different yeast species. Theseanalyses indicated that the negative regulation ofYHB1 by Yap7 arose by neofunctionalization after thewhole genome duplication which led to the C. glabrataand S. cerevisiae extant species. This work describesa new aspect of the regulation of fungal nitric oxidaseand provides detailed insights into its functioning andevolution.

Introduction

Flavohemoglobins (FlavoHbs) are dual proteins whichhave a globin, heme containing, domain in N-terminal anda flavin-containing oxidase domain in C-terminal (reviewedin Forrester and Foster, 2012). FlavoHbs function as nitricoxide (NO) oxidoreductases, efficiently turning NO intonitrate in aerobic conditions. NO is a normal by-product ofmetabolic activities in oxygen and nitrogen rich environ-ments. At high concentrations, it is toxic to the cells bydirectly or indirectly (through the formation of reactivenitrogen species) damaging proteins, lipids and DNA, andinterfering with iron metabolism and the activity of hemecontaining proteins (reviewed in Ischiropoulos and Gow,2005). Moreover, NO is one of the chemical weapons usedby the cells of the innate immune system to fight againstpathogenic microbes (Forrester and Foster, 2012). Fla-voHbs are present in most bacteria, fungi and protozoaespecies, but have been lost in archea and in higher eukary-otes. They have been shown to actually confer NO resist-ance in many different species of bacteria (Vasudevanet al., 1991; Crawford and Goldberg, 1998; Pathania et al.,2002), fungi (Liu et al., 2000; de Jesus-Berrios et al., 2003;Ullmann et al., 2004) and protozoae (Iijima et al., 2000).The expression of FlavoHbs is increased in response tonitrosative stress in all microorganisms in which it wasexamined. In bacteria, FlavoHb expression is constitutivelyrepressed by strong negative transcriptional regulators.This repression is relieved by NO which dissociates therepressor from the DNA (Bodenmiller and Spiro, 2006;

Accepted 24 February, 2015. *For correspondence. E-mail [email protected]; Tel. (+33) 1 4427 8140; Fax: (+33) 1 4427 7336.

Molecular Microbiology (2015) ■ doi:10.1111/mmi.12983

© 2015 John Wiley & Sons Ltd

mailto:[email protected]

mailto:[email protected]

Tucker et al., 2008; 2010). The situation is markedly dis-tinct in fungi, in which nitric oxidase expression is inducedby positive regulators in response to NO. The yeastsCandida glabrata and Saccharomyces cerevisiae haveone flavoHb encoding gene which is called YHB1 andwhich is about 30% identical to its bacterial homologues(Zhu et al., 1992). Candida albicans has three flavoHbencoding genes (YHB1, YHB4 and YHB5) but only Yhb1seems to harbor NO dioxygenase activity (Ullmann et al.,2004; Hromatka et al., 2005). YHB1 has been shown to beinduced by NO in C. albicans, S. cerevisiae and Aspergil-lus nidulans (Ullmann et al., 2004; Hromatka et al., 2005;Sarver and DeRisi, 2005) and by the presence of mac-rophages in C. albicans (Lorenz et al., 2004). In S. cerevi-siae, a single transcription factor involved in the NOup-regulation of YHB1 has been described: the C2H2 zincfinger Fzf1. The deletion of FZF1 abolishes the NO tran-scriptional response of YHB1 but does not modify its basalexpression (Sarver and DeRisi, 2005). The direct bindingof Fzf1 to the YHB1 promoter has not been formallyestablished but a DNA consensus specific to all Fzf1dependent genes (called CS2) is present 500 base pairsupstream of the YHB1 ATG (Sarver and DeRisi, 2005). InC. albicans, the Zn2Cys6 zinc finger transcription factorCta4 is the functional equivalent of Fzf1. Deletion of CTA4prevents the induction of YHB1 by NO and causes NOhypersensitivity (Chiranand et al., 2008). Cta4 is a pro-moter resident protein which binds the NO responseelement (NORE, position -240/-220) in the YHB1 promoter,even in the absence of NO (Chiranand et al., 2008). Artifi-cial activation of Cta4 by fusion with the Gal4 activatingdomain leads to a moderate increase of YHB1 expression(Schillig and Morschhauser, 2013). In C. albicans, thegeneral transcriptional regulator Ndt80 also binds theYHB1 promoter and contributes to its NO driven induction(Sellam et al., 2010; Yang et al., 2012). Additionally, theZn2Cys6 transcription factor Cwt1 of C. albicans binds tothe promoter of YHB1 with or without NO and has beendescribed as a negative regulator of YHB1 induction(Sellam et al., 2012). However, the deletion of CWT1 hasno effect on the basal expression of YHB1 and only tran-siently and moderately increases its induction in responseto NO (Sellam et al., 2012). Beside its regulation by NO,YHB1 level of expression has been shown to be decreasedby oxidative stress and repressed by hypoxia in S. cerevi-siae, repressed by iron starvation in C. albicans andinduced by hypoxia and nitrate assimilation in Aspergillusspecies (Buisson and Labbe-Bois, 1998; Lan et al., 2004;Hickman and Winston, 2007; Schinko et al., 2010; Hsuet al., 2011; Vodisch et al., 2011).

In the present work, we show that Yap7, a transcriptionfactor of previously unknown function, is an importantregulator of YHB1 expression in C. glabrata and S. cerevi-siae. Yap7 belongs to the Yap subfamily of basic leucine

zipper (bZIP) transcription factors. S. cerevisiae has 8members of the Yap family. Seven out of these 8 Yapfactors have been shown to be involved in variousstress responses (reviewed in Rodrigues-Pousada et al.,2010). Yap7 is the only one for which no function hadbeen proposed. The Yap families of S. cerevisiae andC. glabrata have expanded due to the whole genomeduplication (WGD) which occurred in one of their commonancestors (Scannell et al., 2007) and Yap7 belongs to apair of ohnologues, which means that it has a paraloguewhich resulted from the WGD. This ohnologue, calledYap5, contributes to the cell response to iron overload inS. cerevisiae (Li et al., 2008; Pimentel et al., 2012). Yap5and Yap7 are orthologues to the C. albicans transcriptionfactor Hap43 and to the HapX regulators found in filamen-tous ascomycetes (e.g. Aspergillus and Fusarium species)and in basidiomycetes (e.g. Cryptococcus neoformans). InC. albicans, Hap43 controls iron homeostasis by repress-ing the expression of iron consuming genes (Homannet al., 2009; Hsu et al., 2011; Singh et al., 2011). Similarly,the deletion of HapX impairs iron homeostasis in thehuman pathogens Aspergillus fumigatus and Cryptococ-cus neoformans (Jung et al., 2010; Schrettl et al.,2010) and in the plant pathogen Fusarium oxysporum(Lopez-Berges et al., 2012). The particularity of the HapXproteins is to contain a bipartite HAP4L-bZIP domain,which is also found in the Yap7 protein from S. cerevisiae(Singh et al., 2011). Hap4 is the regulatory subunit of theyeast CCAAT box binding complex (CBC), which recruitsthe Hap2, Hap3 and Hap5 proteins to control the expres-sion of genes involved in the respiratory pathway in S. cer-evisiae (McNabb et al., 1995). The Hap4L domain is a 20amino acid motif which allows Hap4 to interact with CBC(McNabb and Pinto, 2005). The role of the HAP4L-bZIPbipartite domain has been documented in the cases ofHap43 in C. albicans and HapX in filamentous ascomy-cetes (Hortschansky et al., 2007; Singh et al., 2011). LikeHap4, Hap43 and HapX use their HAP4L domain to inter-act with the CBC complex. This interaction is necessaryto fully repress the expression of their target genes(Hortschansky et al., 2007; Jung et al., 2010; Schrettlet al., 2010; Hsu et al., 2011; Singh et al., 2011). Recently,the HapX protein of Aspergillus fumigatus and its orthologsfrom Aspergillus. nidulans and Fusarium oxysporum werefound to be involved also in iron resistance by activatingvacuolar iron import genes. This function also requiresinteraction with the Hap/CBC complex (Gsaller et al.,2014).

In this work, we show that Yap7 strongly represses YHB1expression in C. glabrata and S. cerevisiae by directlybinding its promoter. C. glabrata cells in which Yap7 hasbeen inactivated were highly resistant to NO. This regula-tion occurs in optimal growth conditions and is largelyindependent of the presence of external NO. To exert

2 J. Merhej et al. ■

© 2015 John Wiley & Sons Ltd, Molecular Microbiology

repression, Yap7 recruits the general repressor Tup1, bothin S. cerevisiae and C. glabrata. Our results suggest that,in S. cerevisiae, YHB1 repression also involves the Yap7evolutionarily conserved HAP4Like-bZIP bipartite domain,which allows interaction with the Hap2/3/5 regulatorycomplex. This mechanism is similar to the one used byHap43 and HapX proteins to repress iron consuminggenes (Hortschansky et al., 2007; Jung et al., 2010; Hsuet al., 2011; Singh et al., 2011). Finally, analyses of the roleof Yap7 and Yap5 in several yeast species indicate thatYHB1 repression is a relatively recent acquisition whicharose from the neofunctionalization of Yap7 after the wholegenome duplication.

Results

YAP7 is a strong repressor of nitric oxide oxidaseexpression in S. cerevisiae

Our initial aim was to analyze the function of Yap7 in theframe of a global analysis of the Yap transcription factors inS. cerevisiae and in C. glabrata (Merhej et al., in prepara-tion). We first compared the protein sequences of Yap7from C. glabrata, S. cerevisiae and other post-WGD yeastspecies (Figure 1A). The ScYap7 reference sequence(coming from the S288c laboratory strain) was surprisinglyshort compared with its orthologues, even when thespecies were very closely related to S. cerevisiae. Moreprecisely, the ScYap7 sequence contains the typical YapbZIP DNA binding site, but lacks the regulatory domainspresent in the C-terminal part of the other Yap7 proteins(Figure 1A) (Rodrigues-Pousada et al., 2010). Lookingmore carefully at the S288c ScYap7 sequence and at itsgenomic environment in the Saccharomyces GenomeDatabase, we found that this was due to a frame-shift inScYAP7 caused by a single nucleotide deletion in position708 of the coding sequence (Supplementary Data S1).Hence, the neighboring ORF YOL029c actually corre-sponds to the C-terminal part of the Yap7 proteins whichare found in other species (Supplementary Data S1). Tostate if this mutation was a general feature of S. cerevisiaeor a specificity of the S288c strain, we looked at the DNAsequences of the YAP7/YOL029c loci from 38 differentS. cerevisiae strains whose genome sequences wereavailable on the web (Liti et al., 2009; Bergstrom et al.,2014).Among the five ‘pure’ (i.e. non mosaic) S. cerevisiaelineages that were previously defined (Liti et al., 2009),three forms of ScYap7 were observed, which differed bytheir size (Figure 1B). The short form (245 AA) harboringthe frame-shift described above was found in the wine/European lineage and in most of the commonly usedlaboratory strains. In contrast, the Malaysian, North Ameri-can and African lineages encode a much longer ScYap7protein (518 AA), which is not interrupted by the frame-shift

and which size is close to the sizes of the Yap7 proteinsfound in other Saccharomyces sensu stricto species(Figure 1Aand B). Finally, the ScYap7 proteins found in thesake lineage are of intermediate lengths (with two possiblevariants of 439 and 468 AA respectively) due to a prema-ture stop codon close to the C-terminus of the codingsequence (Figure 1B).

To study the role of the short and full-length ScYap7proteins, transcriptome analyses were conducted compar-ing Scyap7Δmutants constructed either in the S288C(short Yap7) or in the North-American (full-length Yap7)genetic backgrounds to their respective isogenic wild typestrains (Figure 1C and D). The deletion of the full-lengthYap7 protein in the North-American background led to adramatic increase (about 64 fold) in YHB1 expressionlevel, which was by far the largest effect measured(Figure 1C), and which was not observed when the shortform of Yap7 was deleted in the S288C background(Figure 1D). These microarray results were validated byQ-RTPCR analyses (Supplementary data S2). Based onthese results, we focused our work on this strong andpreviously unknown regulatory interaction between Yap7and YHB1. To confirm that the difference observedbetween the two S. cerevisiae genetic backgrounds wasmostly due to mutations in the YAP7 gene and not todifferences in other genes, the S288C version of YAP7 wasreplaced by the ScYAP7 full-length gene coming from theNorth American lineage (YPS128). Notably, the short andlong variants found in these lineages are 97% identicalover the 245 amino acids shared by the two forms and100% identical between position 3 and 238, which containsthe DNA binding domain. As a control, a S288C mutantstrain was constructed, in which the native short ScYap7was replaced by another short form of ScYap7 coming froma wine/European strain (DVBPG6765). The expression ofthe long form of ScYap7 in S288c severely decreased theexpression levels of ScYHB1 (Figure 1E).As expected, theshort form coming from DVBPG6765 had no effect com-pared with the native S288C short form. ChIP-PCR ex-periments with myc-tagged versions of these proteins con-firmed that the short and full-length Yap7 versions wereable to bind to the ScYHB1 promoter (Figure 1F), which isconsistent with the fact that they all contain the Yap7 bZIPDNA binding domain, which is 100% identical between thedifferent variants (figure 1A). These experiments indicatedthat the full-length ScYap7, which is found in the Malaysian,North-American and African lineages, is a strong constitu-tive repressor of YHB1. We concluded that the loss offunction observed in S288c was only due to mutation(s) inYAP7 in this strain, since the replacement of the short formby the long one restores the repression.

In conclusion of this part, Yap7 functions as a strongrepressor of YHB1 in most S. cerevisiae lineages bydirectly binding to the YHB1 promoter. This role has

New insights in the regulation of nitric oxidase 3




been lost, possibly due to the frame-shift mutation, in theS. cerevisiae laboratory strains and in the wine/Europeanlineage.

YAP7 functions independently of NO and of Fzf1, themajor regulator of NO response in S. cerevisiae

As mentioned in the introduction, the zinc finger transcrip-tion factor Fzf1 has been described as being responsiblefor YHB1 induction in response to NO in S. cerevisiae.However, these results were obtained in the S288C back-ground, in which the Yap7 regulation of YHB1 wasdisabled. To assess the potential functional interactionsbetween full-length Yap7 and Fzf1, we constructed fzf1Δ

and fzf1Δyap7Δ mutants in the S. cerevisiae North Ameri-can genetic background and measured the NO-mediatedinduction of YHB1 in these strains (Figure 2A). Asexpected, the results obtained in the absence of YAP7recapitulated what was previously described in the S288Cbackground (Sarver and DeRisi, 2005): YHB1 had a highbasal expression level, it was moderately induced by NO(about two fold, compare the orange histograms at 0 and50 minutes) and this induction was totally impaired by thedeletion of FZF1, which had no effect on the YHB1 basallevel (compare the green and orange histograms). In thepresence of YAP7, the basal level of YHB1 was very low,YHB1 was strongly induced by NO (about 200 fold, redhistograms) and this induction was severely reduced by

Fig. 1. Yap7 role in the regulation of YHB1.A. Schematic representation of Yap7 proteins from 9 post whole genome duplication yeast species: V. polyspora (XP_001646788.1),K. africana (XP_003956919.1) S. bayanus (GenBank: AACA01000018.1), S. kudriavsevii (GenBank: EJT43076.1), S. mikatae (GenBank:AABZ01000441.1), S. cerevisiae (NP_014614.1), C. glabrata (XP_446029.1), C. nivariensis and N. delphensis (Gabaldon et al., 2013). ThebZIP domain is indicated by the dark blue box. The cystein-rich domains are indicated by yellow boxes. The species in green and blue belongto Saccharomyces and Nakazeomyces clades, respectively. The size of each protein is indicated on the right of the cartoon.B. The neighbor joining tree of S. cerevisiae strains was generated based on the whole-genome SNPs between 38 S. cerevisiae strainscollected from different geographic origins and hosts as described in (Liti et al., 2009). The strains with clean lineages are highlighted in grey.The color of the boxes at the end of the tree branches indicates the size of the Yap7 protein in the corresponding strain.C. Genome-wide transcriptional analysis of the yap7Δ mutant compared to S. cerevisiae wild type strain from North American background(YLF131) grown in YPD medium. The values for each gene are the means of two biological replicates.D. Genome-wide transcriptional analysis of the yap7Δ mutant compared to S. cerevisiae wild type strain from S288C background grown inYPD medium. The values for each gene are the means of two biological replicates.E. Expression of YHB1 measured by Q-RTPCR in S. cerevisiae strains containing different forms of ScYAP7; Sc::WT: BY4742 (S288C) wildtype strain containing its own endogenous short form of ScYAP7 (245 aa), Sc::YAP7-sh: BY4742 containing the coding sequence of a shortform of ScYAP7 from DVBPG6765 European strain 245 aa. Sc::YAP7-ln: BY4742 containing the coding sequence of a long form (518 aa) ofScYAP7 from YPS128 North American strain, Sc::CgAP7: BY4742 containing YAP7 from C. glabrata. For transcriptional analysis, Q-RTPCRwas performed using primers located in the coding sequence of ScYHB1 and ScACT1. The values are the expression values of YHB1 relativeto the ACT1 values used as an internal normalization control.F. ChIP-PCR performed on the Yap7 versions described just above. For ChIP-PCR analysis, Q-RTPCR was performed using primers locatedin the promoters of ScYHB1 and ScACT1 (used as the reference for normalization) genes.

Fig. 2. Yap7 acts independently of NO and of the major NO regulator Fzf1 in S. cerevisiae.A. Expression of YHB1 measured by Q-RTPCR in the indicated strains. All strains are from the S. cerevisiae North American YLF131background. Cells were grown in minimal medium buffered to pH7.4 and DPTA-NONOate was added to a final concentration of 1 mM at timezero. Cultures were harvested after 50 min of DPTA-NONOate treatment. Q-RTPCR was performed using primers located in the codingsequence of YHB1 and ACT1 (as endogenous control). Error bars represent the standard error of the mean.B. ChIP-PCR was performed using anti c-myc antibody on cultures of S. cerevisiae BY4742 untagged wild type strain (mock IP) andmyc-tagged Sc::YAP7-ln strain with or without 50 min treatment with 1 mM of DPTA-NONOate. Q-RTPCR was performed using primerslocated in the promoters of YHB1 and ACT1. Error bars represent the standard error of the mean.



the single deletion of Fzf1 (to about 1% of the wild typeinduction, compare the red and blue histograms at 50minutes). Notably, the level of expression of YHB1 inpresence of NO and absence of Fzf1 was about 100 foldlower than in the absence of YAP7 (compare the blue andgreen histograms). According to ChIP-qPCR experiments(Figure 2B), NO treatment led to a small decrease of Yap7binding to YHB1 (by about 1.5 fold) which may be reliableto the very modest NO-driven increase of YHB1 expres-sion in the absence of Fzf1 (Figure 2A, blue histograms).

In conclusion of this part, the Yap7 repression of YHB1is mostly insensitive of the presence of NO and Fzf1 isconfirmed as the major regulator of NO-driven YHB1induction in S. cerevisiae.

The transcriptional repressor Tup1co-immunoprecipitates with the full-length Yap7 and isrequired for YHB1 repression

To screen for proteins which could contribute to the tran-scriptional repression of YHB1 together with Yap7, weconstructed protein A-tagged versions of the short andfull-length forms of ScYap7 that we inserted at the ScYap7locus in the S288C background. Immunoprecipitationexperiments (IP) were performed using IgG and theco-immunoprecipitating protein partners were then identi-fied by mass spectrometry. A control experiment was per-formed using an untagged strain in order to identifypotential contaminants. A total of 296 proteins weredetected in the IP targeting the full-length ScYap7. Fromthose, only 61 were absent from the untagged control andwere considered as being specifically enriched in the IP.Because we were interested in finding potential partners ofYap7 that act at the YHB1 promoter, we focused on the 46proteins which were annotated as being localized in thenucleus (Supplementary Data S3). This list was enriched inproteins involved in chromatin remodeling, in general tran-scription factors and in few specific transcription factors.However, the vast majority of these proteins were alsofound in the IP targeting the short form of Yap7, whichbinds to YHB1 but is unable to repress its expression(supplementary Data S3). Only two nuclear proteinsco-immunoprecipitated with the full-length form of ScYap7but not with the short one: Yol029c, which corresponds tothe C-terminal part of the full-length ScYap7 and which isnot present in the short form of ScYap7, and the transcrip-tional repressor Tup1. We confirmed these results by per-forming Western blot analyses of our input and IP samples,using anti-Tup1 antibodies. In accordance with the massspectrometry results, Tup1 could only be detected in the IPwhen the long form of Yap7 was used as bait (Figure 3Aand B). To assess the role of Tup1 in the repression ofYHB1 in S. cerevisiae, we deleted ScTUP1 in the NorthAmerican S. cerevisiae background. RT-QPCR analyses

of YHB1 expression in the tup1Δ mutants showed that theinactivation of TUP1 leads to a strong derepression ofYHB1 expression which mimics the inactivation of YAP7(Figure 3C). Importantly, in S. cerevisiae, the doubletup1Δyap7Δ mutant showed YHB1 levels similar to the twosingle mutants, supporting the fact that Yap7 and Tup1collaborate in the constitutive repression of YHB1(Figure 3C). To gain further insights into this regulation, weperformed ChIP-PCR experiments targeting Tup1 inS288C strains expressing either the long or the short formof Yap7. These experiments showed that Tup1 bound theYHB1 promoter only in the presence of the full-length Yap7(Figure 3D). This result ruled out the possibility that Tup1could be recruited at the YHB1 promoter independently ofYap7 and supports the existence of an interaction betweenthe two proteins. Additionally, we tested the influence ofnitric oxide on Tup1 binding to YHB1. NO treatment mod-erately diminished (2 fold) Tup1 occupancy at YHB1 pro-moter (Figure 3D), which is consistent with our previousresults on Yap7 binding (Figure 2B).

The CCAAT binding complex subunit Hap5 contributesto the repression of YHB1 in S. cerevisiae

As mentioned in the introduction, Yap7 is orthologue tothe fungal HapX transcription factors (Singh et al., JBC2011). The particularity of these proteins is to contain abipartite HAP4L-bZIP domain, which they use to interactwith the CCAAT binding complex (CBC) Hap2/3/5 andregulate transcription (Hortschansky et al., 2007; Junget al., 2010; Schrettl et al., 2010; Hsu et al., 2011; Singhet al., 2011; Lopez-Berges et al., 2012; Gsaller et al.,2014). Examination of the HAP4L-bZIP domains (Figure4A) or of the full length Yap5 and Yap7 orthologues indifferent fungal species (Figure 5, Supplementary data S4and S5) indicated that the HAP4L-bZIP domain is con-served from basidiomycetes to the pre-WGD yeastspecies. After the whole genome duplication, the HAP4Lmotif was lost in the two resulting ohnologues, Yap7 andYap5, with the notable exceptions of the Yap7 proteinsfound in the Saccharomyces species (Figures 4A and 5;Supplementary data S5). The presence of a HAP4L-bZIPdomain in ScYap7 raised the interesting possibility that itcould repress YHB1 expression by interacting with theCBC, using a mechanism similar to the one described forHapX proteins. Notably, Hap5 is one of the proteins whichco-immunoprecipitated with the short and full-length Yap7proteins in S. cerevisiae, which both contains the Hap4Ldomain (Supplementary data S3). To test the potentialinvolvement of the CBC complex in the repression ofYHB1, a S. cerevisiae HAP5 knock-out strain was con-structed in the North American genetic background.Q-RTPCR analyses showed that YHB1 expression levelsincreased 34 fold in the hap5Δ strain compared with the



wild type (Figure 4B). Hence, Hap5 contributes to theconstitutive repression of YHB1. Notably, the derepres-sion observed in the hap5Δ strain is much lower as com-pared to the ones measured in the yap7Δ strain (250 foldincrease of YHB1 expression; Figure 4B), or in the tup1Δstrain (Figure 3C). The deletion of HAP5 in the yap7Δstrain did not lead to an additional increase in the YHB1levels (Figure 4B), suggesting that the YAP7 deletion isepistatic on the HAP5 deletion. These results indicate thatthe two proteins cooperate in this process but that only alimited fraction of the repressor activity of Yap7 towardsYHB1 is mediated by Hap5. In this model, Yap7 would

interact with Hap5 through its HAP4L domain, taking theplace of the Hap4 regulatory subunit. Hence, Hap4 is notexpected to play a role in the regulation of YHB1. To testthis hypothesis, we deleted HAP4 in the North AmericanS. cerevisiae background. This inactivation of HAP4 hadno effect on YHB1 expression (Figure 4B), reinforcing theidea that Yap7 may directly interact with the CBC complexto regulate YHB1.

In conclusion of this part, the Yap7 proteins from theSaccharomyces species differ from their orthologues inother post-WGD yeast clades by having conserved aHAP4L-bZIP domain. This domain confers to ScYap7 the

Fig. 3. Tup1 and Yap7 are co-repressors of YHB1 in S. cerevisiae.A. Western-blot analysis using protein extracts from the co-immunoprecipitation of Yap7 to confirm the physical interaction of Yap7 full-length(YAP7-ln) with Tup1. Protein extracts of IP and inputs from Protein A tagged Yap7 short and full-length forms and from untagged sample wereimmunoblotted using the rabbit IgG-HRP antibody to validate the success of the co-imunoprecipitation experiment. Tup1 was immunoblottedusing Anti-TUP1 antibody. Ded1 was immunoblotted with Anti-DED1 antibody and was used as a control to confirm the specificity of theco-immunoprecipitation.B. The quantification of the signals shown in Figure 3A was performed using the Image J software. The values represented on thesehistograms are IP/Input ratios, averaged from two independent measurements.C. Expression of ScYHB1 was measured by Q-RTPCR in the indicated strains. The wild type strain is the S. cerevisiae North AmericanYLF131 strain. The mutants used were constructed in S. cerevisiae North American background. Q-RTPCR was performed using primerslocated in the coding sequence of ScYHB1 and ScACT1 (as endogenous control). Error bars represent the standard error of the mean.D. ChIP measurements of Tup1 enrichments at YHB1 promoter in different contexts. ChIP-PCR was performed using rabbit anti Tup1antibodies and cultures of S. cerevisiae S288c strains expressing either the long (YAP7-ln) or the short (YAP7-sh) protein. The mock IP wasperformed using the dynabeads with the anti-rabbit IgG but no anti-Tup1 antibody (mock). Tup1 enrichment was also performed on S288Ccells containing the full-length Yap7 50 minutes after the addition of 1 mM DPTA-NONOate (YAP7-ln NO). The enrichment of the YHB1promoter in the Tup1 IP compared to the Input (IP/Input ratio) was measured by Q-RTPCR and normalized with the enrichment of the ACT1promoter. Error bars represent the standard error of the mean.



possibility to interact with the CBC complex and this inter-action apparently contributes, although moderately, to therepression of ScYHB1.

The role of Yap7 is conserved in C. glabrata, but doesnot involve CBC

As mentioned above, the Yap7 protein of the human patho-gen C. glabrata (named Cgap7 thereafter) differs from itsS. cerevisiae orthologue by the absence of any recogniz-able HAP4L domain in its sequence (Figures 4A and 5).This prompted us to test the role of Yap7 in this species. Todo so, the transcriptome of a CgAP7 knock-out strain(called cgap7Δ) was compared to the wild type strain usingDNAmicroarrays. Similarly to what was observed in S. cer-evisiae, CgYHB1 showed an increase of 20 to 30 fold inexpression in the mutant, which was by far the strongesteffect detected (Figure 6A). Quantitative real time PCR(Q-RTPCR) analyses of CgYHB1 expression levels con-

firmed this overexpression in the cgap7Δ mutant(Figure 6B). To establish if Cgap7 was directly regulatingCgYHB1 expression, a myc-tagged version of Cgap7 wascloned in an expression plasmid under the control of itsown promoter. This plasmid was used to transform cgap7Δcells. We controlled that this C-terminal tagging of thetranscription factor did not impair its activity by measuringthe level of expression of CgYHB1 in this cgap7-myc strain(Figure 6B). Chromatin immunoprecipitation experimentsfollowed by Q-RTPCR (ChIP-PCR) showed that theCgYHB1 promoter was enriched 60 fold in the Cgap7-mycIP compared with the CgACT1 promoter (Figure 6C). Thisenrichment was not detected in a mock ChIP-PCR experi-ment conducted on the wild type, untagged, strain. Aspreviously observed in S. cerevisiae, NO treatment did notseverely impair Yap7 binding in C. glabrata (1.5 folddecrease, Figure 6C). Because the role of CgYHB1 in NOresistance had never been addressed directly, we testedthe resistance of wild type cells and cgap7Δcells to differ-

Fig. 4. The CCAAT-binding complex contributes to the repressing activity of Yap7 in S. cerevisiae.A. Alignment of the bipartite domain containing the HAP4L and the bZIP domains in ScHap4, ScYap5, ScYap7, Cgap5 and Cgap7 proteinsand their homologues from 5 pre-whole genome duplication species: L. kluyveri (LkYap5/7; Lakl0F05464g), L. thermotolerans (LtYap5/7;Klth0F12496g), K. lactis (KlYap5/7; Klla0D14399g), C. albicans (Hap43) and A. nidulans (HapX). The alignment was performed using ‘BioEdit’sequence Alignment editor (Hall, 1999). The conserved amino-acids are highlighted. The functional domains (bZIP and HAP4L) of the proteinswere searched using PFAM database and the e-value scores of each domain is given in Supplementary Data S5.B. Expression of ScYHB1 measured by qPCR in the indicated strains. The wild type strain is the S. cerevisiae North American YLF131 strain.The mutants used were all constructed in this S. cerevisiae North American background. Q-RTPCR was performed using primers located inthe coding sequence of ScYHB1 and ScACT1 (as endogenous control). The y-axis is discontinuous and the missing part is indicated by 2parallel slanting bars drawn on the axis and the histograms. Error bars represent the standard error of the mean.



ent doses of external NO (Figure 6D). At doses rangingfrom 1 mM to 3 mM, the cell survival rate in the wild typewas below 50%. In contrast, mutant cells exhibited asurvival rate of more than 80%, even with NO concentra-tions up to 3 mM.

To assess the conservation of the role of CBC and TUP1in YHB1 repression, CgHAP5 and CgTUP1 were deletedand the CgYHB1 expression was measured in the resulting

strains mutants (Figure 6B). As observed in S. cerevisiae,the inactivation of TUP1 led to a strong increase in YHB1expression which mimicked the inactivation of CgAP7(Figure 6B). In contrast, the deletion of CgHAP5 had noeffect on the expression of CgYHB1 (Figure 6B). To assessthe conservation of the mechanisms that allows Yap7 torepress YHB1 in S. cerevisiae and C. glabrata, the inactiveshort form of ScYAP7 was replaced by CgAP7 in the S288c

Fig. 5. Phylogenetic analyses of Yap5 and Yap7. Neighbor-joining protein tree constructed based on the multiple sequence alignment of theYap5 and Yap7 proteins of species from Ascomycota and Basidiomycota species including the basidiomycetes C. Cryptococcus, somePezizomycotina filmentous Ascomycetes and Hemiascomycetes species from Candida, Kluyveromyces, Lachancea, Nakazeomyces andSaccharomyces clades using MEGA6 software. The Whole Genome Duplication is indicated. Post-WGD species belonging to theSaccharomyces clade (red dots) or the Nakazeomyces clade (blue triangle) are indicated. The names of the YAP7 homologues are indicatedfor each group of species. The complete multiple sequence alignment is available in Supplementary Data S4. The accession numbers of theprotein sequences are listed in Supplementary Data S4. The cartoon for each group of proteins is represented on the right side. CRD:Cystein-rich domains. The particular CRD sequence (CGFCX5CXC) which was shown to contribute to iron sensing in Yap5, Yap5/7 and HapXproteins (Gsaller et al., 2014; Rietzschel et al., 2015) is represented in dark blue. LCR: Low complexity domains (Serine-rich domains).Presence or absence of the HAP4L, bZIP and the regulatory domains are presented on the cartoons of the proteins. Note that the HapXprotein tree does not follow the phylogeny of the species.



background. This led to a decrease in ScYHB1 expression,which was similar to the effect obtained with the full-lengthScYap7 (Figure 1E). This repression was associated withdirect binding of Cgap7 to the ScYHB1 promoter, as shownby ChIP-PCR (Figure 1F).

In conclusion of this part, Cgap7 constitutively andstrongly represses the expression of CgYHB1. Impairingthis regulation by deleting CgAP7 made the cells stronglyresistant to NO exposure in vitro. As in S. cerevisiae, thefull repression of YHB1 requires Tup1. In contrast, it doesnot involve Hap5, which may be reliable to the absence of

the HAP4L domain in Cgap7. Despite this difference, thefunctioning of Yap7 in C. glabrata and S. cerevisiae wasconserved enough, so that Cgap7 could restore YHB1repression in a Δyap7 S. cerevisiae mutant.

The Yap7-dependent regulation of YHB1 arose fromneofunctionalization after the whole genome duplication

As emphasized in Figure 5, YAP7 constitutes with YAP5 apair of ohnologues which resulted from the whole genomeduplication (WGD) that occurred in a common ancestor of

Fig. 6. Cgap7 is a repressor of YHB1 in C. glabrata.A. Genome-wide transcriptional analysis of CgAP7 deletion mutant (cgap7Δ) compared to C. glabrata wild type strain grown in YPD medium.B. Expression of CgYHB1 measured by Q-PCR in wild type, cgap7Δ, myc-tagged Cgap7 (CgAP7-myc), tupΔ, hap5Δ strains in YPD medium.C. ChIP-PCR was performed using anti c-myc antibody on cultures of C. glabrata untagged wild type (mock IP) and CgAP7-myc strains grownin standard YPD medium (CgAP7-myc) or 50 minutes after the addition of 1 mM DPTA-NONOate (CgAP7-myc+NO). The enrichment of theCgYHB1 promoter in the IP compared to the Input was measured by Q-RTPCR and normalized with the enrichment of the CgACT1 promoter(as endogenous control).D. C. glabrata wild type and cgap7Δ were grown in minimal medium buffered to pH 7.4. Cultures were treated or not with the indicated finalconcentrations of DPTA-NONOate and incubated for 6 hours. After incubation, dilutions were plated on YDP-agar plates. Colonies werecounted after 24 to 48 hours. The numbers of colonies obtained for each strain in the untreated condition (0 mM) were considered as 100% ofcell viability. Error bars represent the standard error of the mean.



S. cerevisiae and C. glabrata. To get further insights intothe evolutionary path that led to the present role of Yap7 inYHB1 expression, the roles of Cgap5 (CAGL0K08756g)and of the Yap5/7 orthologues in pre-WGD yeast specieswere investigated.

In S. cerevisiae, Yap5 is known to be responsible for theinduction of GRX4 (encoding a glutaredoxin) and CCC1(encoding an iron vacuolar transporter) in response to ironexcess (Li et al., 2008; Pimentel et al., 2012). To analyzethe potential role of Cgap5 in the regulation of the ironoverload response, the expression levels of CgGRX4 and

CgCCC1 were analyzed by Q-RTPCR in wild type andCgAP5 knock-out (cgap5Δ) strains, in normal growth con-ditions or in iron excess (Figure 7A and B). As previouslydescribed in S. cerevisiae (Li et al., 2008; Pimentel et al.,2012), these two genes were induced by iron excess inthe wild type C. glabrata strain. This induction was abol-ished by the deletion of CgAP5 (Figure 7A and B). Toestablish if Cgap5 was directly regulating these twogenes, a myc-tagged version of Cgap5 was cloned in anexpression plasmid under the control of its own promoter.This plasmid was used to transform cgap5Δcells. This

Fig. 7. Cgap5 is an iron stress regulator in C. glabrata. Cultures of C. glabrata wild type (CgWT), CgAP5 deleted (cgap5Δ) andcomplemented myc-tagged Cgap5 (CgAP5-myc) strains were grown in minimal synthetic medium (SC). Expression of CgGRX4 (A) andCgCCC1 (B) was measured by Q-RTPCR in the indicated strains after 20 min of treatment 5 mM FeSO4 (colored histograms) or after mocktreatment (white histograms). Q-RTPCR was performed using primers located in the coding sequence of CgGRX4, CgCCC1 and CgACT1 (asendogenous control). ChIP-PCR in CgGRX4 (C) and CgCCC1 (D) promoter regions was performed in C. glabrata wild type untagged strain(mock IP) and CgAP5-myc without iron treatment (CgAP5-myc) or after 20 min of iron treatment (CgAP5-myc+Fe). Q-RTPCR was alsoperformed using primers located in the promoters of CgACT1 gene as an endogenous control. Error bars represent the standard error of themean. The C. glabrata orthologue of ScCCC1 is CAGL0C00693g and is called here CgCCC1. ScGRX4 belongs to a pair of ohnologues whichalso contains ScGRX3. The orthologue pair in C. glabrata is composed of CAGL0G08151g and CAGL0L11990g. CAGL0L11990g showed noevidence of being regulated by Cgap5 in our preliminary experiments (data not shown). Hence the CAGL0G08151g is called CgGRX4 in thefigure and in the main text since our results strongly suggest that it is the functional homologue of ScGRX4.



C-terminal tagging of the transcription factor did not impairits activity, as demonstrated by measuring the responsesof CgGRX4 and CgCCC1 to iron excess in this cgap5-myc strain (Figure 7A and B). ChIP-PCR showed thatCgap5 binds to the promoters of these two genes inde-pendently of the presence of iron excess (Figure 7C andD), in accordance with what was previously described inthe case of ScYap5 (Li et al., 2008; Pimentel et al., 2012).

Finally, Cgap5 was able to restore a normal ScGRX4 ironresponse when transferred into a yap5Δ S. cerevisiaestrain (Figure 8C). In conclusion, the Yap5-dependent ironresponse which was previously described in S. cerevisiaeis conserved in C. glabrata.

Considering this strong conservation of the roles ofYap5 and Yap7 in the two post-WGD species that westudied, we aimed at determining whether these regula-

Fig. 8. Basal repression of YHB1 by Yap7 does not occur in pre-WGD yeast species. All the experiments performed to investigate YHB1regulation (right panel) were done in YPD medium. All the experiments performed to investigate GRX3/4 regulation (left panel) were done inminimal synthetic medium (SC). For iron treatment conditions, cells were collected 20 min after adding FeSO4 to a final concentration of 5 mM.A. Expression of KlGRX3/4 in K. lactis wild type and Δklyap5/7 strains with (green histograms) or without (white histograms) iron treatment.B. Expression of LkGRX3/4 in L. kluyveri wild type and Δlkyap5/7 strains with (green histograms) or without (white histograms) iron treatment.C. Expression of GRX4 in S. cerevisiae BY4742 strains in which a transgenic YAP5/7 gene was introduced at the ScYAP5 locus: ScYAP5(BY4742 wild type strain), Sc::CgAP5 (BY4742 containing the C. glabrata CgAP5), Sc::KLTH (BY4742 containing LtYAP5/7 fromL. thermotolerans), Sc::LAKL (BY4742 containing LkYAP5/7 from L. kluyveri), Sc::KLLA (BY4742 containing KlYAP5/7 from K. lactis). Strainscontaining YAP5/7 from pre-whole genome duplication species are included in the orange box.D. Expression of KlYHB1 in K. lactis wild type and Δklyap5/7 strains.E. Expression of LkYHB1 in L. kluyveri wild type and Δlkyap5/7 strains.



tions appeared after or before the whole genome duplica-tion (WGD). For this, we investigated the role of theYap5/7 orthologues in two pre-WGD species that werecalled ‘protoploid’ because they diverged from post-WGDspecies prior to the whole genome duplication: Lachan-cea kluyveri (Yap5/7 orthologue: SAKL0F05434g, calledLkYAP5/7 thereafter) and Kluyveromyces lactis (yap5/7orthologue: KLLA0D14399g, called KlYAP5/7 thereafter)(Figure 5; Genolevures et al., 2009). We examined theiron overload response and the YHB1 basal expressionlevel in strains deleted for KlYAP5/7 or LkYAP5/7, ascompared to their isogenic wild types. In the K. lactis andL. kluyveri wild type strains, transcript levels of KlGRX3/4and LkGRX3/4 increased in response to iron excess whilethe deletion of KlYAP5/7 or LkYAP5/7 impaired the iron-dependent induction of these two genes (Figure 8A andB). Hence, KlYap5/7 and LkYap5/7 have a role in theresponse to iron excess, which is very similar to Cgap5and ScYap5. Considering this conservation, we testedwhether the molecular mechanisms governing the regu-lation of GRX4 were conserved between pre- and post-WGD species. We replaced YAP5 in S. cerevisiae by theYAP5/7 genes coming from L. Kluyverii or K. lactis and weexamined the response of ScGRX4 to iron excess in thewild type and in the transgenic strains (Figure 8C). NeitherLkYap5/7, nor KlYap5/7 proteins were able to complementthe absence of Yap5 in S. cerevisiae suggesting that,although the role of Yap5 is conserved between pre- andpost-WGD species, the molecular mechanisms by which itoperates have diverged.

The deletion of KlYAP5/7 or LkYAP5/7 had no effect onthe basal expression of klYHB1 and LkYHB1 (Figure 8Dand E), suggesting that the regulation of YHB1 by Yap7 isspecific to post-WGD species. We also verified that thedeletion of YAP5/7 in these species had no effect on theinduction of YHB1 by NO treatment (supplementaryFigure S6).

In conclusion of this part, in the pre-WGD yeast speciesK. lactis and L. kluyverii, the Yap5/7 proteins have a rolein the response to iron excess, which is conserved in theYap5 proteins found in the post-WGD species S. cerevi-siae and C. glabrata. In contrast, Yap5/7 proteins do notcontribute to the basal repression of YHB1, suggestingthat this regulation appeared after the WGD, by neofunc-tionalization of the duplicate paralogue which gave rise tothe Yap7 lineage.

Discussion

Yap7: a new regulator of nitric oxide oxidase expressionin C. glabrata and S. cerevisiae

The central finding of this work is the characterization of thebZip transcription factor Yap7 as being a strong repressor

of the YHB1 basal level of expression in the human patho-gen C. glabrata and in the model yeast S. cerevisiae,which are two post-Whole Genome Duplication yeastspecies. This regulation is very specific since transcrip-tome analyses conducted in S. cerevisiae and C. glabrataindicated that YHB1 is the only gene to be subjected to thisstrong repression in the conditions studied. This dramaticregulation had been missed in previous studies of YHB1expression because it was lost in the laboratory strains ofS. cerevisiae. This loss of function can be due either to theframe-shift in the YAP7 gene or to some of the 5 aminoacids which differ between the N-terminal sequences of theshort and long forms (which are 97% identical). The trun-cated Yap7 protein has a DNA binding domain and DNAbinding properties which are identical to the long form, butwas unable to repress YHB1 expression. The frame-shifted YAP7 was found only in the wine/European lineageof S. cerevisiae and in mosaic strains arising from thislineage. This suggests that this mutational event is rela-tively recent in the history of the S. cerevisiae species (Litiet al., 2009; Bergstrom et al., 2014). An important corollaryof this finding was that all previous experiments on YHB1regulation conducted in S. cerevisiae had been done in aYap7 loss of function context (Sarver and DeRisi, 2005;Zhu et al., 2006; Hickman and Winston, 2007). This led usto question the published models of NO response in S. cer-evisiae and the actual role of Fzf1 as being the majorregulator of this response. In bacteria, NsrR is a globalrepressor which shuts down FlavoHb expression in theabsence of NO. This protein contains a 2Fe-2S clusterwhich forms a cysteine-dinitrosyl-iron complex (DISC) inpresence of NO. The NsrR-DISC complex has a reducedaffinity for the FlavoHb promoter, which relieves the tran-scriptional repression of the FlavoHb gene (Bodenmillerand Spiro, 2006; Tucker et al., 2008; 2010). Interestingly,the full-length Yap7 protein contains conserved cysteinerich domains (CRDs), which are a common feature of theYap transcription factors involved in redox-homeostasis(Rodrigues-Pousada et al., 2010). In the cases of Yap1,Yap2, Yap5 and Yap8, which sense oxidative stress,cadmium, iron and arsenic respectively, the CRDs havebeen shown to encounter redox modifications which dras-tically change the regulatory properties of the proteins atthe level of DNA binding or nucleo-cytoplasmic shuttling(Menezes et al., submitted manuscript; Delaunay et al.,2000; Wood et al., 2004; Azevedo et al., 2007; Rietzschelet al., 2015). In Aspergillus and Fusarium species, theHapX proteins also contain CRDs which are important fortheir role in iron homeostasis (Gsaller et al., 2014). Hence,a seducing hypothesis was that Yap7 could be functionallyanalogous to the bacterial NsrR system and that its CRDscould play a role similar to the NsrR iron-sulfur clusters inNO sensing. However, our analyses of the NO response inthe S. cerevisiae North-American genetic background did



not support this hypothesis. The additive effects of theYHB1 derepression caused by deletion of YAP7 and of itsactivation by Fzf1 following NO treatment strongly sug-gests that these two pathways act independently fromeach other (compare Δyap7, Δyap7+NO, WT+NO andΔyap7Δfzf1+NO on Figure 2A). Considering this, then if NOwould severely impair Yap7 repression, we should haveseen an increase of YHB1 expression following NO expo-sure in the Δfzf1 strain, which should be equivalent to theYHB1 increase observed in the Δyap7 strain in absence ofNO. Actually, this is absolutely not the case (Figure 2A). Inaccordance with this model, we observed only a verymodest decrease of Yap7 promoter occupancy followingNO treatment, both in S. cerevisiae and C. glabrata(Figures 2B and 6C). In conclusion, our analyses of YHB1expression in different genetic background following NOtreatment unambiguously confirmed the major role of Fzf1in the NO response of YHB1 and suggest that the Yap7activity and DNA binding properties are poorly sensitive tothe presence of NO.

YHB1 repression requires the general co-repressorTup1 in S. cerevisiae and in C. glabrata

Although the Yap7 protein sequence is poorly conservedbetween S. cerevisiae and C. glabrata (less than 25%identity and 40% similarity), Cgap7 was able to bind andrepress YHB1 in S. cerevisiae, suggesting that at leastpart of its mechanisms of action are conserved betweenthese two species. Importantly, the bZIP DNA bindingdomain of Yap7 and Cgap7 are almost identical, suggest-ing that they recognize very similar Yap ResponseElements (YREs). Also, our co-immunoprecipitationexperiments and our functional analyses indicated thatYap7 could repress YHB1 through the recruitment of thegeneral co-repressor Tup1. According to our CgTUP1deletion experiments, this role is conserved in C. glabrata.Notably, the Tup1 sequence is very similar betweenS. cerevisiae and C. glabrata (60% identity, 75% similar-ity). This Tup1 sequence conservation, together with thehigh conservation of the Yap7 bZIP motif mentionedabove, may explain that Cgap7 was able to efficiently bindYHB1 and repress its expression in a Δyap7 mutant ofS. cerevisiae (Figure 1E and F). Tup1 is the yeast homo-logue of the Groucho transcriptional repressor in highereukaryotes (Courey and Jia, 2001). It functions incomplex with the Ssn6 protein. Tup1/Ssn6 does notdirectly contact DNA but is rather recruited at its targetpromoters by the interaction of Tup1 with various specificDNA binding transcription factors (Malave and Dent,2006). Upon its recruitment, the Tup1/Ssn6 acts on thechromatin state to repress the transcription of its targetgenes (Malave and Dent, 2006; Chen et al., 2013). Thelist of the Tup1 co-factors identified in S. cerevisiae

includes more than ten proteins (Hanlon et al., 2011).Thus, Tup1 is involved in the control of many importantprocesses, including catabolic repression, flocculation,DNA damage response, hypoxia, etc. We showed that, inthe presence of the full-length Yap7 protein, Tup1 wasrequired for the repression of YHB1 in standard growthconditions and that this effect was not increased in adouble tup1Δyap7Δ mutant (Figure 3C). Moreover, Tup1binding to YHB1 promoter requires the full-length Yap7since this binding was abolished when only the short formwas present (Figure 3D). Together with the co-IP results,these findings strongly suggest that Yap7 and Tup1 col-laborate to repress YHB1 expression. The interactionbetween Yap7 and Tup1 apparently requires the 250amino acids in C-terminus of Yap7, since the short form ofYap7 was not able to interact with Tup1 and to recruit it tothe YHB1 promoter. Our work identified Yap7 as a newTup1 co-factor, the Tup1-Yap7 interaction being appar-ently mostly dedicated to the repression of YHB1, in theconditions that we studied.

FlavoHb expression is tightly controlled in yeasts

Yap7 may not be the only Tup1 co-factor to repress YHB1.In anaerobic conditions, YHB1 is strongly repressed byHap1, a heme dependent transcription factor, the activity ofwhich is modulated by oxygen availability in S. cerevisiae(Hickman and Winston, 2007). Hap1 activates manyaerobic genes in the presence of oxygen (Becerra et al.,2002; Ter Linde and Steensma, 2002) and repressesseveral genes, including YHB1, in hypoxia (Hickman andWinston, 2007). Interestingly, Hap1 also requires the Tup1/Ssn6 complex to switch off its target genes (Hickman andWinston, 2007). However, in aerobic conditions, Hap1does not repress the expression of YHB1, either in theS288C background (Hickman and Winston, 2007), or in theNorth-American strain (Supplementary Data S7). Hence,in S. cerevisiae, at least two Tup1 cofactors, Yap7 andHap1, are devoted to the repression of YHB1, dependingon the oxygen availability. Remarkably, the basal level ofYHB1 expression was low in all yeast species that weexamined, even those in which Yap7 orthologues do notcontrol YHB1 expression in standard growth conditions (K.lactis, L. kluyverii or C. albicans), (Supplementary DataS8). This suggests that, in pre-WGD yeast species, YHB1is also constitutively repressed, though by a differentmechanism. Because, in yeasts, the NO induction of YHB1is due to transcriptional activators, the regulation of YHB1in fungi was thought to be mostly positive, in contrast towhat was found in bacteria. According to our work, itseems that, from bacteria to yeasts, microorganisms haveevolved complex regulatory systems to efficiently repressnitric oxide oxidase expression when it is not absolutelyrequired.



Contribution of the CCAAT binding complex to YHB1repression in S. cerevisiae: a relic from ancient times

The CCAAT binding complex (CBC) recognizes theCCAAT motif, which is present in the promoters of morethan 30% of eukaryotic genes (Bucher, 1990). The Hap2/3/5 subunits of CBC are conserved from yeasts to plantsand mammals, in which CBC is called NY-Y (nuclear factorY) (Mantovani, 1999). In mammals and plants, these threesubunits are sufficient for both DNA binding and transcrip-tional regulation (Mantovani, 1999). In fungi, however,CBC works as a DNA binding platform which sometimesrequires a fourth regulatory subunit for regulatinggene expression. The first CBC regulatory subunit tobe described was Hap4 in S. cerevisiae (Forsburg andGuarente, 1989). The interaction between Hap4 and CBCmediates the transcriptional activation of genes encodingrespiratory complex subunits. However, Hap4 is poorlyconserved and clear orthologues of Hap4 have been foundonly in few yeast species (Bourgarel et al., 1999; Sybirnaet al., 2005). Instead, fungal genomes encode severalproteins which only share with Hap4 the 20 amino acidmotif which is necessary for its interaction with the othersubunits of CBC [called hap4-like interaction domain(HAP4L)]. For instance, proteins containing a HAP4L-bZIPbipartite domain have been found in 46 fungal genomes,encompassing a phylogenetic distance which, in terms ofgenome sequence divergence, is much larger than thewhole vertebrate tree (Sybirna et al., 2010; Singh et al.,2011). From C. neoformans to C. albicans, the HAP4L-bZIP proteins (named Hap43 in C. albicans and HapX inthe other species) have a remarkably conserved role incontrolling the response to iron starvation (Hortschanskyet al., 2007; Baek et al., 2008; Jung et al., 2010; Hsu et al.,2011; Singh et al., 2011; Lopez-Berges et al., 2012) andwere more recently shown to act in the high iron responseas well (Gsaller et al., 2014). The interaction between theCBC and HapX proteins lead to either strong transcrip-tional repression or moderate transcriptional activation,depending on the genes considered or on the growthconditions used (Hortschansky et al., 2007; Baek et al.,2008; Jung et al., 2010; Hsu et al., 2011; Singh et al., 2011;Lopez-Berges et al., 2012; Gsaller et al., 2014). The CBCand the HAP4L domain are absolutely required for the fullrepression of HapX and Hap43 targets in low iron condi-tions (Hortschansky et al., 2007; Baek et al., 2008; Junget al., 2010; Hsu et al., 2011; Singh et al., 2011;Lopez-Berges et al., 2012) and for the HapX mediatedactivation of iron vacuolar import genes in high iron condi-tions (Gsaller et al., 2014). The role of the bZIP motif is lessclear. In A. nidulans, HapX is dependent on CBC for DNAbinding in vitro (Hortschansky et al., 2007). In C. albicans,the HAP4L domain of Hap43 is absolutely required for itsfunction and the DNAmotif enriched in the promoters of the

Hap43 targets is the CCAAT sequence, not the YRE (Chenet al., 2011; Singh et al., 2011). These data suggested that,in these species, HapX proteins bind DNA exclusivelythrough their interaction with the CBC, and not directlythrough their bZIP domain. However, this model has beenchallenged by the observation that the bZIP motif of Hap43is also essential for its function (Singh et al., 2011) and byrecent results which showed that, in vitro, HapX interactswith the promoter of some of its targets, through an evolu-tionary conserved DNA sub-motif located in 3’ of theCCAAT motif (Gsaller et al., 2014; Hortschansky et al.,2015).

The Yap5/7 proteins are the orthologues of the HapXproteins in the Saccharomycetacea yeast species(Figures 5 and 9). The HAP4L-bZIP domain is nicely con-served in the pre-WGD species (Figures 4A and 5). Inpost-WGD species, the HAP4L motif has degenerated inthe Yap5 lineage and it disappeared in the Yap7 lineage,with the notable exception of the Saccharomyces sensustricto species, in which it was conserved (Figures 4A and5). The detection of Hap5 in the ScYap7 co-IP experi-ments and the derepression of YHB1 in the S. cerevisiaeHAP5 mutants indicate that ScYap7 may interact withCBC to repress YHB1, using a mechanism similar toHapX proteins. In accordance with this model, Hap4 didnot participate to the regulation of YHB1 expression(Figures 4B). Both the short and the full-length forms ofYap7 were able to co-immunoprecipitate with Hap5, whichcan be connected to the presence of the HAP4L-bZIPdomain in the two proteins. The fact that the short Yap7form is not able to repress YHB1 tells us that this interac-tion is not enough for transcriptional repression and thatthe C-terminal part of Yap7 is necessary for the Yap7-CBCcomplex to be fully functional. Notably, the loss of repres-sion observed in the hap5Δ mutant was less than 20% ofthe one measured for the yap7Δ and tup1Δ strains. Thisclearly indicated that Yap7 can bind to the YHB1 promoterand recruit Tup1 independently of CBC. This marginal roleof CBC in YHB1 repression has been lost in C. glabrataand, probably, in most post-WGD species. In thesespecies, the HAP4L motif is absent and it is very likely thattheir Yap7 proteins do not interact with CBC anymore(Figure 9). These results are in accordance with the factthat the DNA binding motif bound by Yap7 in S. cerevisiaeand, very likely, in C. glabrata is the YRE, and not theCCAAT motif ( (Tan et al., 2008), Merhej et al., unpub-lished results), which indicates that, in contrast with theHapX proteins, Yap7 and Cgap7 can bind DNA throughtheir bZIP motif, independently of the CBC. The ortho-logues of Yap5/7 in pre-WGD species do not repressYHB1 expression in standard growth conditions. This wasobserved in K. lactis, L. kluyverii (this work) and C. albi-cans (Hsu et al., 2011). These results suggest that thedirect and constitutive regulation exerted by Yap7 proteins



on YHB1 seems to be a specificity of the S. cerevisiaeand C. glabrata species. This led us to the conclusion thatthe role of Yap7 in C. glabrata and S. cerevisiae is aninnovation of post-WGD species, which may haveappeared by neofunctionalization of one Yap5/7 duplicateafter the whole genome duplication (Figure 9).

A model for YHB1 basal repression in S. cerevisiae

Putting together the data discussed above, our model ofYHB1 basal repression in S. cerevisiae is the following.The long form of Yap7 represses the expression of

YHB1 (Figure 1C, E and F) by binding to its promoter(Figure 1F) and recruiting the general repressor Tup1(Figure 3). Hap5 also moderately contributes to thisrepression and interacts with the long form of Yap7(Figure 4B and supplementary data S3). The fact thatthe effect of the deletion of the long form of YAP7 onYHB1 expression is exactly the same as the deletion ofTUP1 and not an addition of the tup1Δ and hap5Δeffects (Figures 3B and 4B) indicates that Tup1 is nec-essary, but not totally sufficient, for the full repression ofYHB1 by Yap7. The short form of Yap7, which is foundin the west European clade and in all laboratory strains,

Fig. 9. Evolution of the HapX-Yap5/7 family members in the fungus kingdom. The role and activity of the Yap7 relatives, the presence orabsence of the HAP4L domain in the protein, the contribution of the CCAAT-binding complex to the activity of the protein and the DNA bindingmotif of each protein are represented. Note that the tree on the left is a schematic representation of the protein tree of Figure 5. The distancesbetween the tree branches do not represent the real phylogenetic distances between the proteins. The phylogenetic groups of the indicatedspecies are indicated by the vertical bars on the right of the figure. The whole genome duplication event is indicated with a red dot on thetree. The letters within brackets indicate the publication from which the information comes: [a] (Li et al., 2008), [b] (Pimentel et al., 2012), [c](Hsu et al., 2011), [d] (Jung et al., 2010), [e] (Schrettl et al., 2010), [f] (Lopez-Berges et al., 2012), [g] (Hortschansky et al., 2007), [h] (Chenet al., 2011), [i] (Singh et al., 2011), [j] (Sybirna et al., 2005), [k] (Kuo et al., 2010), [l] (Tan et al., 2008) [m] (Gsaller et al., 2014).



still contains a HAP4L-bZIP domain which is 100% iden-tical to the long form. Hence, it is able to bind the YHB1promoter (Figure 1F) and to interact with Hap5 (supple-mentary data S3). However, it lost its ability to interactwith Tup1 (Figure 3A, B and D). Because the interactionwith Tup1 is necessary for the repression of YHB1 byYap7 (Figure 3C), this short form is not able to repressYHB1 anymore (Figure 1D and E).

Cgap5 plays a conserved role in the response toiron excess

In the frame of this work, we also analyzed the function ofCgap5, the ohnologue of Cgap7, in Candida glabrata. Weshowed that, like in S. cerevisiae, Cgap5 is required for theresponse to iron overload. More precisely, Cgap5 is atranscriptional activator of the C. glabrata homologues ofGRX4 and CCC1 in the presence of iron excess. As forYap7 and Cgap7, the mechanisms of action of Cgap5 andYap5 seemed to be nicely conserved, since Cgap5 wasable to replace Yap5 in S. cerevisiae. Our experiments inK. lactis and L. kluyverii showed that this role is evenconserved in pre-WGD yeast species. However, theKlYap5/7 and LkYap5/7 proteins could not replace Yap5,indicating that their mechanisms of action are different.This connection between the Yap5/7 proteins and ironhomeostasis is ancient but has significantly divergedthrough evolution (Figure 9). As mentioned above, thehomologues of Yap5/7 in Aspergillus nidulans, Aspergillusfumigatus, Fusarium oxysporum, Cryptococcus neofor-mans and Candida albicans are major regulators of theresponse to iron starvation and principally act by repress-ing transcription (Figure 9) (Jung et al., 2010; Schrettlet al., 2010; Hsu et al., 2011; Singh et al., 2011;Lopez-Berges et al., 2012). Additionally, they play a role inresponse to iron excess by activating the expression ofgenes involved in the vacuolar transport of iron (Gsalleret al., 2014). From K. lactis to S. cerevisiae, the responseto iron starvation is under the control of the Aft1/2 zincfinger transcriptional activators (Courel et al., 2005; Condee Silva et al., 2009; Goncalves et al., 2014). In thesespecies, the Yap5 proteins are transcriptional activatorswhich control the induction of a handful of genes inresponse to iron excess (Figure 9) (Li et al., 2008;Pimentel et al., 2012), this work). Remarkably, the proteinmotifs which allow Yap5 in S. cerevisiae and the HapXproteins in filamentous fungi sensing iron excess are verysimilar, indicating that this role probably has the sameevolutionary origin and is very ancient (Gsaller et al., 2014;Rietzschel et al., 2015).

In conclusion, this work described and dissected a newaspect of the regulation of nitric oxide oxidase in yeasts.Additionally, it described Cgap5 and Cgap7, as the firsttranscription factors identified in the human pathogen

C. glabrata to be involved in the control of iron homeosta-sis and nitric oxide oxidase expression, respectively.

Experimental procedures

Culture conditions

Plasmids were amplified in Escherichia coli strain DH5α in LBmedium with 20 μg/ml of ampicillin. For yeast transformantsselection and screening, cells were plated on selectivemedium containing 2% glucose, 0.67% Yeast nitrogen base(DIFCOTM), 2% agar and 0.074% of CSM mixture (DIFCOTM)lacking uridine, tryptophane or histidine when using URA3,TRP1 or HIS3 markers or on YDP-agar containing 200 μg/mlof Geneticin® (Life Technologies) when using KanMX6marker.

For yeast transformation, transcriptome studies, chromatinimmunoprecipitation, RNA expression for evolutionary studyand co-immunoprecipitation, cells were grown in YPD liquidmedium (DIFCOTM) containing 1% yeast extract, 2% peptoneand 2% Dextrose.

For iron overload analysis, cultures were grown in syn-thetic medium (SC) (0.67% Yeast nitrogen base containingammonium sulfate and 2% glucose) supplemented withappropriate selective amino acids. Cells were treated withFeSO4 to a final concentration of 5 mM.

For liquid growth in nitrosative stress conditions, cultureswere carried out in synthetic medium buffered to pH 7.2 with3% succinic acid and 0.28% of KOH and supplemented withappropriate selective amino acids. To induce nitrosativestress, Dipropylenetriamine NONOate (DPTA NONOate;Cayman Chemicals) was resuspended to 0.5 M in 10 mMNaOH immediately before use (alkaline inactive solution) andadded to BSC cultures at a final concentration of 1 mM. Onemolecule of DPTA NONOate spontaneously releases twomolecules of NO per amine molecule in a pH-dependentmanner (pH 7.0–7.4).

For C. glabrata and S. cerevisiae viability assays, cultureswere performed in BSC medium. At A600 = 0.1, cells weretreated with increasing concentrations of DPTA-NONOate.After 6 hours of incubation at 30°C with shaking, 10−4 and 10−5

dilutions of each culture were plated on YPD agar medium.Colonies were counted after 24h to 48h.

All liquid cultures were performed at 30°C with shaking at160 rpm. Additional experiment-specific growth conditionsare given in the respective paragraphs.

Sequences sources, multiple sequence alignment andphylogenetic tree

DNA and protein sequences of YAP5 (YIR018w) andYAP7 (YOL028C) of the wild type strain S288c (BY4742)were downloaded from Saccharomyces Genome Databasewebsite (http://www.yeastgenome.org/) and those ofCgAP5 (CAGL0K08756g) and CgAP7 (CAGL0F01265g)from Candida Genome Database website (http://www.candidagenome.org). The homologues of YAP7 and YAP5were identified by position-specific iterated BLAST PSI-BLAST® in NCBI database (http://www.ncbi.nlm.nih.gov/).Their sequences were downloaded from Génolevures website



http://www.yeastgenome.org/

http://www.candidagenome.org

http://www.candidagenome.org

http://www.ncbi.nlm.nih.gov/

(http://www.genolevures.org). The genomic sequences ofYAP7 from 38 S. cerevisiae strains were collected from Sac-charomyces genome resequencing project (SGRP) website(http://www.moseslab.csb.utoronto.ca/sgrp/). The sequenceswere compared to identify SNP and frame-shift mutations.

The multiple sequence alignment of YAP5 and YAP7 homo-logues was conducted using the ‘Molecular EvolutionaryGenetics Analysis’ (MEGA6) software (Tamura et al., 2013).Based on this alignment, a neighbor-joining phylogenetic treewas built using the bootstrap method with 1000 replications.The substitutions method was the Poisson Model with uniformrate among sites and complete deletion of gaps.

Strains and plasmids construction

The yeast strains used throughout this study can be foundin Supplementary Data S9. The parental strains were:C. glabrata ΔHTU, which was kindly provided by Cécile Fair-head, S. cerevisiae BY4742 (Euroscarf collection) andYLF131 [derived from the North-American wild type strainYPS128 and suitable for genetic manipulation (Louvel et al.,2014)] and L. kluuveri LAKL011 derived from CBS3082,which were offered by G. Fisher and K. lactis MWL9S1, whichwas a gift of Marc Lemaire. A complete list of the primersused for yeast mutants construction, plasmid construction,mutant validation and qPCR are listed in Supplementary DataS10. The plasmids used or generated during this study aredescribed in Supplementary Data S11.

C. glabrata. CgAP5 and CgAP7 gene replacement cassetteswere amplified from pFA6a-TRP1 plasmid (Longtine et al.,1998) using pair of primers containing in 5’, sequences iden-tical to the 80 bases upstream the target gene (Forwardprimer) and 80 bases downstream the target gene (reverseprimer). The resulting DNA cassette containing the TRP1 orHIS3 markers with the target gene flanking regions wereused for subsequent transformation (Schiestl and Gietz,1989) of C. glabrata ΔHTU. Correct integration of the cas-sette and the absence of the deleted gene were verified byPCR and southern blot.

The plasmid pCU-PDC1 contains a chimeric C. glabrataCEN/ARS sequence, an URA3 cassette and a multi cloningsite downstream the constitutive PDC1 promoter (Zordanet al., 2013). The plasmid pGRB2.1 was generated by digest-ing pCU-PDC1 with BamHI to eliminate a 1.4 kb fragmentcontaining the PDC1 promoter. A 2.1 kb fragment containingthe 13xMyc-HIS5 cassette was amplified using primers 435and 437 (containing several restriction sites including XhoI),digested with XhoI restriction enzyme and cloned into theunique XhoI site of pCU-PDC1 and pGRB2.1 to generatepYR29-Myc-HIS and pGRB2.1-Myc-His, respectively.

To generate expression plasmids containing a myc-taggedversion of CgAP5 and CgAP7, PCR-amplified DNA fragmentscontaining the coding sequences of CgAP5 and CgAP7together with 1 kb of their respective promoters were digestedwith SmaI and NotI and cloned into the pGB2.1-Myc-Hisplasmid upstream the myc epitope sequence yielding inpGRB2.1-CgAP5-Myc-His and pGRB2.1-CgAP7-Myc-Hisplasmids, respectively. The resulting plasmids were used totransform cgap5Δ and cgap7Δ strains thus generating thestrains CgAP5-myc and CgAP7-myc, respectively. The RNA

expression and protein translation of the CgAP5 and CgAP7myc-tagged proteins were verified by qPCR and Western blot.

S. cerevisiae. The same procedure used for gene deletion inC. glabrata was also applied in S. cerevisiae with minor modi-fications. The homology regions flanking the selection markergene were 50 bp long. The marker genes were KanMX orHIS3 amplified from pFA6a-kanMX6 and pFA6a-HIS3MX6(Longtine et al., 1998) or URA3 amplified from pRS416 plas-mids, respectively.

For heterologous expression of YAP5 homologues in S. cer-evisiae, YAP5 sequences from S. cerevisiae BY4742strain, C. glabrata CgAP5 and YAP5/7 homologues fromK. lactis MWL9S1 and L. kluyveri CBS3082 were clonedinto pGRB2.1-Myc-His or pGRB2.1-Myc-His plasmids asdescribed in Supplementary Data S11. The resulting plasmidswere used as DNA matrix to amplify gene replacement cas-settes containing YAP5 coding sequences upstream the myctag and HIS3 marker gene flanked by 50 bp corresponding tothe 5’ and 3’ sequences bordering the Yap5 locus. Thesecassettes were transformed into S. cerevisiae BY4742 toreplace the wild-type YAP5 by homologous recombination.The correct expression and translation of the transgenic Yap7proteins was validated by Western blot using anti-myc anti-body (Roche).

To generate proteinA (ProtA)-tagged version of long andshort YAP7, YAP7 from BY4742 and YAP7 from YPS128were amplified and cloned into pGRB2.1-ProtA-His (Supple-mentary Data S11). The resulting plasmids were used toamplify DNA fragments containing YAP7 coding sequencesfollowed by ProtA tag and HIS3 marker gene. These cas-settes were used to transform S. cerevisiae BY4742. Thecorrect expression of the ProtA-tagged version of YAP7factors in the resulting strains was verified by Western blotusing Anti-IgG Antibody.

K. lactis and L. kluyveri. For deletion of YAP5/7 in K. lactisMWL9S1 and L. kluyveri LAKL011, deletion cassettes carry-ing the URA3 marker gene flanked by 500 bp DNA sequenceshomologue to the sequences located at the 5’ and 3’ ends ofeach target gene were constructed by using double-joint PCRstrategy as previously described (Yu et al., 2004). The primersused for the construction of the deletion cassettes are listed inSupplementary Data S10. The transformation and correctintegration of the cassettes were performed as describedabove for C. glabrata.

RNA extraction, cDNA synthesis and qPCR

Cells were grown overnight to an A600 0.6–0.8 in the appropri-ate medium, exposed or not to the indicated concentration ofdrug (Selenite, FeSO4 or DPTA-NONOate), snap-frozen incold ethanol and collected by centrifugation. Cells lysiswas mechanically performed with glass beads using aFastprep®-24 bead beater (MP Biomedicals). Total RNAextraction was carried out using the RNeasy extraction kit(Qiagen) following the manufacturer’s instructions.Absence ofRNA degradation was verified on agarose gel and the concen-tration of each sample was determined using NanoPhotom-eter® spectrometer (IMPLEN). For each sample, 1 μg of thetotal RNA were DNAse treated using Turbo DNA-free kit



http://www.genolevures.org

http://www.moseslab.csb.utoronto.ca/sgrp/

(Ambion). After DNAse treatment, 0.2 μg of total RNA wereused to perform cDNA synthesis using Superscript II ReverseTranscriptase according to the manufacturer’s instructions(Invitrogen). The resulting cDNA were purified using QIAquickPCR purification kit (Qiagen) and diluted to three differentconcentrations (1:10, 1:20 and 1:40). Quantitative PCR reac-tions were performed on a C1000TM Thermalcycler (Bio-rad)with a 2X SYBR Green master mix (Promega). The qPCRreaction mixture contained 0.5 μM of each primer and 4 μL ofone of the three dilutions of the cDNA. These dilutions servedi) to construct the relative standard curve for each and ii) astriplicate for each sample. The primers used for qPCR arelisted in Supplementary Data S10. The relative expression fora given gene was generated by calculating the difference inthe abundance between the transcripts of this gene comparedto the transcripts of the ACT1 gene, used as an endogeneousreference, based on the ΔCt method following the formula‘Efficiency (Ct target – Ct Act1)’. Finally, the expression values werenormalized to the expression of CgYHB1 in CgWT strain giventhe arbitrary value 1 (Figure 6B).

Microarray experiments and data analysis

Strains were grown overnight to A600 of 0.6–0.8 in YPD. Cellswere collected and total RNA was extracted, quality controlledand quantified as described above. During extraction, TotalRNA was cleaned up from DNA contamination using DNAse I(Qiagen) according to the manufacturer’s instructions. One μgof total RNA was used for fluorescent cDNA synthesis accord-ing to the amino-allyl protocol. The cDNA were labeled withCy3 and Cy5 and hybridization was performed according tothe protocol described at transcriptome.ens.fr/ sgdb/protocols/. A minimum of two biologically independent experi-ments were performed for each condition, using dye switch. Aset of 1 to15 specific probes per gene matching the ORFs ofS. cerevisiae or C. glabrata, was designed using the Teolensoftware, with default parameters (Jourdren et al., 2010). Thisprobe design was deposited on the Array express database(A-MEXP-2365 for S. cerevisiae and A-MEXP-2402 forC. glabrata) and was used to produce Agilent arrays in an 8 x60k format (Agilent Technologies). After overnight hybridiza-tion and washing, the slides were scanned using a 2-micronAgilent microarray scanner. The images were analyzed usingthe feature extraction software and normalized using Goul-phar software (Lemoine et al., 2006). For each gene, theCy5/Cy3 ratios corresponding to the different probes wereaveraged. The mean of the biological replicates was calcu-lated.Agene was considered as being differentially expressedif its absolute Log2(fold change) value was more than 1 and ifits expression variation was considered as being statisticallysignificant using the TMEV version of SAM with a FDR of 5%,a S0 calculated by the Tusher method and using the exactnumber of permutation. The complete microarray data areavailable at Array express database under the accessionnumber E-MTAB-2653.

Chromatin-immunoprecipitation followed by qPCR(ChIP-qPCR)

Overnight cultures of tagged and wild type strains weregrown overnight to A600 of 0.8–1. 50 ml of each culture wasfixed by adding 1% formaldehyde for 15 min at room tem-

perature with occasional agitation. Cross-linking wasstopped by adding glycine to a final concentration of340 mM and incubating for 5 min at room temperature.Cells were disrupted in lysis buffer [50 mM HEPES-KOH pH7.5, 140 mM NaCl, 1 mM, EDTA, 1% Triton X-100 and 0.1%Na-deoxycholate, 1 mM phenylmethylsulfonyl fluoride andprotease inhibitor cocktail (Roche)], DNA was sheared usinga Bioruptor® standard sonication device (Diagenode) andimmunoprecipitation (IP) was performed using a mousemonoclonal anti-myc antibody (Roche) coupled to magneticbeads (pan-mouse IgG Dynabeads; Dynal Biotech), asdescribed in (Merhej et al., 2014). About 1% of total solublefraction before IP was retained for subsequent DNA extrac-tion (INPUT). The correct immunoprecipitation of the targetprotein was systematically controlled by Western blot usinganti-myc antibody (Roche). The cross-linking reversal of theIP complexes and of the INPUT was then processed over-night at 65°C as previously described (Merhej et al., 2014).The samples were treated with proteinase K (Roche) andRNAse A (ThermoFisher - France) to eliminate any residualproteins or RNA. Total (IN) or immunoprecipitated (IP) DNAwere extracted using a standard phenol/Chloroform proto-col, purified using QIAquick PCR purification kit (Qiagen)and eluted in 50 μl H2O. The ChIP targeting Tup1 was per-formed exactly as described except that we used rabbitanti-Tup1 antibodies (Gift from Joseph Reese) and goatanti-rabbit IgG Dynabeads from Dynal Biotech.

Quantification of specific promoter regions was performedby qPCR. The primers used for qPCR are listed in Supple-mentary Data S10. Three serial dilutions (1:10, 1:20, 1:40) ofIP samples were simultaneously processed together with INsamples used for normalization. Standard curves, used toassess the PCR efficiency for each couple of primers, weregenerated using the three serial dilutions. The enrichment ofthe ACT1 promoter from the corresponding species was usedas endogenous control. The relative enrichment of a specificlocus in the immunoprecipitated DNA was determined usingthe ΔΔCt method using the formula Efficiency (Ct IPtarget – Ct IPAct1)

- Efficiency (Ct INtarget – Ct INAct1).

Co-immunoprecipitation, mass spectrometry andWestern-blot analysis

The co-immunoprecipitation experiments were performedfollowing the protocol described in (Bretes et al., 2014)using cells collected from cultures of S. cerevisiae wild type,YAP7-sh-ProtA and YAP7-ln-ProtA strains in YPD liquidmedium. Briefly, the pellet were frozen in liquid nitrogen andthen disrupted in cryogenic conditions. Two grams of theresulting lysate were thawed slightly on ice in a semi-stringent extraction buffer (20 mM Hepes pH 7.4, 110 mMKOAc, 2 mM MgCl2, 0.5% Triton X-100, 0.1% Tween-20,1 mM DTT, 4 μg/ml pepstatin A, 0.18 mg/ml PMSF, 1/5000Antifoam B). The lysate was dissolved in the extractionbuffer using a Polytron mixer and the solution was filteredusing 1.2 μm filters. The resulting filtered solution was incu-bated with Dynabeads® (Dynal Biotech.) conjugated toRabbit IgG antibadies for 30 min at 4°C on a rotating wheel.The beads were then washed once with the extration buffer,once with a last wash buffer (0.1 M NH4Ac, 0.02% Tween-20, 0.1 mM MgCl2) and 4 times with the same last wash



buffer without Tween-20. The immunoprecipitated proteinswere eluted with the elution buffer (100 mM NH4OH,0.5 mM EDTA) and lyophilized overnight. The proteins wereresuspended in 25 mM Ammonium Carbonate and 3⁄4 of thevolume was used for mass spectrometry analysis while theremaining volume was used for Western blot analysis. Thetryptic digestion of the eluted proteins, the chromatographicseparation of the resulting peptides, the fragmentation andidentification by mass spectrometry and the data analysiswere performed as previously described (Bretes et al.,2014). For Western-blot analysis, fractions from IP andinput samples were separated on 10% SDS-Polyacrylamidegel electrophoresis (SDS-PAGE). Proteins were then trans-ferred to Whatman® Protan® BA83 nitrocellulose membrane(GE Healthcare). Immunoblotting were performed using1:30000 rabbit IgG Anti-DED1, 1:5000 rabbit IgG anti-TUP1(Gift from Joseph Reese), 1:10000 rabbit IgG-HRP (target-ing ProteinA epitope) antibodies. Detection was performedusing ImageQuant LAS4000 (GE Healthcare Life Sciences)following incubation with UptiLightTM HRP blot chemilumi-nescent ECL substrate (Interchim).

Acknowledgements

We are grateful to Rebecca Zordan, Brendan Cormack, SadriZnaidi, Christophe D’enfert, Gilles Fischer, Nicolas Agier,Marc Lemaire, Cecile Fairhead, Monique Bolotin-Fukuhara,Hugo Lavoie, Joseph Reese and Suzanne Noble for giving usplasmids, strains, antibodies and useful advices. We thankthe ‘Plateforme de protéomique structurale et fonctionnellede l’Institut Jacques Monod’ for mas spectrometry analyses.We thank Stéphane Le Crom and Jean-Charles Cadoret forgranting us access to microarray facilities. This work wasfunded by the STRUDYEV project of the Agence nationalepour la recherche and by the Emergence program of Univer-sité Pierre et Marie Curie.

References

Azevedo, D., Nascimento, L., Labarre, J., Toledano, M.B.,and Rodrigues-Pousada, C. (2007) The S. cerevisiae Yap1and Yap2 transcription factors share a common cadmium-sensing domain. FEBS Lett 581: 187–195.

Baek, Y.U., Li, M., and Davis, D.A. (2008) Candida albicansferric reductases are differentially regulated in response todistinct forms of iron limitation by the Rim101 and CBFtranscription factors. Eukaryot Cell 7: 1168–1179.

Becerra, M., Lombardia-Ferreira, L.J., Hauser, N.C.,Hoheisel, J.D., Tizon, B., and Cerdan, M.E. (2002) Theyeast transcriptome in aerobic and hypoxic conditions:effects of hap1, rox1, rox3 and srb10 deletions. Mol Micro-biol 43: 545–555.

Bergstrom, A., Simpson, J.T., Salinas, F., Barre, B., Parts, L.,Zia, A., et al. (2014) A high-definition view of functionalgenetic variation from natural yeast genomes. Mol BiolEvol 31: 872–888.

Bodenmiller, D.M., and Spiro, S. (2006) The yjeB (nsrR) geneof Escherichia coli encodes a nitric oxide-sensitive tran-scriptional regulator. J Bacteriol 188: 874–881.

Bourgarel, D., Nguyen, C.C., and Bolotin-Fukuhara, M. (1999)HAP4, the glucose-repressed regulated subunit of the

HAP transcriptional complex involved in the fermentation-respiration shift, has a functional homologue in the respira-tory yeast Kluyveromyces lactis. Mol Microbiol 31: 1205–1215.

Bretes, H., Rouviere, J.O., Leger, T., Oeffinger, M., Devaux,F., Doye, V., and Palancade, B. (2014) Sumoylation of theTHO complex regulates the biogenesis of a subset ofmRNPs. Nucleic Acids Res 42: 5043–5058.

Bucher, P. (1990) Weight matrix descriptions of four eukary-otic RNA polymerase II promoter elements derived from502 unrelated promoter sequences. J Mol Biol 212: 563–578.

Buisson, N., and Labbe-Bois, R. (1998) Flavohemoglobinexpression and function in Saccharomyces cerevisiae. Norelationship with respiration and complex response to oxi-dative stress. J Biol Chem 273: 9527–9533.

Chen, C., Pande, K., French, S.D., Tuch, B.B., and Noble,S.M. (2011) An iron homeostasis regulatory circuit withreciprocal roles in Candida albicans commensalism andpathogenesis. Cell Host Microbe 10: 118–135.

Chen, K., Wilson, M.A., Hirsch, C., Watson, A., Liang, S., Lu,Y., et al. (2013) Stabilization of the promoter nucleosomesin nucleosome-free regions by the yeast Cyc8-Tup1 core-pressor. Genome Res 23: 312–322.

Chiranand, W., McLeod, I., Zhou, H., Lynn, J.J., Vega, L.A.,Myers, H., et al. (2008) CTA4 transcription factor mediatesinduction of nitrosative stress response in Candida albi-cans. Eukaryot Cell 7: 268–278.

Conde e Silva, N., Goncalves, I.R., Lemaire, M., Lesuisse, E.,Camadro, J.M., and Blaiseau, P.L. (2009) KlAft, theKluyveromyces lactis ortholog of Aft1 and Aft2, mediatesactivation of iron-responsive transcription through thePuCACCC Aft-type sequence. Genetics 183: 93–106.

Courel, M., Lallet, S., Camadro, J.M., and Blaiseau, P.L.(2005) Direct activation of genes involved in intracellulariron use by the yeast iron-responsive transcription factorAft2 without its paralog Aft1. Mol Cell Biol 25: 6760–6771.

Courey, A.J., and Jia, S. (2001) Transcriptional repression:the long and the short of it. Genes Dev 15: 2786–2796.

Crawford, M.J., and Goldberg, D.E. (1998) Regulation of theSalmonella typhimurium flavohemoglobin gene. A newpathway for bacterial gene expression in response to nitricoxide. J Biol Chem 273: 34028–34032.

Delaunay, A., Isnard, A.D., and Toledano, M.B. (2000) H2O2sensing through oxidation of the Yap1 transcription factor.EMBO J 19: 5157–5166.

Forrester, M.T., and Foster, M.W. (2012) Protection fromnitrosative stress: a central role for microbial flavohemo-globin. Free Radic Biol Med 52: 1620–1633.

Forsburg, S.L., and Guarente, L. (1989) Identification andcharacterization of HAP4: a third component of the CCAAT-bound HAP2/HAP3 heteromer. Genes Dev 3: 1166–1178.

Gabaldon, T., Martin, T., Marcet-Houben, M., Durrens, P.,Bolotin-Fukuhara, M., Lespinet, O., et al. (2013) Compara-tive genomics of emerging pathogens in the Candidaglabrata clade. BMC Genomics 14: 623.

Genolevures, C., Souciet, J.L., Dujon, B., Gaillardin, C.,Johnston, M., Baret, P.V., et al. (2009) Comparativegenomics of protoploid Saccharomycetaceae. GenomeRes 19: 1696–1709.



Goncalves, I.R., Conde e Silva, N., Garay, C.L., Lesuisse, E.,Camadro, J.M., and Blaiseau, P.L. (2014) The basis forevolution of DNA-binding specificity of the Aft1 transcriptionfactor in yeasts. Genetics 196: 149–160.

Gsaller, F., Hortschansky, P., Beattie, S.R., Klammer, V.,Tuppatsch, K., Lechner, B.E., et al. (2014) The Janus tran-scription factor HapX controls fungal adaptation to bothiron starvation and iron excess. EMBO J 33: 2261–2276.

Hall, T.A. (1999) BioEdit: a user-friendly biological sequencealignment editor and analysis program for Windows 95/98/NT. Nucleic Acids Symp 41: 95–98. Oxford UniversityPress.

Hanlon, S.E., Rizzo, J.M., Tatomer, D.C., Lieb, J.D., and Buck,M.J. (2011) The stress response factors Yap6, Cin5, Phd1,and Skn7 direct targeting of the conserved co-repressorTup1-Ssn6 in S. cerevisiae. PLoS ONE 6: e19060.

Hickman, M.J., and Winston, F. (2007) Heme levels switch thefunction of Hap1 of Saccharomyces cerevisiae betweentranscriptional activator and transcriptional repressor. MolCell Biol 27: 7414–7424.

Homann, O.R., Dea, J., Noble, S.M., and Johnson, A.D.(2009) A phenotypic profile of the Candida albicans regu-latory network. PLoS Genet 5: e1000783.

Hortschansky, P., Eisendle, M., Al-Abdallah, Q., Schmidt,A.D., Bergmann, S., Thon, M., et al. (2007) Interaction ofHapX with the CCAAT-binding complex–a novel mecha-nism of gene regulation by iron. EMBO J 26: 3157–3168.

Hortschansky, P., Ando, E., Tuppatsch, K., Arikawa, H.,Kobayashi, T., Kato, M., et al. (2015) Deciphering the com-binatorial DNA-binding code of the CCAAT-bindingcomplex and the iron-regulatory basic region leucine zipper(bZIP) transcription factor HapX. J Biol Chem 290: 6058–6070.

Hromatka, B.S., Noble, S.M., and Johnson, A.D. (2005) Tran-scriptional response of Candida albicans to nitric oxide andthe role of the YHB1 gene in nitrosative stress and viru-lence. Mol Biol Cell 16: 4814–4826.

Hsu, P.C., Yang, C.Y., and Lan, C.Y. (2011) Candida albicansHap43 is a repressor induced under low-iron conditionsand is essential for iron-responsive transcriptional regula-tion and virulence. Eukaryot Cell 10: 207–225.

Iijima, M., Shimizu, H., Tanaka, Y., and Urushihara, H. (2000)Identification and characterization of two flavohemoglobingenes in Dictyostelium discoideum. Cell Struct Funct 25:47–55.

Ischiropoulos, H., and Gow, A. (2005) Pathophysiologicalfunctions of nitric oxide-mediated protein modifications.Toxicology 208: 299–303.

de Jesus-Berrios, M., Liu, L., Nussbaum, J.C., Cox, G.M.,Stamler, J.S., and Heitman, J. (2003) Enzymes that coun-teract nitrosative stress promote fungal virulence. Curr Biol13: 1963–1968.

Jourdren, L., Duclos, A., Brion, C., Portnoy, T., Mathis, H.,Margeot, A., and Le Crom, S. (2010) Teolenn: an efficientand customizable workflow to design high-quality probesfor microarray experiments. Nucleic Acids Res 38: e117.

Jung, W.H., Saikia, S., Hu, G., Wang, J., Fung, C.K.,D’Souza, C., et al. (2010) HapX positively and negativelyregulates the transcriptional response to iron deprivation inCryptococcus neoformans. PLoS Pathog 6: e1001209.

Kuo, D., Licon, K., Bandyopadhyay, S., Chuang, R., Luo, C.,Catalana, J., et al. (2010) Coevolution within a transcrip-tional network by compensatory trans and cis mutations.Genome Res 20: 1672–1678.

Lan, C.Y., Rodarte, G., Murillo, L.A., Jones, T., Davis, R.W.,Dungan, J., et al. (2004) Regulatory networks affected byiron availability in Candida albicans. Mol Microbiol 53:1451–1469.

Lemoine, S., Combes, F., Servant, N., and Le Crom, S.(2006) Goulphar: rapid access and expertise for standardtwo-color microarray normalization methods. BMC Bioin-formatics 7: 467.

Li, L., Bagley, D., Ward, D.M., and Kaplan, J. (2008) Yap5 isan iron-responsive transcriptional activator that regulatesvacuolar iron storage in yeast. Mol Cell Biol 28: 1326–1337.

Liti, G., Carter, D.M., Moses, A.M., Warringer, J., Parts, L.,James, S.A., et al. (2009) Population genomics of domes-tic and wild yeasts. Nature 458: 337–341.

Liu, L., Zeng, M., Hausladen, A., Heitman, J., and Stamler,J.S. (2000) Protection from nitrosative stress by yeast fla-vohemoglobin. Proc Natl Acad Sci USA 97: 4672–4676.

Longtine, M.S., McKenzie, A., 3rd, Demarini, D.J., Shah,N.G., Wach, A., Brachat, A., et al. (1998) Additionalmodules for versatile and economical PCR-based genedeletion and modification in Saccharomyces cerevisiae.Yeast 14: 953–961.

Lopez-Berges, M.S., Capilla, J., Turra, D., Schafferer, L.,Matthijs, S., Jochl, C., et al. (2012) HapX-mediated ironhomeostasis is essential for rhizosphere competence andvirulence of the soilborne pathogen Fusarium oxysporum.Plant Cell 24: 3805–3822.

Lorenz, M.C., Bender, J.A., and Fink, G.R. (2004) Transcrip-tional response of Candida albicans upon internalization bymacrophages. Eukaryot Cell 3: 1076–1087.

Louvel, H., Gillet-Markowska, A., Liti, G., and Fischer, G.(2014) A set of genetically diverged Saccharomyces cer-evisiae strains with markerless deletions of multiple aux-otrophic genes. Yeast 31: 91–101.

McNabb, D.S., and Pinto, I. (2005) Assembly of the Hap2p/Hap3p/Hap4p/Hap5p-DNA complex in Saccharomycescerevisiae. Eukaryot Cell 4: 1829–1839.

McNabb, D.S., Xing, Y., and Guarente, L. (1995) Cloning ofyeast HAP5: a novel subunit of a heterotrimeric complexrequired for CCAAT binding. Genes Dev 9: 47–58.

Malave, T.M., and Dent, S.Y. (2006) Transcriptional repres-sion by Tup1-Ssn6. Biochem Cell Biol 84: 437–443.

Mantovani, R. (1999) The molecular biology of the CCAAT-binding factor NF-Y. Gene 239: 15–27.

Merhej, J., Frigo, A., Le Crom, S., Camadro, J.M., Devaux, F.,and Lelandais, G. (2014) bPeaks: a bioinformatics tool todetect transcription factor binding sites from ChIPseq datain yeasts and other organisms with small genomes. Yeast31: 375–391.

Pathania, R., Navani, N.K., Gardner, A.M., Gardner, P.R.,and Dikshit, K.L. (2002) Nitric oxide scavenging and detoxi-fication by the Mycobacterium tuberculosis haemoglobin,HbN in Escherichia coli. Mol Microbiol 45: 1303–1314.

Pimentel, C., Vicente, C., Menezes, R.A., Caetano, S.,Carreto, L., and Rodrigues-Pousada, C. (2012) The role ofthe Yap5 transcription factor in remodeling gene expres-



sion in response to Fe bioavailability. PLoS ONE 7:e37434.

Rietzschel, N., Pierik, A.J., Bill, E., Lill, R., and Muhlenhoff, U.(2015) The Basic Leucine Zipper Stress Response Regu-lator Yap5 Senses High-Iron Conditions by Coordination of[2Fe-2S] Clusters. Mol Cell Biol 35: 370–378.

Rodrigues-Pousada, C., Menezes, R.A., and Pimentel, C.(2010) The Yap family and its role in stress response. Yeast27: 245–258.

Sarver, A., and DeRisi, J. (2005) Fzf1p regulates an inducibleresponse to nitrosative stress in Saccharomyces cerevi-siae. Mol Biol Cell 16: 4781–4791.

Scannell, D.R., Butler, G., and Wolfe, K.H. (2007) Yeastgenome evolution–the origin of the species. Yeast 24:929–942.

Schiestl, R.H., and Gietz, R.D. (1989) High efficiency trans-formation of intact yeast cells using single stranded nucleicacids as a carrier. Curr Genet 16: 339–346.

Schillig, R., and Morschhauser, J. (2013) Analysis of afungus-specific transcription factor family, the Candida albi-cans zinc cluster proteins, by artificial activation. Mol Micro-biol 89: 1003–1017.

Schinko, T., Berger, H., Lee, W., Gallmetzer, A., Pirker, K.,Pachlinger, R., et al. (2010) Transcriptome analysis ofnitrate assimilation in Aspergillus nidulans reveals connec-tions to nitric oxide metabolism. Mol Microbiol 78: 720–738.

Schrettl, M., Beckmann, N., Varga, J., Heinekamp, T.,Jacobsen, I.D., Jochl, C., et al. (2010) HapX-mediatedadaption to iron starvation is crucial for virulence of Asper-gillus fumigatus. PLoS Pathog 6: e1001124.

Sellam, A., Askew, C., Epp, E., Tebbji, F., Mullick, A.,Whiteway, M., and Nantel, A. (2010) Role of transcriptionfactor CaNdt80p in cell separation, hyphal growth, andvirulence in Candida albicans. Eukaryot Cell 9: 634–644.

Sellam, A., Tebbji, F., Whiteway, M., and Nantel, A. (2012) Anovel role for the transcription factor Cwt1p as a negativeregulator of nitrosative stress in Candida albicans. PLoSONE 7: e43956.

Singh, R.P., Prasad, H.K., Sinha, I., Agarwal, N., andNatarajan, K. (2011) Cap2-HAP complex is a critical tran-scriptional regulator that has dual but contrasting roles inregulation of iron homeostasis in Candida albicans. J BiolChem 286: 25154–25170.

Sybirna, K., Guiard, B., Li, Y.F., Bao, W.G., Bolotin-Fukuhara,M., and Delahodde, A. (2005) A new Hansenula polymor-pha HAP4 homologue which contains only the N-terminalconserved domain of the protein is fully functional in Sac-charomyces cerevisiae. Curr Genet 47: 172–181.

Sybirna, K., Petryk, N., Zhou, Y.F., Sibirny, A., and Bolotin-Fukuhara, M. (2010) A novel Hansenula polymorpha tran-scriptional factor HpHAP4-B, able to functionally replacethe S. cerevisiae HAP4 gene, contains an additional bZipmotif. Yeast 27: 941–954.

Tamura, K., Stecher, G., Peterson, D., Filipski, A., andKumar, S. (2013) MEGA6: molecular evolutionary geneticsanalysis version 6.0. Mol Biol Evol 30: 2725–2729.

Tan, K., Feizi, H., Luo, C., Fan, S.H., Ravasi, T., and Ideker,T.G. (2008) A systems approach to delineate functions ofparalogous transcription factors: role of the Yap family inthe DNA damage response. Proc Natl Acad Sci USA 105:2934–2939.

Ter Linde, J.J., and Steensma, H.Y. (2002) A microarray-assisted screen for potential Hap1 and Rox1 target genesin Saccharomyces cerevisiae. Yeast 19: 825–840.

Tucker, N.P., Hicks, M.G., Clarke, T.A., Crack, J.C., Chandra,G., Le Brun, N.E., et al. (2008) The transcriptional repres-sor protein NsrR senses nitric oxide directly via a [2Fe-2S]cluster. PLoS ONE 3: e3623.

Tucker, N.P., Le Brun, N.E., Dixon, R., and Hutchings, M.I.(2010) There’s NO stopping NsrR, a global regulator of thebacterial NO stress response. Trends Microbiol 18: 149–156.

Ullmann, B.D., Myers, H., Chiranand, W., Lazzell, A.L., Zhao,Q., Vega, L.A., et al. (2004) Inducible defense mechanismagainst nitric oxide in Candida albicans. Eukaryot Cell 3:715–723.

Vasudevan, S.G., Armarego, W.L., Shaw, D.C., Lilley, P.E.,Dixon, N.E., and Poole, R.K. (1991) Isolation and nucleotidesequence of the hmp gene that encodes a haemoglobin-likeprotein in Escherichia coli K-12. Mol Gen Genet 226: 49–58.

Vodisch, M., Scherlach, K., Winkler, R., Hertweck, C., Braun,H.P., Roth, M., et al. (2011) Analysis of the Aspergillusfumigatus proteome reveals metabolic changes and theactivation of the pseurotin A biosynthesis gene cluster inresponse to hypoxia. J Proteome Res 10: 2508–2524.

Wood, M.J., Storz, G., and Tjandra, N. (2004) Structural basisfor redox regulation of Yap1 transcription factor localiza-tion. Nature 430: 917–921.

Yang, Y.L., Wang, C.W., Leaw, S.N., Chang, T.P., Wang, I.C.,Chen, C.G., et al. (2012) R432 is a key residue for themultiple functions of Ndt80p in Candida albicans. Cell MolLife Sci 69: 1011–1023.

Yu, J.H., Hamari, Z., Han, K.H., Seo, J.A., Reyes-Dominguez, Y., and Scazzocchio, C. (2004) Double-jointPCR: a PCR-based molecular tool for gene manipulationsin filamentous fungi. Fungal Genet Biol 41: 973–981.

Zhu, J., Jambhekar, A., Sarver, A., and DeRisi, J. (2006) ABayesian network driven approach to model the transcrip-tional response to nitric oxide in Saccharomyces cerevi-siae. PLoS ONE 1: e94.

Zhu, L., Gunn, C., and Beckman, J.S. (1992) Bactericidalactivity of peroxynitrite. Arch Biochem Biophys 298: 452–457.

Zordan, R.E., Ren, Y., Pan, S.J., Rotondo, G., De Las Penas,A., Iluore, J., and Cormack, B.P. (2013) Expression plas-mids for use in Candida glabrata. G3 (Bethesda) 3: 1675–1686.

Supporting information

Additional supporting information may be found in the onlineversion of this article at the publisher’s web-site.



Yap7 is a Transcriptional Repressor of Nitric Oxide Oxidase in Yeasts, which arose from Neofunctionalization after Whole Genome Duplication

Documents