the transcriptional response to alkaline ph in saccharomyces cerevisiae: evidence for...

Molecular Microbiology (2002)

46

(5) 1319ndash1333

copy 2002 Blackwell Publishing Ltd

Blackwell Science LtdOxford UKMMIMolecular Microbiology0950-382XBlackwell Science 200246Original Article

Alkaline response in yeastR Serrano et al

Accepted 4 September 2002 For correspondence E-mail JoaquinArinouabes Tel (

+

34) 93 581 2182 Fax (

+

34) 93 581 2006

The transcriptional response to alkaline pH in

Saccharomyces cerevisiae

evidence for calcium-mediated signalling

Raquel Serrano

1

Amparo Ruiz

1

Dolores Bernal

1

James R Chambers

2

and Joaquiacuten Arintildeo

1

1

Departament de Bioquiacutemica i Biologia Molecular Facultat de Veterinagraveria Universitat Autogravenoma de Barcelona Bellaterra E-08193 Barcelona Spain

2

Invitrogen Corporation Carlsbad CA 92008 USA

Summary

The short-time transcriptional response of yeast cellsto a mild increase in external pH (76) has been inves-tigated using DNA microarrays A total of 150 genesincreased their mRNA level at least twofold within45 min Alkalinization resulted in the repression of232 genes The response of four upregulated genes

ENA1

(encoding a Na

++++

-ATPase also induced by salinestress) and

PHO84

PHO89

and

PHO12

(encodinggenes upregulated by phosphate starvation) wascharacterized further The alkaline response of

ENA1

was not affected by mutation of relevant genesinvolved in osmotic or oxidative signalling but wasdecreased in calcineurin and

rim101

mutantsMapping of the

ENA1

promoter revealed two pH-responsive regions The response of the upstreamregion was fully abolished by the drug FK506 ormutation of

CRZ1

(a transcription factor activated bycalciumcalcineurin) whereas the response of thedownstream region was essentially calcium indepen-dent

PHO84

and

PHO12

responses were unaffectedin

crz1

cells but required the presence of Pho2 andPho4 In contrast part of the alkali-induced expres-sion of

PHO89

was maintained in

pho4

or

pho2

cells but was fully abolished in a

crz1

strain or inthe presence of FK506 Heterologous promoters car-rying the minimal calcineurin-dependent responseelements found in

ENA1

or

FKS2

were able to drivealkaline pH-induced expression These results dem-onstrate that the transcriptional response to alkalinepH involves different signalling mechanisms and thatcalcium signalling is a relevant component of thisresponse

Introduction

Fast adaptation to stressing environmental changes isoften a requirement for cell survival Although post-transcriptional mechanisms can play an important rolein stress response in many cases adaptation involvesremodelling gene expression and is mediated by inductionor repression of a more or less specific set of genes(Estruch 2000)


grows well in a relativelywide range of external pH although it proliferates morerapidly at acidic pH An important factor for the mainte-nance of an acidic environment is the yeast plasma mem-brane H

+

-ATPase which actively extrudes protons Thisactivity generates a proton gradient which drives theimport of many nutrients and cations (van der Rest

et al

1995 Serrano 1996) and is essential for survival There-fore even exposure to a mild alkaline environment can beconsidered as a stress condition

Regulation of gene expression by alkaline pH was firstidentified in the fungi

Aspergillus nidulans

which facili-tates adaptation for growth at a very wide pH range andit has also been described in other fungi such as

Candidaalbicans

in this case linked to filamentous growth (for areview see Denison 2000) The response to alkaline pHin

A nidulans

relies on the proteolytic activation of thePacC transcription factor Homologues of this factor havebeen found in

S cerevisiae

(Rim101) Y

arrowia lipolytica

and

C albicans

(Lambert

et al

1997 Li and Mitchell1997 Ramoacuten

et al

1999) At least in part the transduc-tion pathway that leads to the activation of the transcrip-tion factor is also conserved (Denison

et al

1998)However some aspects of the signalling mechanisms in

A nidulans

do not reproduce fully in other fungi Forinstance cleavage of Rim101 occurs under both acidicand alkaline growth environments (Li and Mitchell 1997)and the possibility of Rim101-independent pathways hasbeen postulated recently in

C albicans

(Davis

et al

2000)and


(Lamb

et al

2001)In contrast to other fungi the study of alkaline response

in

S cerevisiae

has received little attention This is some-what surprising as the powerful genetics and genomicsof budding yeast make this model organism a great toolfor increasing our relatively scarce knowledge about how

Used Mac Distiller 50x Job Options
This report was created automatically with help of the Adobe Acrobat Distiller addition Distiller Secrets v105 from IMPRESSED GmbH13You can download this startup file for Distiller versions 405 and 50x for free from httpwwwimpressedde1313GENERAL ----------------------------------------13File Options13 Compatibility PDF 1213 Optimize For Fast Web View Yes13 Embed Thumbnails Yes13 Auto-Rotate Pages No13 Distill From 13 Distill To Page All Pages13 Binding Left13 Resolution [ 600 600 ] dpi13 Paper Size [ 595 842 ] Point1313COMPRESSION ----------------------------------------13Color Images13 Downsampling Yes13 Downsample Type Average Downsampling13 Downsample Resolution 120 dpi13 Downsampling For Images Above 180 dpi13 Compression Yes13 Compression Type JPEG13 JPEG Quality Medium13 Bits Per Pixel As Original Bit13Grayscale Images13 Downsampling Yes13 Downsample Type Average Downsampling13 Downsample Resolution 120 dpi13 Downsampling For Images Above 180 dpi13 Compression Yes13 Compression Type JPEG13 JPEG Quality Medium13 Bits Per Pixel As Original Bit13Monochrome Images13 Downsampling Yes13 Downsample Type Average Downsampling13 Downsample Resolution 600 dpi13 Downsampling For Images Above 900 dpi13 Compression Yes13 Compression Type ZIP13 Anti-Alias To Gray No1313 Compress Text and Line Art Yes1313FONTS ----------------------------------------13 Embed All Fonts Yes13 Subset Embedded Fonts Yes13 Subset When Percent Of Characters Used is Less 100 13 When Embedding Fails Warn and Continue13Embedding13 Always Embed [ Courier-BoldOblique Helvetica-BoldOblique Courier Helvetica-Bold Times-Bold Courier-Bold Helvetica Times-BoldItalic Times-Roman ZapfDingbats Times-Italic Helvetica-Oblique Courier-Oblique Symbol ]13 Never Embed [ ]1313COLOR ----------------------------------------13Color Management Policies13 Color Conversion Strategy Leave Color Unchanged13 Intent Default13Device-Dependent Data13 Preserve Overprint Settings Yes13 Preserve Under Color Removal and Black Generation No13 Transfer Functions Preserve13 Preserve Halftone Information Yes1313ADVANCED ----------------------------------------13Options13 Use Prologueps and Epilogueps No13 Allow PostScript File To Override Job Options Yes13 Preserve Level 2 copypage Semantics Yes13 Save Portable Job Ticket Inside PDF File Yes13 Illustrator Overprint Mode No13 Convert Gradients To Smooth Shades No13 ASCII Format No13Document Structuring Conventions (DSC)13 Process DSC Comments Yes13 Log DSC Warnings No13 Resize Page and Center Artwork for EPS Files No13 Preserve EPS Information From DSC Yes13 Preserve OPI Comments No13 Preserve Document Information From DSC Yes1313OTHERS ----------------------------------------13 Distiller Core Version 500013 Use ZIP Compression Yes13 Deactivate Optimization No13 Image Memory 524288 Byte13 Anti-Alias Color Images No13 Anti-Alias Grayscale Images No13 Convert Images (lt 257 Colors) To Indexed Color Space Yes13 sRGB ICC Profile sRGB IEC61966-211313END OF REPORT ----------------------------------------1313IMPRESSED GmbH13Bahrenfelder Chaussee 491322761 Hamburg Germany13Tel +49 40 897189-013Fax +49 40 897189-7113Email infoimpressedde13Web wwwimpressedde
Adobe Acrobat Distiller 50x Job Option File
ltlt13 ColorSettingsFile ()13 LockDistillerParams false13 DetectBlends false13 DoThumbnails true13 AntiAliasMonoImages false13 MonoImageDownsampleType Average13 GrayImageDownsampleType Average13 MaxSubsetPct 10013 MonoImageFilter FlateEncode13 ColorImageDownsampleThreshold 1513 GrayImageFilter DCTEncode13 ColorConversionStrategy LeaveColorUnchanged13 CalGrayProfile (None)13 ColorImageResolution 12013 UsePrologue false13 MonoImageResolution 60013 ColorImageDepth -113 sRGBProfile (sRGB IEC61966-21)13 PreserveOverprintSettings true13 CompatibilityLevel 1213 UCRandBGInfo Remove13 EmitDSCWarnings false13 CreateJobTicket true13 DownsampleMonoImages true13 DownsampleColorImages true13 MonoImageDict ltlt K -1 gtgt13 ColorImageDownsampleType Average13 GrayImageDict ltlt VSamples [ 2 1 1 2 ] HSamples [ 2 1 1 2 ] Blend 1 QFactor 076 ColorTransform 1 gtgt13 CalCMYKProfile (US Web Coated (SWOP) v2)13 ParseDSCComments true13 PreserveEPSInfo true13 MonoImageDepth 113 AutoFilterGrayImages false13 SubsetFonts true13 GrayACSImageDict ltlt HSamples [ 2 1 1 2 ] VSamples [ 2 1 1 2 ] Blend 1 QFactor 09 gtgt13 ColorImageFilter DCTEncode13 AutoRotatePages None13 PreserveCopyPage true13 EncodeMonoImages true13 ASCII85EncodePages false13 PreserveOPIComments false13 NeverEmbed [ ]13 ColorImageDict ltlt VSamples [ 2 1 1 2 ] HSamples [ 2 1 1 2 ] Blend 1 QFactor 076 ColorTransform 1 gtgt13 AntiAliasGrayImages false13 GrayImageDepth -113 CannotEmbedFontPolicy Warning13 EndPage -113 TransferFunctionInfo Preserve13 CalRGBProfile (sRGB IEC61966-21)13 EncodeColorImages true13 EncodeGrayImages true13 ColorACSImageDict ltlt HSamples [ 2 1 1 2 ] VSamples [ 2 1 1 2 ] Blend 1 QFactor 09 gtgt13 Optimize true13 ParseDSCCommentsForDocInfo true13 GrayImageDownsampleThreshold 1513 MonoImageDownsampleThreshold 1513 AutoPositionEPSFiles false13 GrayImageResolution 12013 AutoFilterColorImages false13 AlwaysEmbed [ Courier-BoldOblique Helvetica-BoldOblique Courier Helvetica-Bold Times-Bold Courier-Bold Helvetica Times-BoldItalic Times-Roman ZapfDingbats Times-Italic Helvetica-Oblique Courier-Oblique Symbol ]13 ImageMemory 52428813 OPM 013 DefaultRenderingIntent Default13 EmbedAllFonts true13 StartPage 113 DownsampleGrayImages true13 AntiAliasColorImages false13 ConvertImagesToIndexed true13 PreserveHalftoneInfo true13 CompressPages true13 Binding Left13gtgt setdistillerparams13ltlt13 PageSize [ 5760 7920 ]13 HWResolution [ 600 600 ]13gtgt setpagedevice

1320

R Serrano

et al


Molecular Microbiology

46

1319ndash1333

pH changes are detected and how this information isinternalized and transmitted Furthermore as external pHis very easy and inexpensive to monitor and modify thepromoters of alkali-inducible genes may represent usefultools for large-scale gene expression However when thiswork was initiated the number of

Saccharomyces

genesknown to be responsive to alkalinization of the mediumwas surprisingly small

ENA1

encoding a P-type Na

+

-ATPase essential for sodium detoxification (Garciadeblaacutes

et al

1993 Mendoza

et al

1994)

SHC1

and

SCY1

(Hong

et al

1999) and only during its developmenthas more extensive information on the transcriptionalresponse of

S cerevisiae

cells to alkaline conditionsbecome available (Causton

et al

2001 Lamb

et al

2001) The first report evaluated the genomic response toa rather large pH shift (pH 4ndash8) and found 71 genes thatincreased expression at least 21-fold after 2 h The sec-ond report provided a detailed time course analysis of thetranscriptional response after shifting the cells from pH 6to 79 identifying almost 500 genes induced at leastthreefold These studies were based on different technol-ogies and their results showed a relatively poor overlapIn addition we had experimental evidence for a numberof genes that large strain-to-strain differences in the tran-scriptional response to alkaline pH could exist suggestingthat there could be additional pH regulation targets thatcould be missed Furthermore we considered that a moredetailed study of the signalling pathways responsible forthis transcriptional response was required to understandhow budding yeast identifies and adapts to alkaline pHstress

Consequently we report here a genome-wide analysisof the short-term transcriptional response of yeast cells toa mild alkalization of the extracellular medium followed bya detailed analysis of the response of the

ENA1

PHO84

PHO89

and

PHO12

genes revealing that the alkalineresponse involves different components and that calciumsignalling participates in this response

Results

Genome-wide analysis of the yeast transcriptional response to alkaline pH

The response of the genome of

S cerevisiae

to mildalkaline conditions has been evaluated by DNA microar-ray analysis It has been shown that transcriptionalresponses differ substantially in their timing after stressand that monitoring at different time points is essentialFor that reason analysis was carried out using mRNAfrom wild-type yeast cells exposed to pH 76 for 5 25 and45 min Our results indicate that 150 genes were inducedat least twofold at one or more of the time points testedThey can be found grouped into functional families inTable 1 A significant number of genes involved in phos-phate transport and metabolism as well as in ion trans-port and homeostasis (particularly of copper and iron)appear to be upregulated by mild alkaline conditionsA remarkable effect on genes involved in transcriptionand RNA processing was also observed More thanone-third of the induced genes (57) did not have a well-characterized function The response to alkaline pH wasfast and transient in some cases but slower and moresustained in others About one-third of the genes (42cases denoted as lsquoErsquo in Table 1) showed a peak responseafter 5 min This is a very fast response comparable withthat obtained for specific responsive genes after exposureof cells to osmotic stress It should be noted however thatonly a subset of genes was responsive to both alkalineand saline stresses (see specific examples in Fig 1where the response of several genes of interest is docu-mented by Northern blot analysis and

Discussion

)The response of other genes such as

FRE1

ARN3

and

FIT2

was also confirmed using

LacZ

fusion assays (notshown)

We have also detected an at least threefold decreasein 232 genes which are listed in Table 2 grouped intofunctional families It is remarkable that the expression of

Fig 1

Northern blot analysis of representative genes Total RNA from cells grown in YPD at pH 64 (NI) exposed for the indicated times to pH 76 or for 10 min to 04 NaCl (S) was elec-trophoresed transferred to membranes and probed with an

ENA1 Xba

Indash

Xba

I fragment (16 kbp) or the entire ORF of

VTC3

PHM2

(25 kbp)

PHO84

(17 kbp)

PHO89

(16 kbp)

RPL28A

(096 kbp)

ADE12

(13 kbp) and

ADE13

(14 kbp) The signal of

RPL28A

unmodified after the treatments is included for reference

Alkaline response in yeast

1321



46

1319ndash1333

Table 1

Classification into functional families of ORFs whose transcripts are induced at least twofold after alkaline pH stress

Data presented correspond to the maximal induction observed Asterisks denote that the intensity of the hybridization signal was out of the linearrange of detection and therefore induction might be stronger than that presented here The lsquoTimersquo column indicates the kinetics of response Genesshowing the highest induction after 5 min at alkaline pH are denoted lsquoErsquo (Early) lsquoIrsquo (Intermediate) indicates a peak at 25 min and lsquoLrsquo (Late)corresponds to genes that show the highest induction after 45 min at alkaline pH Functional categories were assigned based on informationprovided by MIPS (Mewes

et al

1999) Boldface indicates genes identified as responsive to a wide variety of stresses (Gasch

et al

2000 Causton

et al

2001)

1322

R Serrano

et al



46

1319ndash1333

a large number of amino acid biosynthetic genes wasreduced The effect seems to be quite general as path-ways for the biosynthesis of most amino acid familiesare affected In addition the expression of most genesinvolved in purine biosynthesis a pathway responsible for

de novo

generation of AMP was also severely affected(Table 2 and Fig 1) This is not a general effect on nucle-otide metabolism as the pyrimidine biosynthetic pathwaywas not affected

Characterization of the response of genes

ENA1

PHO84

PHO89

and

PHO12

to alkaline pH

Four alkali-upregulated genes were selected as modelsfor further characterization on the basis of the differenttiming of their response and their transcriptional respon-siveness to other stress situations such as saline (

ENA1

)or phosphate starvation stress (

PHO84

PHO89

and

PHO12

) As shown in Table 1 and Fig 1 the kinetics ofthe transcriptional response to alkaline pH can be ratherdifferent from gene to gene For instance

PHO89

and

ENA1

show a very rapid (and transient) response withinthe first 5ndash10 min whereas

PHO84

and

PHO12

are only

induced after 30 min of exposure to alkaline pH Thesedifferential kinetics were confirmed using

LacZ

transla-tional fusions (not shown) The same reporter constructswere used to demonstrate that the sensitivity to alkalinestress can also differ from gene to gene (Fig 2) A sub-stantial increase in expression can be observed for

PHO84

even below neutral pH and most of the responseis achieved barely above pH 70 with no further increaseabove pH 77ndash79 In contrast the

ENA1

and

PHO89

pro-moters require a stronger signal For instance

ENA1

expression remains quite low until pH rises above 75The

ENA1PMR2A

gene has a complex transcriptionalresponse as it is rapidly induced by osmotic shock aswell as by exposure to toxic concentrations of Na

+

or Li

+

and by alkaline pH The Northern blot experiment shownin Fig 1 suggested that the increase in

ENA1 expressionupon shift to pH 76 could be as strong as that obtainedafter osmotic shock This was tested further by comparingthe change in b-galactosidase activity of an ENA1ndashLacZtranscriptional fusion (pKC201) in wild-type cells shiftedfrom pH 64 (13 plusmn 2 Units) to pH 76 (200 plusmn 9) pH 80(393 plusmn 49) or made 04 M NaCl (276 plusmn 53 Units) There-fore the response to high pH of the ENA1 promoter can

Table 2 Classification into functional families of ORFs whose transcripts are repressed at least threefold after alkaline pH stress

Amino acid metabolismGCV1 GCV2 GCV3 ARG56 ARG4 ARG3 ARG1 ARG81 ARO9 ASN1 ASN2 BAT1 GDH1 CPA2 GLN1 HIS1 HIS4 HIS5 HIS7 ILV2 ILV3 LEU1 LEU4 LYS20 LYS4 LYS12 LYS1 LYS9 MET6 STR2 MET22 SER1 SRY1 SHM1 SHM2 GAT1 SAM4 HOM2 TRP5

Purine biosynthesisADE1 ADE57 ADE3 ADE13 ADE4 ADE12 ADE2 IMD1 MTD1

Carbohydrate metabolism and energeticsGSY1 ADH5 PFK1 PGM2 PYK2 MAE1 PYC2 IDP1 ACO1 ATP15 ATP20 QCR6

Other metabolismSUR2 ERG6 UPC2 OYE3 ACS2 DFR1Cell growth and divisionCLA4 CRN1 RNR4 RFC3 HHF1 CDC15 CKS1 SWI4 SIM1 FAR1 FAR3 MUM2 WTM1 NDD1 IME2

TranscriptionHMRA1 BTT1 DAL81 TAF19 MCM1 RPA49 EGD1 DAT1 TAD2 SEN54 DIS3 FAL1 SMX2 CLF1 PSP2 DBP5Protein synthesis and degradationAOS1 UBA1 CCT7 HSP82 RPL34A RPS7B TUF1 RPL33B RPL5 TIF5 YDR341c

TransportCTP1 AGP1 YHM1 HXT12 HXT9 HXT8 GAL2 MDL1 AAC1 TAT2 HXT11 YOR383c PMA2

Endocytosis and secretionSEC4 KEX1 SEC9 SCD5 SEC16 YAP1801 ENT1 TLG2

OthersGFA1 EXG1 RLM1 SPR1 TIR1 SSA4 FLO8 GAL80 RTG1 EFD1 HAL5 RSE1 RGA2 YAK1 MTL1 HAC1 VMA13 SFL1

UnknownYAR010C YBR053C YBL098W YRO2 YBL101W-A YBR012W-A YBR043C YBR147W YBR134W YBR113W NGR1 YCL020W COS2 YCL042W YCL019W OSH2 YDR380W YEL045C YER160C YFL032W YFR045W YGL010W YGL117W YGL072C YGL157W YGR031W YGR102C YGR182C YHL021C YHR029C YIL039W YIL041W YIL105C YIL122W YJL160C YJL200C YJR026W YKL105C YLL013C YLR037C YLR089C YLR177W YLR186W ECM22 YLR257W YML039W YML040W GIS4 YLR427W YML045W YLR437C YML010W YML095C-A YMR051C YMR046C YMR048W SNZ1 YMR124W HAS1 YMR325W YMR323W YNL101W YNL246W YNR047W YNL339C YOL015W YOL063C INP54 YOL022C MSB4TIR2 PTC5 YOR108W YOR138C YOR203W YOR223W YOR192C YPL133C YPR011C YPR012W YPL281C YPR014C YPR022C YPL229W YPR037C

Boldface indicates genes identified as responsive to a wide variety of stresses (Gasch et al 2000 Causton et al 2001)

Alkaline response in yeast 1323

copy 2002 Blackwell Publishing Ltd Molecular Microbiology 46 1319ndash1333

be similar or even higher than that produced under equiv-alent conditions by an osmotic shock However theobservation that a mutant lacking the Hog1 mitogen-activated protein (MAP) kinase which mediates re-sponses to osmotic stress (Maacuterquez and Serrano 1996)does not show a reduced alkaline pH response (Fig 3)suggests that the osmotic sensing pathway is not involvedin the response of ENA1 to alkaline pH Mutation of theMSN2MSN4 genes encoding transcription factorsinvolved in the general stress response through bindingto the stress response element (STRE) did not reduceENA1 expression after exposure to high pH (in fact the

response was even stronger in this mutant) nor did theabsence of Yap1 a transcription factor required for effi-cient response to oxidative stress In addition a strain witha non-functional SLT2MPK1 gene which codes for aMAP kinase involved in the maintenance of cell integritydid not result in impaired expression In contrast we con-firmed the previous finding (Lamb et al 2001) that cellslacking the RIM101 gene a homologue of the A nidulansPacC required for alkaline response in this organism dis-play reduced ENA1 expression in response to alkaline pH(it must be noted however that the degree of Rim101dependence can be variable as we have found that muta-tion of this gene in an otherwise wild-type backgroundyields an essentially full response) Finally alkaline pHresponse was consistently reduced in cells lacking theCNB1 gene encoding the regulatory subunit of the Ca2+calmodulin-dependent SerThr protein phosphatase cal-cineurin This was indicative that calcium signalling couldplay a role in the alkaline response

A similar analysis was carried out for the PHO84PHO89 and PHO12 promoters As shown in Fig 4 lackof Pho2 or Pho4 required for the expression of genes thatrespond to phosphate starvation fully abolished the alka-line response of PHO84 and PHO12 whereas in the caseof PHO89 a partial response could still be detected Dele-tion of both PHO24 genes had an additive effect leadingto an almost complete lack of PHO89 induction For allthree genes deletion of PHO81 encoding a positive reg-ulatory protein of the phosphate pathway mimicked thedeletion of PHO2 or PHO4 Interestingly we have verifiedthat cells lacking the PHO2 PHO4 or PHO81 genesdisplay a growth defect under alkaline conditions (notshown) A very relevant result was derived from the useof a cnb1 mutant Alkaline induction of PHO84 andPHO12 was essentially unmodified by this mutation Incontrast the response from the PHO89 promoter was

Fig 2 Differential sensitivity of genes ENA1 PHO84 PHO89 and PHO12 to alkaline stress Wild-type DBY746 cells were transformed with plasmids pKC201 (squares) pPHO84-LacZ (circles) pPHO89-LacZ (diamonds) or pPHO12-LacZ (triangles) Cells were shifted from pH 55 to the indicated pH and then grown for the periods of time and under the conditions stated in Experimental procedures (except that for pH lt 72 50 mM Bis-Tris was used as buffer) before assay b-galactosidase activity Data are expressed as the percentage of the maximal response obtained for each plasmid and correspond to a representative experiment

Fig 3 Response of the ENA1 gene to alkaline stress in cells lacking essential components of different stress response pathways The indi-cated strains were transformed with plasmid pKC201 Cultures were subjected to alkaline stress (pH 80) and b-galactosidase activity was measured after 60 min in non-stressed (filled bars) or stressed (empty bars) cells Data are mean plusmn SEM from three to six independent clones

1324 R Serrano et al


completely lost suggesting that calcium signalling playsa fundamental role in alkaline pH activation of the PHO89promoter

A remarkable outcome from these experiments was thatcalcineurin appeared to mediate the effects of alkalinestress on the expression of ENA1 (in part) and PHO89(fully) suggesting a role for calcium signalling in the alka-line response To verify this possibility further we testedall four promoters under alkaline pH or high calcium eitherin the presence of FK506 a drug that inhibits calcineurin

or in cells lacking Crz1Tcn1Hal8 the transcription factorthat mediates a number of calcium-dependent cal-cineurin-mediated transcriptional responses (Matheoset al 1997 Stathopoulos and Cyert 1997) As shown inFig 5 PHO84 and PHO12 did not respond to calciumand their alkaline response was not compromised by incu-bation with FK506 or in the absence of Crz1 As expectedthe response of ENA1 to calcium was fully abolished byFK506 and by lack of CRZ1 whereas the alkalineresponse was only partially lost in both cases FinallyPHO89 was able to respond with roughly equal potencyto alkaline pH and high calcium and these responseswere completely lost when cultures were grown in thepresence of FK506 or when cells were deficient for Crz1function These results confirm the notion that calciummediates the alkaline response of ENA1 and PHO89partially in the first case and completely in the secondone

Functional mapping of the ENA1 promoter reveals a calcium-dependent and a calcium-independent pH-responsive region

To identify the region(s) in the ENA1 promoter responsiblefor the alkaline pH response a preliminary mappingbased on CYC1ndashLacZ reporter constructs containing dif-ferent regions of the ENA1 gene promoter (Alepuz et al1997) was made (not shown) This allowed us to identifythe region between nucleotides (nt) -853 to -358 asresponsible for pH response A more detailed mapping(Fig 6) showed that a region between nt -742 and -490(pMP211) was fully able to elicit the alkaline responseThis region is rather complex and contains most of theregulatory elements relevant for ENA1 expressiondescribed so far (see Discussion) Within this region afragment from nt -751 to -667 (pMR706) did produce asignificant response whereas an adjacent downstreamregion (-681 to -565 pMR605) containing the AGGGGsequence matching a typical STRE was essentially inac-tive The evidence that relevant elements might lie furtherdownstream in the promoter was obtained by testing con-structs pMP212 pMP209 and pMP213 which showedthat a small region from nt -573 to -490 was sufficientto elicit a very substantial part of the ENA1 alkalineresponse Therefore it appears that the Na+-ATPase pro-moter contains at least two separate elements that couldact as targets for signal(s) triggered by exposure to alka-line conditions For descriptive purposes the upstreamregion will be denoted as ARR1 (for alkaline responsiveregion) and the downstream segment as ARR2 (seeFig 6 for sequence information)

The evidence that part of the alkaline pH response ofthe ENA1 promoter may be mediated by calcium and the

Fig 4 Characterization of PHO84 PHO89 and PHO12 response to high pH in different mutant strains The indicated strains were trans-formed with plasmids pPHO84ndashLacZ pPHO89ndashLacZ or pPHO12ndashLacZ and cultures were subjected to alkaline stress (pH 80) for 90 min (pPHO84ndashLacZ and pPHO89ndashLacZ) or 120 min (pPHO12ndashLacZ) b-Galactosidase activity was measured in non-stressed (filled bars) or stressed (empty bars) cells Data are mean plusmn SEM from six independent clones



presence of a CDRE element within ARR1 prompted usto undertake a more detailed analysis To this end theregions from nt -742 to -577 (containing the CDRE ele-ment) and from -573 to -490 (ARR2) were cloned intoplasmid pSLFD-178K to yield plasmids pMRK212 and

pMRK213 respectively and tested for the ability to drivetranscription under conditions that would impair a cal-cium-mediated response As shown in Fig 7 plasmidpMRK212 exhibited an alkaline- and calcium-dependentresponse that was in both cases fully abolished in the

Fig 5 Dependence of a functional calcineurin pathway on high pH response of ENA1 PHO84 PHO89 and PHO12 Wild-type DBY746 or the isogenic strain EDN92 (crz1) was transformed with plasmids pKC201 (ENA1) pPHO84ndashLacZ (PHO84) pPHO89ndashLacZ (PHO89) or pPHO12ndashLacZ (PHO12) and subjected to alkaline stress (pH 80 empty bars) or CaCl2 addition to the medium (02 M dotted bars) as described in Experimental pro-cedures in the presence or absence of the drug FK506 (15 mg ml-1) Filled bars denote b-galactosidase activity in the absence of any treatment Data are mean plusmn SEM from three to six independent clones

Fig 6 Functional mapping of the ENA1 promoter in response to alkaline stress Wild-type DBY746 cells were transformed with the indicated constructs and exposed to alkaline stress (pH 80) for 60 min The inducednon-induced b-galactosidase activity ratio is presented as mean plusmn SEM from six independent clones The segments of the ENA1 promoter included in each plasmid are denoted by boxes and their relative position is indicated by numbers (nt positions from the initiating Met codon) flanking each box Relevant known regulatory elements are depicted schemat-ically as well as underlined within the nucleotide sequence of regions ARR1 and ARR2 (see text for further information) Details on the construction of plasmids can be found in Experimental procedures



presence of the calcineurin inhibitor FK506 or when theexperiment was performed in cells lacking calcineurin orthe calcineurin-activated transcription factor Crz1 In con-trast pMRK213 promoted a strong response to pH thatwas essentially unaltered in the presence of FK506 or incells lacking CNB1 or CRZ1 These results demonstratethat ENA1 response to alkaline pH due to ARR1 is trig-gered by calcium whereas the response involving ARR2is essentially calcium independent

Blocking of the Rim101- and calcineurin-dependent pathways does not fully abolish alkaline pH response from the ENA1 promoter

The evidence that rim101 cells display reduced expres-sion from the ENA1 promoter under alkaline stressprompted us to evaluate whether the calcineurin-insensitive response driven from the ARR2 region of thispromoter could be Rim101 dependent As shown in Fig 8

Fig 7 Evidence for calcium-dependent and calcium-independent pH-responsive regions in the ENA1 promoter The indicated strains were transformed with the reporter plasmids pMRK212 (-742-577) and pMRK213 (-573-490) which contain the ARR1 and ARR2 regions of the ENA1 promoter respectively b-Galactosidase activity was tested under condi-tions described in the legend to Fig 5 Data are mean plusmn SEM from three to six independent clones It should be noted that the pMRK reporter constructs have a lower basal activity than the pMP or pMR series which accounts for the apparent differences in induction when compared with equivalent promoter fragments shown in Fig 6

Fig 8 Evidence for Rim101- and calcineurin-independent activation of the ENA1 promoter under alkaline stress The indicated strains were transformed with the plasmids depicted and b-galactosidase activity was tested in the absence of any treatment (filled bars) or after shifting to pH 80 in the absence (empty bars) or the presence (dotted bars) of the drug FK506 (15 mg ml-1) Data are mean plusmn SEM from six independent clones



lack of Rim101 diminishes the expression of the reportergene from both ARR1 (pMRK212) and ARR2 (pMRK213)regions The activity driven from region ARR1 in rim101cells was fully abrogated by incubation with the calcineurininhibitor FK506 but interestingly in the presence of thisdrug a significant response from region ARR2 was stillobserved in rim101 cells This suggests that a Rim101-and calcineurin-independent pathway able to modulatethe ENA1 promoter must exist and that this pathway actsthrough the ARR2 region

Calcineurin-dependent response elements are sufficient to elicit an alkaline pH response

The evidence that CDRE-containing promoters such asENA1 and PHO89 can respond to alkaline pH in a cal-cium-dependent manner led us to test whether such ele-ments would be sufficient to generate a response toalkaline stress This was tested using the reporter plas-mids pLA containing only nt -732 to -711 from the ENA1promoter (ie the downstream CDRE described by Men-dizabal et al 2001) and pAMS366 which includes fourtandem copies of the CDRE found in the FKS2 promoter(Stathopoulos and Cyert 1997) As shown in Fig 9 bothcalcium-responsive elements were able to drive expres-sion of the LacZ gene and this ability was fully abolishedby incubation of the cells with FK506 Furthermore amutated version of the FKS2 CDRE unable to bind Crz1(Stathopoulos and Cyert 1997) and to respond in thepresence of calcium ions (pAMS364) also failed to induceLacZ expression under alkaline pH stress

Discussion

The analysis of the transcriptional response of yeast tochanging environmental conditions can provide not onlyvery relevant information about fundamental cellular func-tions but also useful tools such as regulatable promoterssuitable for gene expression studies or large-scale protein

production As modification of the pH of the culturemedium is a very simple and inexpensive procedure weconsidered that it would be interesting to learn how manyS cerevisiae genes would develop a transcriptionalresponse under mild alkaline stress (ie under conditionsthat would not compromise cell growth) and to gain insightinto the molecular basis of the alkaline response in theyeast S cerevisiae

We have found that about 150 genes can be inducedat least twofold under mild alkaline conditions A relevantquestion is how many of the pH-responsive genesdetected in this work are also induced in most other stressconditions (ie heat shock saline reductive oxidativestarvation) and can therefore be considered members ofthe set of genes involved in the general stress responseBy pooling our information with different sets of dataregarding the transcriptional response to other stress con-ditions such as heat shock saline oxidative etc (Gaschet al 2000 Posas et al 2000 Rep et al 2000 Caustonet al 2001) we have found that 28 genes induced underour conditions (Table 1) can be considered members ofthe environmental stress response (ESR) or the commonenvironmental response (CER) gene categories (Gaschet al 2000 Causton et al 2001) This does not meanthat the remaining genes display a fully specific responseto alkaline exposure as the expression of a significantnumber of them can also be modified to some extentunder a particular type of stress For instance about 65of the genes listed in Table 1 are also induced by salinestress (Posas et al 2000 Rep et al 2000) This impliesthat although there is a relatively large number of genesin which the alkaline and saline stress responses coin-cide these are most probably independent events assuggested by Lamb et al (2001) and by our own results

Our data show that a number of genes involved inphosphate transport and metabolism appear to be upreg-ulated by alkaline pH These include not only PHO84 andPHO89 encoding high-affinity phosphateH+ and phos-phateNa+ co-transporters respectively (for a review see

Fig 9 Calcineurin-dependent response ele-ments are sufficient to elicit an alkaline pH response Wild-type DBY746 cells were trans-formed with the reporter plasmid pLA which contains the -732-711 CDRE found in the ENA1 promoter Plasmids pAMS366 contain-ing a tandem of four copies of the CDRE found in FKS2 and pAMS364 carrying a mutated non-functional version of the tandem (Statho-poulos and Cyert 1997) were introduced into wild-type or crz1 cells Cells were grown in high pH or high calcium in the presence or absence of FK506 as indicated in the legend to Fig 5 Data are mean plusmn SEM from three to six inde-pendent clones



Persson et al 1999) and PHO12 a repressible acidphosphatase but also the recently characterized genesVTC1PHM4 VTC3PHM2 and VTC4PHM3 Thesegenes have been identified as members of a yeast chap-erone family involved in the distribution of the vacuolarH+-ATPase (Cohen et al 1999) as well as possiblyresponsible for polyphosphate synthesis at the vacuole(Ogawa et al 2000) Therefore alkaline pH seems totrigger a transcriptional response quite similar to that ofphosphate starvation Our observation that lack of com-ponents of the phosphate starvation response pathway(such as PHO81) abolishes the alkaline response ofPHO12 and PHO84 suggests that exposure to alkali actu-ally generates a situation of phosphate starvation thatwould be responsible for the induction of some of thesegenes This hypothesis is reinforced by our observationsthat pho2 pho4 and pho81 mutants display an alkaline-sensitive phenotype and that for PHO84 and PHO12 thekinetics of transcriptional response to alkaline pH andphosphate starvation are quite similar (not shown) How-ever the situation could be quite different in the case ofPHO89 as deduced from the faster pH response shownin Fig 1 the differential calcineurin-dependence docu-mented in Fig 4 and the observation that exposure to lowphosphate results in an induction almost insignificantwhen compared with the increase in expression after shiftto alkaline pH (data not shown)

An additional function clearly disturbed by alkaline pHconditions refers to copper and iron transport and homeo-stasis Components of both the AFT1RCS1-regulated(ARN1 ARN2 and ARN3) and the high-affinity ferrous(FET3) siderophore-iron uptake system are induced aswell as the predominant membrane-associated ferric andcupric reductase encoded by FRE1 Most of these genesare also induced when the availability of copper or iron ismarkedly decreased (Askwith and Kaplan 1998) As cop-per and iron solubility can be reduced by alkaline pH theobserved increase in expression might simply reflect adecreased concentration of these cations in the mediumAlthough quantitative determination by atomic absorptionspectrometry of both total and 022 mm filterable copperand iron present in the medium did not reveal significantdifferences when control and alkaline media were com-pared (data not shown) it is still possible that the actualavailability of these cations could be impaired by alkalineconditions

While this work was in progress two reports haveaddressed the study of changes in transcriptionalresponse after alkaline pH exposure One of these (Lambet al 2001) identified 71 candidate alkaline-responsivegenes that increased their mRNA level at least 21-foldafter 2 h of exposure to pH 80 Fourteen of these genesare also identified in our study The relatively low numberof genes found by these authors can possibly be attributed

to the transient nature of the mRNA increase a featurecommon to many responsive genes under a variety ofstresses This is supported by the observation that withthe only exception of PHO89 and ENA1 all 14 commongenes display an intermediate or late response in ourstudy The second study (Causton et al 2001) addressedthe evaluation of changes in expression after differentperiods (from 10 to 100 min) of exposure to pH 79 andreported 476 genes with at least a threefold increase atany given time point Interestingly comparison of this setof data with that presented in Table 1 reveals only 33common genes As microarray data are expressed as theratio between basal and stress conditions a possibleexplanation for this relatively low level of concordancecould be both the lower initial pH (60 versus 64 in ourcase) and higher final pH used in this study (79 versus76) In addition the different yeast genetic backgroundused in these experiments could have a substantial influ-ence at least for a number of specific genes In thisregard we have tested the response of PHO84 andPHO89 promoters to pH 80 in five distinct yeast strainsusing LacZ fusion assays and found large differences(more than 10-fold in some cases) in the non-inducedinduced ratio essentially as a result of strain-to-straindifferences in the level of expression under non-inducedconditions (not shown)

Exposure of yeast to alkaline pH also results in adecrease in the mRNA level of a set of genes It shouldbe noted that this alkaline-induced decrease in expressionis largely independent of the general stress response Infact only 21 out of 232 repressed genes (Table 2) are alsorepressed by a variety of stresses (Gasch et al 2000Causton et al 2001) As exposure to different stress con-ditions such heat shock oxidative or osmotic stress hasbeen shown to slow growth and induce G1 arrest tran-siently (Rowley et al 1993 Lee et al 1996 Belliacute et al2001) it could be argued that the observed decrease inexpression of the amino acid and purine biosyntheticgenes could be part of a general downregulation effect onthe biosynthetic machinery associated with slow growthHowever this is probably not the case as this decreasehas been observed under mild alkaline conditions(pH 76) which do not alter growth in this strain as judgedby monitoring cell growth rate and budding index (notshown)

The selection of several alkaline pH-responsive genesalso involved in other stress responses combined with theuse of yeast mutants defective in different stress signallingtransduction pathways has provided a useful insight intothe molecular basis of the alkaline response Our dataindicate that the alkaline effect on the ENA1 promoterinvolves at least two different signalling pathways thatexclude the osmotic responsive the general stress andthe oxidative response pathways These pathways have



different regions of the promoter as targets The upstreamtarget (ARR1) essentially corresponds to the 3cent CDREelement (nt -727 to -719) recently dissected by Mendiz-abal et al (2001) We show in agreement with theseauthors that the upstream low-affinity Crz1-binding ele-ment found in the ENA1 promoter (nt -826 to -807) doesnot contribute significantly to the alkaline response (com-pare constructs pMR805 and pMR705 in Fig 6) There-fore calcium signalling represents one of the componentsof the ENA1 response to alkaline pH which would explainthe previous finding that calcineurin was required for thefull response of ENA1 to saline and alkaline conditions(Mendoza et al 1994) The downstream alkaline pH-responsive region of ENA1 (ARR2 nt -573 to -490)which accounts for 60ndash70 of the response is rathercomplex and contains most of the regulatory elementsrelevant for ENA1 expression described so far such asthe Mig1-binding motif (-533 to -544) which was charac-terized as an upstream-repressing sequence regulated bycarbon source as well as the Sko1-binding motif (-502 to-513) which is responsible for transcriptional repressionregulated by osmotic stress (Proft and Serrano 1999)However the data presented here (Fig 3) indicate that theHog1-mediated osmotic stress signal is not necessary foralkaline induction of the ENA1 promoter and publishedevidence suggested that a mig1 strain retains the abilityto induce an ENA1ndashLacZ reporter when shifted at high pH(Alepuz et al 1997) Furthermore we have verified thatmost of the response derived from the ARR2 region ispreserved in sko1 or mig1 mutants (not shown) There-fore it seems reasonable to assume that this region maycontain an unidentified regulatory element involved inalkaline response

The functional analysis of PHO84 PHO89 and PHO12promoters reveals that ENA1 is not the only gene able tobe induced through a calcium-mediated mechanism afterexposure to high pH Although response of PHO84 andPHO12 is not affected by conditions that would blockcalcium signalling we confirm here the recent finding(Yoshimoto et al 2002) that expression of PHO89 can beinduced by external calcium in agreement with the exist-ence in its promoter of a CDRE-like consensus sequenceat positions -515 to -504 We also show that induction ofPHO89 after alkaline stress is roughly equal in potency tothat observed after exposure to calcium and that bothresponses are fully dependent on the Cnb1 and Crz1functions Therefore in the case of PHO89 the only sig-nalling component for high pH seems to be calcium medi-ated which is in agreement with our finding that CDREsequences are sufficient to promote expression underalkaline pH stress From these observations it would beexpected that a burst of intracellular calcium should followexposure to alkaline pH An ongoing project in our labo-ratory aims to monitor this hypothetical change in calcium

levels and to identify the source of this cation It is worthnoting that a very recent genome-wide survey on cal-cineurin-mediated response to high calcium and sodium(Yoshimoto et al 2002) has identified about 160 genesshowing calcineurin dependence (defined as a gt50response in cnb1 cells compared with the wild-typestrain) In addition these authors provided a DNA consen-sus sequence for Crz1 binding (GNGGC[GT]CA) Wehave found that 38 genes from our Table 1 correspond togenes experimentally determined as calcineurin depen-dent or containing the consensus Crz1-binding sequencewithin the 1 kbp upstream untranslated region This indi-cates that although we prove that placing a Crz1p site infront of a reporter can make it pH responsive not all genesknown to be activated through calcineurin or containingputative Crz1 binding sites do respond to high pH This isperhaps because of the presence of additional regulatoryelements in these promoters

The early finding that calcineurin mutants are alkalisensitive (Nakamura et al 1993) might lead to specula-tion about the functional significance of calcium signallingin adaptation to high pH conditions However it has beenreported that crz1 mutants are not sensitive to alkaline pH(Stathopoulos and Cyert 1997) and in our hands lack ofCrz1 results in only a very mild phenotype much lessapparent that than produced by deletion of CNB1 (notshown) As it has been shown that Crz1 is the major (ifnot the only) effector of calcineurin-regulated geneexpression in yeast (Yoshimoto et al 2002) it must beconcluded that in budding yeast the calcium-induced andcalcineurin-mediated transcription response is probablynot responsible for survival at high pH This implies thatadditional targets for calcineurin with a relevant role inhigh pH adaptation must exist

In addition to those cases in which the transcriptionalresponse could be secondary to alkaline-derived depriva-tion of nutrients (such as perhaps PHO84 and PHO12)from the data presented here and from recent evidence(Lamb et al 2001) it is clear that transcriptional responseto high pH in S cerevisiae involves different signallingpathways One of these pathways would be mediated byRim101 as mutation of RIM101 results in poor growthunder high pH and failure to induce expression of severalgenes under alkaline conditions such as ARN4 andYOL154w (Lamb et al 2001) These same authors how-ever reported other cases in which the responseappeared to be only partially Rim101 dependent or evenfully independent of the presence of the transcription fac-tor In fact the existence of RIM101-independent alkalinepH responses has also been proposed recently in thecase of C albicans (Davis et al 2000) We documenthere the existence of a second calcium-mediated path-way that can be fully (PHO89) or partially (ENA1) respon-sible for high pH response The case of ENA1 is



particularly interesting as we also show that blockingof the calcium-mediated pathway in a rim101 mutantdoes not fully abolish the transcriptional ENA1 responseto high pH suggesting that additional Rim101- andcalcineurin-independent pathways involved in the alkalinepH response may exist in S cerevisiae

Experimental procedures

Yeast strains and growth conditions

Yeast cells were grown at 28infinC in YPD medium (10 g l-1 yeastextract 20 g l-1 peptone and 20 g l-1 dextrose) or when indi-cated in synthetic minimal (SD) or complete minimal (CM)medium (Adams et al 1997) The relevant genotype of thestrains described in this work can be found in Table 3 Dis-ruption of HOG1 and RIM101 was made by replacingsequences from +40 to +1311 of the HOG1 open readingframe (ORF) and sequences from +44 to +1865 of theRIM101 ORF by the heterologous marker KANMx4 accord-ing to the short flanking homology replacement strategy(Wach et al 1994) Deletion of CRZ1 was accomplished byreplacing the entire ORF exactly as described by Stathopou-los and Cyert (1997) Mutation of CNB1 was made by poly-merase chain reaction (PCR) amplification of the mutantlocus present in the cnb1 strain generated by Winzeler et al(1999) This DNA fragment was subsequently used to trans-form the wild-type strains used in this work Disruption ofPHO2 was accomplished as follows a BamHIndashBamHI frag-ment containing the TRP1 gene (from plasmid YDp-W) wasinserted into the BglII site of the PHO2 gene (cloned into thep2AAH plasmid) to give pDPPHO2 A DraI 17 kbp fragmentwas excised from pDPPHO2 and used to transform cellsDisruption of PHO4 with a 22 kbp LEU2 marker gene wasperformed by transforming cells with a 38 kbp AvaIndashHindIIIfragment derived from plasmid pDPPHO4 Plasmids p2AAHand pDPPHO4 were generous gifts from W Houmlrz Universityof Munich Deletion of PHO81 was performed as follows A41 kbp region of the PHO81 genomic locus spanning fromnt -428 to -160 downstream of the stop codon was amplifiedby PCR and cloned into the KpnI site of pUC19 to giveplasmid pPHO81 Then a 18 kbp BglII fragment of the ORF

was replaced by a 16 kbp LEU2 marker recovered fromplasmid YDp-L by digestion with BamHI yielding plasmidpPHO81D-La This plasmid was digested with AvaI and theresulting 29 kbp fragment was used to transform cells Allgene deletions generated in this work were confirmed byPCR

Genomic microarray analysis

Genomic yeast DNA fragments were amplified by PCR puri-fied as described by Posas et al (2000) and arrayed ontoSuperAldehyde substrates (TeleChem International) ForRNA preparation DBY746 cells were grown in YPD to anoptical density of 08 and split into aliquots Cells were cen-trifuged and resuspended in either fresh YPD (pH 64) or YPDcontaining 20 mM TAPS (pH 76) Cultures were centrifugedfor 5 min at 7000 rpm and total RNA was extracted usinghot phenol and glass beads as described previously (Koumlhrerand Dombey 1991) Poly (A)+ RNA was purified from totalRNA with an mRNA isolation kit (Roche) based on biotin-labelled oligo (dT) probes and streptavidin bound to magneticparticles according to the manufacturerrsquos instructions BeforecDNA synthesis RNAs were tested by Northern blot usingspecific probes

Fluorescently labelled cDNA was prepared from poly (A)+

RNA by oligo dT-primed polymerization using Superscript IIreverse transcriptase (LTI) and Cy3-dCTP (Amersham Phar-macia Biotech) essentially as described by Posas et al(2000) Probes were purified using Qiaquick PCR purificationkits (Qiagen) Yeast microarrays were hybridized in duplicatefor 4 h under cover slides with a Cy3-dCTP-labelled cDNAprobe After hybridization microarrays were washed driedand subsequently scanned using a confocal laser ScanArray3000 system (Packard) Data were collected using IMAGENE

software (BioDiscovery) The complete set of data will beavailable through the following web address httpquirouabesJarino

Northern blot analysis

For Northern blot analysis total RNA (20 mg per lane) waselectrophoresed on 07 agarosendashformaldehyde gels and

Table 3 Saccharomyces cerevisiae strains used in this study

Name Relevant genotype Sourcereference

DBY746 MATa ura3-52 leu2-3112 his3-D1 trp1-D239 D BotsteinEDN92 DBY746 crz1KAN This workRSC1 DBY746 hog1KAN This workRSC4 DBY746 pho4LEU2 This workRSC5 DBY746 pho2TRP1 This workRSC6 DBY746 pho2TRP1 pho4LEU2 This workRSC17 DBY746 pho81LEU2 This workRSC21 DBY746 rim101KAN This workMAR15 DBY746 cnb1KAN This workJA100 MATa ura3-52 leu2-3112 his4 trp1-1 can-1r De Nadal et al (1998)JC10 JA100 mpk1LEU2 Vissi et al (2001)W3031A MATa ade-2-2 his3-1115 leu2-3112 ura3-1 trp1-1 R RothsteinPEY70 W3031A msn2HIS3 msn4TRP1 Alepuz et al (1997)Wyap1 W3031A yap1KAN F Estruch



transferred to positively charged nylon membranes (Roche)under vacuum Probes were labelled using the random-primed DNA labelling kit (Roche) Membranes were hybrid-ized at 65infinC in the presence of hybridization buffer (700 mMNaCl 40 mM NaH2PO4 pH 76 4 mM EDTA 02 polyvin-lylpirrolidone 02 Ficoll 01 SDS 02 mg ml-1 salmonsperm DNA) and 106 cpm ml-1 of the appropriate 32P-labelled DNA fragment Filters were washed in 01yen SSC (1yenSSC is 150 mM NaCl 15 mM sodium citrate pH 70) 01SDS at 68infinC

Plasmids construction and b-galactosidase analysis

To evaluate the response of the different promoters underalkaline pH conditions a number of b-galactosidase re-porter plasmids were used Initial mapping of the ENA1response was made using the multicopy pKC plasmid series(Cunningham and Fink 1996 Alepuz et al 1997) Theseconstructs derive from plasmid pKC201 which containsENA1 sequences from -1385 to +35 (relative to the startinginitiating Met) fused to LacZ A more precise mapping wasaccomplished using the pMP (Proft and Serrano 1999) andpMR (this work) series Both series contain different PCR-amplified stretches from the ENA1 promoter with added PstIand XhoI restriction sites to facilitate cloning into the CYC1ndashLacZ reporter plasmid pJS205 (Schuller et al 1992) Plas-mids pMRK212 and 213 were prepared by excising theappropriate promoter fragment from the pMP vector by PstIndashXhoI digestion cloning into these sites of plasmid pBluescriptSK and transfer into the SmaIndashXhoI of plasmid pSLFD-178K(Idrissi et al 1998) Plasmid pMRK1303 was constructed byamplification from genomic DNA of the -1304-361 region ofthe ENA1 promoter with added SmaIndashXhoI sites and cloninginto these sites of pSLFD-178K The pMP and pMR series ofplasmids differ from the pMRK one in that the latter displaysvirtually no reporter activity in the absence of a transcription-ally active DNA insert Plasmid pLA (a gift from I Fernandezde Larrinoa UPV Spain) derives from plasmid pLGD312s(Guarente and Mason 1983) by cloning into the SalI site the-732-711 sequence of the ENA1 promoter This sequencecontains the downstream high-affinity Crz1-binding elementidentified by Mendizabal et al (2001) Plasmids pAMS366and pAMS364 which drive the expression of the lacZreporter gene from four copies in tandem of the CDRE (cal-cineurin-dependent response element) present in the FKS2promoter were a gift from M Cyert (Stanford CA USA)pAMS366 carries a wild-type version of these elementswhereas pAMS364 contains a mutated form unable to bindto Crz1 (Stathopoulos and Cyert 1997) To analyse theexpression of PHO84 PHO89 and PHO12 LacZ transla-tional fusions were constructed as follows The regionscomprising nt -603 to +19 (pPHO84ndashLacZ) -671 to +33(pPHO89ndashLacZ) or -1275 to +78 (pPHO12ndashLacZ) wereamplified by PCR with added EcoRI and HindIII sites andcloned into these sites of plasmid YEp357 (Myers et al1986)

Except where stated otherwise cells were grown to anOD660 of 05ndash10 and 25 ml aliquots were made centrifugedand resuspended in the appropriate media For alkalinestress cells were resuspended in rich medium containing50 mM TAPS adjusted to the appropriate pH with potassium

hydroxide (taking into account that autoclaving leads to adecrease in the pH of the medium) For all promoters non-induced cells were resuspended in the same medium (lowpH medium) adjusted to pH 63 except in the case ofPHO84 in which pH was decreased to 56 to ensure that thestarting pH in all cases would not result in significant activityof the promoter It should be noted that the PHO84 promoterdisplays substantial activity even below neutrality (seeFig 2) For calcium treatment solid CaCl2 was added to thelow-pH medium to achieve a final concentration of 02 MFK506-containing media were made by adding the appropri-ate volume of a stock solution (10 mg ml-1 made in 90ethanol 10 Tween P-20) to give a final concentration of15 mg ml-1 In all cases growth was resumed for 60 min(ENA1) 90 min (PHO84 and PHO89) or 120 min (PHO12) toensure maximal response cells were then centrifuged andb-galactosidase activity was measured as described by Rey-nolds et al (1997)

Acknowledgements

The contribution of F Posas to the earliest steps of this workis acknowledged We thank F Estruch M Proft R SerranoI Loacutepez-Calderoacuten M Cyert and W Houmlrz for plasmids strainsand advice The excellent technical assistance of Anna Vilaltaand Manuel Clemente is acknowledged This work was sup-ported by a research grant from the lsquoFundacioacuten RamoacutenArecesrsquo to JA which also supported a fellowship awardedto R Serrano A Ruiz is the recipient of a fellowship from theCIRIT (Generalitat de Catalunya)

References

Adams A Gottschlings DE Kaiser CA and Stearns T(1997) Methods in Yeast Genetics Cold Spring Harbor NYCold Spring Harbor Laboratory Press

Alepuz PM Cunningham KW and Estruch F (1997)Glucose repression affects ion homeostasis in yeastthrough the regulation of the stress-activated ENA1 geneMol Microbiol 26 91ndash98

Askwith C and Kaplan J (1998) Iron and copper transportin yeast and its relevance to human disease Trends Bio-chem Sci 23 135ndash138

Belliacute G Gariacute E Aldea M and Herrero E (2001) Osmoticstress causes a G1 cell cycle delay and downregulation ofCln3Cdc28 activity in Saccharomyces cerevisiae MolMicrobiol 39 1022ndash1035

Causton HC Ren B Koh SS Harbison CT Kanin EJennings EG et al (2001) Remodeling of yeast genomeexpression in response to environmental changes Mol BiolCell 12 333ndash337

Cohen A Perzov N Nelson H and Nelson N (1999) Anovel family of yeast chaperons involved in the distributionof V-ATPase and other membrane proteins J Biol Chem274 26885ndash26893

Cunningham KW and Fink GR (1996) Calcineurin inhib-its VCX1-dependent H+Ca2+ exchange and induces Ca2+

ATPases in Saccharomyces cerevisiae Mol Cell Biol 162226ndash2237

Davis D Wilson RB and Mitchell AP (2000) RIM101-



dependent and-independent pathways govern pHresponses in Candida albicans Mol Cell Biol 20 971ndash978

De Nadal E Clotet J Posas F Serrano R Goacutemez Nand Arintildeo J (1998) The yeast halotolerance determinantHal3p is an inhibitory subunit of the Ppz1p SerThr proteinphosphatase Proc Natl Acad Sci USA 95 7357ndash7362

Denison SH (2000) pH regulation of gene expression infungi Fungal Genet Biol 29 61ndash71

Denison SH Negrete-Urtasun S Mingot JM Tilburn JMayer WA Goel A et al (1998) Putative membranecomponents of signal transduction pathways for ambientpH regulation in Aspergillus and meiosis in Saccharomy-ces are homologous Mol Microbiol 30 259ndash264

Estruch F (2000) Stress-controlled transcription factorsstress-induced genes and stress tolerance in buddingyeast FEMS Microbiol Rev 24 469ndash486

Garciadeblaacutes B Rubio F Quintero FJ Bantildeuelos MAHaro R and Rodriacuteguez-Navarro A (1993) Differentialexpression of two genes encoding isoforms of the ATPaseinvolved in sodium efflux in Saccharomyces cerevisiaeMol Gen Genet 236 363ndash368

Gasch AP Spellman PT Kao CM Carmel-Harel OEisen MB Storz G et al (2000) Genomic expressionprograms in the response of yeast cells to environmentalchanges Mol Biol Cell 11 4241ndash4257

Guarente L and Mason T (1983) Heme regulates tran-scription of the CYC1 gene of S cerevisiae via anupstream activation site Cell 32 1279ndash1286

Hong SK Han SB Snyder M and Choi EY (1999)SHC1 a high pH inducible gene required for growth atalkaline pH in Saccharomyces cerevisiae Biochem Bio-phys Res Commun 255 116ndash122

Idrissi F-Z Fernandez-Larrea JB and Pintildea B (1998)Structural and functional heterogeneity of Rap1p com-plexes with telomeric and UASrpg-like DNA sequences JMol Biol 284 925ndash935

Lamb TM Xu W Diamond A and Mitchell AP (2001)Alkaline response genes of S cerevisiae and their relation-ship to the RIM101 pathway J Biol Chem 276 1850ndash1856

Lambert M Blanchin-Roland S Le Louedec F LepingleA and Gaillardin C (1997) Genetic analysis of regulatorymutants affecting synthesis of extracellular proteinases inthe yeast Yarrowia lipolytica identification of a RIM101pacC homolog Mol Cell Biol 17 3966ndash3976

Lee J Romeo A and Kosman DJ (1996) Transcriptionalremodeling and G1 arrest in dioxygen stress in Saccharo-myces cerevisiae J Biol Chem 271 24885ndash24893

Li W and Mitchell AP (1997) Proteolytic activation ofRim1p a positive regulator of yeast sporulation and inva-sive growth Genetics 145 63ndash73

Maacuterquez JA and Serrano R (1996) Multiple transductionpathways regulate the sodium-extrusion gene PMR2ENA1 during salt stress in yeast FEBS Lett 382 89ndash92

Matheos DP Kingsbury TJ Ahsan US and Cunning-ham KW (1997) Tcn1pCrz1p a calcineurin-dependenttranscription factor that differentially regulates geneexpression in Saccharomyces cerevisiae Genes Dev 113445ndash3458

Mendizabal I Pascual-Ahuir A Serrano R and de Lar-rinoa IF (2001) Promoter sequences regulated by the

calcineurin-activated transcription factor Crz1 in the yeastENA1 gene Mol Gen Genet 265 801ndash811

Mendoza I Rubio F Rodriacuteguez-Navarro A and PardoJM (1994) The protein phosphatase calcineurin is essen-tial for NaCl tolerance of Saccharomyces cerevisiae J BiolChem 269 8792ndash8796

Mewes HW Heumann K Kaps A Mayer K Pfeiffer FStocker S and Frishman D (1999) MIPS a database forprotein sequences and complete genomes Nucleic AcidsRes 27 44ndash48

Myers AM Tzagoloff A Kinney DM and Lusty CJ(1986) Yeast shuttle and integrative vectors with multiplecloning sites suitable for construction of lacZ fusions Gene45 299ndash310

Nakamura T Liu Y Hirata D Namba H Harada SHirokawa T and Miyakawa T (1993) Protein phos-phatase type 2B (calcineurin)-mediated FK506-sensitiveregulation of intracellular ions in yeast is an importantdeterminant for adaptation to high salt stress conditionsEMBO J 12 4063ndash4071

Ogawa N De Risi J and Brown PO (2000) New compo-nents of a system for phosphate accumulation and poly-phosphate metabolism in Saccharomyces cerevisiaerevealed by genomic expression analysis Mol Biol Cell 114309ndash4321

Persson BL Petersson J Fristedt U Weinander RBerhe A and Pattison J (1999) Phosphate permeasesof Saccharomyces cerevisiae structure function and reg-ulation Biochim Biophys Acta 1422 255ndash272

Posas F Chambers JR Heyman JA Hoeffler JP deNadal E and Arintildeo J (2000) The transcriptionalresponse of yeast to saline stress J Biol Chem 27517249ndash17255

Proft M and Serrano R (1999) Repressors and upstreamrepressing sequences of the stress-regulated ENA1 genein Saccharomyces cerevisiae bZIP protein Sko1p confersHOG-dependent osmotic regulation Mol Cell Biol 19 537ndash546

Ramoacuten AM Porta A and Fonzi WA (1999) Effect ofenvironmental pH on morphological development ofCandida albicans is mediated via the PacC-related tran-scription factor encoded by PRR2 J Bacteriol 181 7524ndash7530

Rep M Krantz M Thevelein JM and Hohmann S(2000) The transcriptional response of Saccharomycescerevisiae to osmotic shock Hot1p and Msn2pMsn4p arerequired for the induction of subsets of high osmolarityglycerol pathway-dependent genes J Biol Chem 2758290ndash8300

van der Rest ME Kamminga AH Nakano A AnrakuY Poolman B and Konings WN (1995) The plasmamembrane of Saccharomyces cerevisiae structure func-tion and biogenesis Microbiol Rev 59 304ndash322

Reynolds A Lundblad V Dorris D and Keaveney M(1997) Yeast vectors and assays for expression of clonedgenes In Current Protocols in Molecular Biology AusubelFM Brent R Kingston RE Moore DD SeidmanJG Smith JA and Struhl K (eds) New York USAJohn Wiley amp Sons pp 1361ndash1366

Rowley A Johnston GC Butler B Werner-WashburneM and Singer RA (1993) Heat shock-mediated cell cycle



blockage and G1 cyclin expression in the yeast Saccharo-myces cerevisiae Mol Cell Biol 13 1034ndash1041

Schuller HJ Hahn A Troster F Schutz A and Sch-weizer E (1992) Coordinate genetic control of yeast fattyacid synthase genes FAS1 and FAS2 by an upstreamactivation site common to genes involved in membranelipid biosynthesis EMBO J 11 107ndash114

Serrano R (1996) Salt tolerance in plants and microorgan-isms toxicity targets and defense responses Int Rev Cytol165 1ndash52

Stathopoulos AM and Cyert MS (1997) Calcineurin actsthrough the CRZ1TCN1-encoded transcription factor toregulate gene expression in yeast Genes Dev 11 3432ndash3444

Vissi E Clotet J De Nadal E Barceloacute A Bakoacute EGergely P et al (2001) Functional analysis of the Neu-

rospora crassa Pzl1 protein phosphatase by expression inbudding and fission yeast Yeast 18 115ndash124

Wach A Brachat A Pohlmann R and Philippsen P(1994) New heterologous modules for classical or PCR-based gene disruptions in Saccharomyces cerevisiaeYeast 10 1793ndash1808

Winzeler EA Shoemaker DD Astromoff A Liang HAnderson K Andre B et al (1999) Functional charac-terization of the Saccharomyces cerevisiae genome bygene deletion and parallel analysis Science 285 901ndash906

Yoshimoto H Saltsman K Gasch AP Li HX OgawaN Botstein D et al (2002) Genome-wide analysis ofgene expression regulated by the calcineurinCrz1p signal-ing pathway in Saccharomyces cerevisiae J Biol Chem277 31079ndash31088

1320

R Serrano

et al



46

1319ndash1333

pH changes are detected and how this information isinternalized and transmitted Furthermore as external pHis very easy and inexpensive to monitor and modify thepromoters of alkali-inducible genes may represent usefultools for large-scale gene expression However when thiswork was initiated the number of

Saccharomyces

genesknown to be responsive to alkalinization of the mediumwas surprisingly small

ENA1

encoding a P-type Na

+

-ATPase essential for sodium detoxification (Garciadeblaacutes

et al

1993 Mendoza

et al

1994)

SHC1

and

SCY1

(Hong

et al

1999) and only during its developmenthas more extensive information on the transcriptionalresponse of

S cerevisiae

cells to alkaline conditionsbecome available (Causton

et al

2001 Lamb

et al

2001) The first report evaluated the genomic response toa rather large pH shift (pH 4ndash8) and found 71 genes thatincreased expression at least 21-fold after 2 h The sec-ond report provided a detailed time course analysis of thetranscriptional response after shifting the cells from pH 6to 79 identifying almost 500 genes induced at leastthreefold These studies were based on different technol-ogies and their results showed a relatively poor overlapIn addition we had experimental evidence for a numberof genes that large strain-to-strain differences in the tran-scriptional response to alkaline pH could exist suggestingthat there could be additional pH regulation targets thatcould be missed Furthermore we considered that a moredetailed study of the signalling pathways responsible forthis transcriptional response was required to understandhow budding yeast identifies and adapts to alkaline pHstress

Consequently we report here a genome-wide analysisof the short-term transcriptional response of yeast cells toa mild alkalization of the extracellular medium followed bya detailed analysis of the response of the

ENA1

PHO84

PHO89

and

PHO12

genes revealing that the alkalineresponse involves different components and that calciumsignalling participates in this response

Results

Genome-wide analysis of the yeast transcriptional response to alkaline pH

The response of the genome of

S cerevisiae

to mildalkaline conditions has been evaluated by DNA microar-ray analysis It has been shown that transcriptionalresponses differ substantially in their timing after stressand that monitoring at different time points is essentialFor that reason analysis was carried out using mRNAfrom wild-type yeast cells exposed to pH 76 for 5 25 and45 min Our results indicate that 150 genes were inducedat least twofold at one or more of the time points testedThey can be found grouped into functional families inTable 1 A significant number of genes involved in phos-phate transport and metabolism as well as in ion trans-port and homeostasis (particularly of copper and iron)appear to be upregulated by mild alkaline conditionsA remarkable effect on genes involved in transcriptionand RNA processing was also observed More thanone-third of the induced genes (57) did not have a well-characterized function The response to alkaline pH wasfast and transient in some cases but slower and moresustained in others About one-third of the genes (42cases denoted as lsquoErsquo in Table 1) showed a peak responseafter 5 min This is a very fast response comparable withthat obtained for specific responsive genes after exposureof cells to osmotic stress It should be noted however thatonly a subset of genes was responsive to both alkalineand saline stresses (see specific examples in Fig 1where the response of several genes of interest is docu-mented by Northern blot analysis and

Discussion

)The response of other genes such as

FRE1

ARN3

and

FIT2

was also confirmed using

LacZ

fusion assays (notshown)

We have also detected an at least threefold decreasein 232 genes which are listed in Table 2 grouped intofunctional families It is remarkable that the expression of

Fig 1

Northern blot analysis of representative genes Total RNA from cells grown in YPD at pH 64 (NI) exposed for the indicated times to pH 76 or for 10 min to 04 NaCl (S) was elec-trophoresed transferred to membranes and probed with an

ENA1 Xba

Indash

Xba

I fragment (16 kbp) or the entire ORF of

VTC3

PHM2

(25 kbp)

PHO84

(17 kbp)

PHO89

(16 kbp)

RPL28A

(096 kbp)

ADE12

(13 kbp) and

ADE13

(14 kbp) The signal of

RPL28A

unmodified after the treatments is included for reference


1321



46

1319ndash1333

Table 1



et al


et al

2000 Causton

et al

2001)

1322

R Serrano

et al



46

1319ndash1333


de novo



ENA1

PHO84

PHO89

and

PHO12

to alkaline pH


ENA1


PHO84

PHO89

and

PHO12


PHO89

and

ENA1


PHO84

and

PHO12

are only


LacZ


PHO84


ENA1

and

PHO89


ENA1


ENA1PMR2A


+

or Li

+
















































Discussion













































Acknowledgements


References























































1321



46

1319ndash1333

Table 1



et al


et al

2000 Causton

et al

2001)

1322

R Serrano

et al



46

1319ndash1333


de novo



ENA1

PHO84

PHO89

and

PHO12

to alkaline pH


ENA1


PHO84

PHO89

and

PHO12


PHO89

and

ENA1


PHO84

and

PHO12

are only


LacZ


PHO84


ENA1

and

PHO89


ENA1


ENA1PMR2A


+

or Li

+
















































Discussion













































Acknowledgements


References






















































1322

R Serrano

et al



46

1319ndash1333


de novo



ENA1

PHO84

PHO89

and

PHO12

to alkaline pH


ENA1


PHO84

PHO89

and

PHO12


PHO89

and

ENA1


PHO84

and

PHO12

are only


LacZ


PHO84


ENA1

and

PHO89


ENA1


ENA1PMR2A


+

or Li

+
















































Discussion













































Acknowledgements


References
























































































Discussion













































Acknowledgements


References





























































































Acknowledgements


References






















































the transcriptional response to alkaline ph in saccharomyces cerevisiae: evidence for...

Documents