engineering of promoter replacement cassettes for fine ... · efﬁciency dh5 chemically competent...

APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Aug. 2006, p. 5266–5273 Vol. 72, No. 80099-2240/06/$08.00�0 doi:10.1128/AEM.00530-06Copyright © 2006, American Society for Microbiology. All Rights Reserved.

Engineering of Promoter Replacement Cassettes for Fine-Tuningof Gene Expression in Saccharomyces cerevisiae

Elke Nevoigt,1,2 Jessica Kohnke,2 Curt R. Fischer,1 Hal Alper,1 Ulf Stahl,2 and Gregory Stephanopoulos1*Department of Chemical Engineering, Massachusetts Institute of Technology, Room 56-469, Cambridge, Massachusetts 02139,1 and

Department of Microbiology and Genetics, Berlin University of Technology, Seestr. 13, D-13353 Berlin, Germany2

Received 6 March 2006/Accepted 21 April 2006

The strong overexpression or complete deletion of a gene gives only limited information about its controlover a certain phenotype or pathway. Gene function studies based on these methods are therefore incomplete.To effect facile manipulation of gene expression across a full continuum of possible expression levels, werecently created a library of mutant promoters. Here, we provide the detailed characterization of our yeastpromoter collection comprising 11 mutants of the strong constitutive Saccharomyces cerevisiae TEF1 promoter.The activities of the mutant promoters range between about 8% and 120% of the activity of the unmutated TEF1promoter. The differences in reporter gene expression in the 11 mutants were independent of the carbon sourceused, and real-time PCR confirmed that these differences were due to varying levels of transcription (i.e.,caused by varying promoter strengths). In addition to a CEN/ARS plasmid-based promoter collection, we alsocreated promoter replacement cassettes. They enable genomic integration of our mutant promoter collectionupstream of any given yeast gene, allowing detailed genotype-phenotype characterizations. To illustrate theutility of the method, the GPD1 promoter of S. cerevisiae was replaced by five TEF1 promoter mutants ofdifferent strengths, which allowed analysis of the impact of glycerol 3-phosphate dehydrogenase activity on theglycerol yield.

In both functional genomics and metabolic engineering,strategies for fine-tuning gene expression are required to studythe control exerted by target genes on phenotypes or metabolicfluxes of interest. Traditionally, gene expression varies amongthree conditions: (i) the wild type, (ii) the gene knockout, and(iii) strong overexpression of the target gene. However, severalexamples where the optimization of target phenotypes wasachieved by moderate rather than strong multicopy expressionof a target gene have been given (12, 13, 30). These exampleshighlight the need for tools that enable the fine-tuning andprecise control of gene expression in Saccharomyces cerevisiae.

Several strategies have been explored to obtain a gradedtarget gene expression. One strategy is to clone native promot-ers of various strengths to drive the expression of a desiredgene (34). However, this method is not robust, since endoge-nous promoters may be subject to different regulation modal-ities, even despite their designation as “constitutive.” More-over, a native promoter of a given desired strength may beunavailable or difficult to identify. A second strategy uses vec-tors of different copy numbers to adjust gene expression levels(12, 13, 30). This approach is limited by the availability ofplasmids of any given desired copy number. Other problemswith this technique are the metabolic burden associated withmaintenance of high-copy number plasmids and the cell-cellheterogeneity in expression caused by copy number variancebetween single cells of a culture. A third strategy for modulat-ing gene expression has been to titrate an inducible promotersystem with various concentrations of its inducer (11, 25).

However, inducible systems often suffer from inducer toxicityand inducer-mediated pleiotropic effects. Furthermore, the useof most inducers is prohibitively expensive at industrial scales.Cell-cell heterogeneity is a problem with this strategy, as well,because it is not clear that the level of induction is homoge-neous in all cells of a population. For example, a single-celllevel investigation of an inducible (araBAD) promoter in Esch-erichia coli uncovered the fact that the dose response of re-porter gene expression to increasing inducer concentration wasactually due to an increasing fraction of induced versus unin-duced cells (21). A final strategy for manipulating gene expres-sion has been the creation of artificial promoter libraries. Thisstrategy has been implemented mainly in bacteria (2, 9, 28).There is also one example in which a synthetic promoter li-brary was created in yeast; however only three members of thislibrary were chosen for further study (11). To date, no broad-range, well-characterized promoter collection is available foryeast which can be used directly in the replacement of anypromoter in the yeast genome.

Our laboratory recently described the creation of a pro-moter library in yeast (2) in which promoters of graduallyincreasing strength were generated by subjecting the TEF1(translation and elongation factor 1) promoter of S. cerevisiaeto error-prone PCR. TEF1 promoter mutants with definedactivities were selected. The present study describes the furthercharacterization of the TEF1 promoter mutant collection, thegeneration of promoter replacement cassettes for genomic in-tegration, and its utility for metabolic pathway analysis.

MATERIALS AND METHODS

Strains, cultivation conditions, and reagents. The yeast strain BY4741 (MATahis3�1 leu2�0 met15�0 ura3�0) (created by Brachmann et al. [4a]) was obtainedfrom EUROSCARF, Frankfurt, Germany. Escherichia coli DH5� (Maximum

* Corresponding author. Mailing address: Department of ChemicalEngineering, Massachusetts Institute of Technology, 77 MassachusettsAvenue, Cambridge, MA 02139. Phone: (617) 258-0398. Fax: (617)253-3122. E-mail: [email protected].

5266

on August 22, 2019 by guest

http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

TABLE 1. Primers used

Use/name Sequence

Error-prone PCR of TEF1 promoterP1...............................................................................................................................ATTGGGACAACACCAGTGAATAATTCTTCACCTTTAGA

CATTTTTCTP2...............................................................................................................................ACGCCAAGCGCGCAATTAACCCTCACTAAAGGGAACAA

AAGCTGGAGC

Real-time reverse transcription PCR of yECitrine mRNAP3...............................................................................................................................ATGGCTGACAAACAAAAGAATGP4...............................................................................................................................CAGATTGATAGGATAAGTAATG

Integration of both selectable genetic markersa

Forward primerb

P5...........................................................................................................................TACGCCAAGCGCGCAATTAACCCTCACTAAAGGGAACAAAAGCTGCAGCTGAAGCTTCGTACGC

Reverse primerc

P6a (unmutated TEF1p and mutant 11) ..........................................................AAAATCTGGAAGAGTAAAAAAGGAGTAGAAACATTTTGAAGCTATGCATAGGCCACTAGTGGATCT

P6b (mutant 2).....................................................................................................AAAACCTGGAGGAGTAAAGGGGGAGCCGAAGCACTTTAGAGCCGTGCATAGGCCACTAGTGGATCT

P6c (mutant 6) .....................................................................................................AAAATCTGGAAGAGTAAAAAAGGAGTAGAAACATTCTGAAGCTATGCATAGGCCACTAGTGGATCT

P6d (mutant 8).....................................................................................................AAGATCTGGAAGAGTAAAAGAGGAGTAGAAACGTTTCGAAGCTATGCATAGGCCACTAGTGGATCT

P6e (mutants 5 and 7) ........................................................................................AAAATCTGGAAGAGTAAAAAAGGAGTAGAGACATTTTGAAGCTATGCATAGGCCACTAGTGGATCT

P6f (mutant 12) ...................................................................................................AGAATCTGGAAGAGTAAAAAAGGAGTAGAAACATTTTGAAGCTATGCATAGGCCACTAGTGGATCT

P6g (mutant 9).....................................................................................................AAAGTCTGGAAGAGTAAAAAAGGAGTAGAAACATTTTGAAGCTATGCATAGGCCACTAGTGGATCT

P6h (mutant 4).....................................................................................................AAAATCTGGAAGAGTAAAAAGGGGGTAGAAGCGTTTTGAAGCTATGCATAGGCCACTAGTGGATCT

P6i (mutant 10)....................................................................................................AAAATCTGGAAGAGTAACAAAGGAGTAGAAACATTTTGAAGCTATGCATAGGCCACTAGTGGATCT

P6j (mutant 3)......................................................................................................AAAATCTGGAAGAGTAAAGAAGGAGTAGAAACATTTTGAGGCTATGCATAGGCCACTAGTGGATCT

Genomic integration of TEF1 promoter versions upstreamof GPD1 gene (Fig. 4)

Forward primerb

P7...........................................................................................................................TCGTTACTCCGATTATTTTGTACAGCTGATGGGACCTTGCCGTCTCAGCTGAAGCTTCGTACGC

Reverse primerc

P8a (unmutated TEF1p and mutants 3, 5, 6, 7, 9, 10, and 11) .....................AGCATTCAAGTGGCCGGAAGTTAAGTTTAATCTATCAGCAGCAGCAGACATTTTTCTAGAAAACTTAG

P8b (mutant 2).....................................................................................................AGCATTCAAGTGGCCGGAAGTTAAGTTTAATCTATCAGCAGCAGCAGACATTTTTCTAGAAAACTTGG

P8c (mutant 4) .....................................................................................................AGCATTCAAGTGGCCGGAAGTTAAGTTTAATCTATCAGCAGCAGCAGACATTTTTCTAGAGAACTTAG

P8d (mutant 8).....................................................................................................AGCATTCAAGTGGCCGGAAGTTAAGTTTAATCTATCAGCAGCAGCAGACATTTTTCTAAAAAACTTAG

P8e (mutant 12)...................................................................................................AGCATTCAAGTGGCCGGAAGTTAAGTTTAATCTATCAGCAGCAGCAGACATTTTTCTAGAAGACTTAG

PCR diagnosis of genomic integration of TEF1 promoterversions upstream of GPD1 gene

P9 (binds upstream GPD1 promoter) ..................................................................CCCAAGGCAGGACAGTTACCP10 (binds in GPD1 coding sequence).................................................................AGCACCAGATAGAGCACCACAP11 (binds in K.l.LEU2) .........................................................................................GGACCACCAACAGCACCTAGT

a loxP-K.l.LEU2-loxP and loxP-KanMX-loxP, upstream of the CEN/ARS plasmid-based TEF1 promoter mutant collection.b Used for both the unmutated TEF1 promoter and all mutant versions.c Varies with regard to TEF1 mutants.

VOL. 72, 2006 FINE-TUNING OF GENE EXPRESSION IN YEAST 5267


http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

Efficiency DH5� Chemically Competent E. coli; Invitrogen, Carlsbad, CA) wasused for retransformation of yeast plasmid DNA.

Yeast and bacterial strains were stored in 20% glycerol at �70°C. E. coli wasgrown in Luria-Bertani medium. Ampicillin at 100 �g/ml was added to themedium when required. Yeast strain BY4741 without plasmid was cultivated inYPD medium (10 g of yeast extract/liter, 20 g of Bacto Peptone/liter, and 20 gglucose/liter). To select and grow yeast transformants using either URA3 orLEU2 as a selectable marker, we used a yeast synthetic complete (YSC) mediumcontaining 6.7 g of Yeast Nitrogen Base (Difco)/liter, 20 g glucose/liter, and amixture of appropriate nucleotide bases and amino acids (CSM-URA or CSM-LEU [Qbiogene, Irvine, CA], respectively), resulting in YSC Leu� or YSC Ura�.For the growth experiments using a respiratory carbon source, 2% ethanol and2% glycerol were added to the medium instead of 2% glucose. Solid media wereas described above but with 1.5% agar. Yeast cells were routinely cultivated at30°C in Erlenmeyer flasks closed with cotton plugs and shaken at 200 rpm(semiaerobic conditions) without pH control.

Taq polymerase was obtained from New England Biolabs (Beverly, MA).Primers used for PCR and sequencing were purchased from Invitrogen (Carls-bad, CA).

Retransformation of plasmid DNA. The generation of the CEN/ARS plasmid-based yECitrine reporter plasmid, the cloning and error-prone PCR of the S.cerevisiae TEF1 promoter using primers 1 and 2 (Table 1), and the fluorescence-activated cell sorting selection of 11 yeast clones exhibiting graded levels ofspecific reporter protein fluorescence have been described elsewhere (2). Isola-tion of plasmid DNAs from these selected yeast clones was performed using theZymoprep Yeast Plasmid Miniprep Kit (ZYMO Research, Orange, CA), and 2.5ml of DNA samples were transformed into 50 ml Maximum Efficiency DH5�competent cells (Invitrogen, Carlsbad, CA) following the instructions of themanufacturer.

Reporter protein fluorescence measurement. Measurements of specific fluo-rescence were performed using cells harvested from the logarithmic phase duringgrowth in shake flasks. The fluorescence of yECitrine (27) was measured indiluted cultures (optical density at 600 nm [OD600], between 0.1 and 0.3) incuvettes using a fluorescence spectrometer (HITACHI F-2500) with an excita-tion wavelength of 502 nm and an emission wavelength of 532 nm. The specificfluorescence is referred to here as the ratio of the fluorescence level measuredand the OD600 measured in the same cuvette.

Reporter mRNA quantification. Total yeast RNA was extracted using theRiboPureTM-Yeast Kit (AMBION), including DNase treatment. The RNAconcentration was quantified by absorbance at 260 nm. Quantification ofyECitrine mRNA was performed using the iSciptTM One-Step RT-PCR Kit withSYBR Green (Bio-Rad) according to the manufacturer’s instructions and aniCycler (Bio-Rad). We used 100 ng total yeast RNA in one reverse transcription(RT)-PCR. Primers P3 and P4 (Table 1) were used in RT-PCR. Data wereanalyzed using the iCycler software (Bio-Rad Laboratories, Hercules, CA).

Construction of plasmids usable as templates for generating promoter re-placement cassettes. To introduce the selectable and removable markers loxP-K.l.LEU2-loxP and loxP-KanMX-loxP upstream of the different TEF1 promoterversions in the CEN/ARS plasmids, the two marker cassettes were amplified byPCR using plasmids pUG6 (8) and pUG73 (7), respectively, as templates. Prim-ers P5/P6a-j (Table 1) were designed for introducing the selectable markers byrecombination-based cloning (22). PCR conditions were 95°C for 45 seconds,54°C for 1 min, and 72°C for 2 min. PCR was performed for 25 cycles. Theelongation time in the last cycle was 15 min. The 12 purified PCR fragments,together with the 12 SacI-linearized CEN/ARS plasmids bearing the correspond-ing TEF1 promoter versions, were transformed into competent cells of strainBY4741 (Frozen-EZ Yeast Transformation II; ZYMO RESEARCH, Orange, CA).

Genomic integration of the TEF1 promoter mutant collection upstream GPD1coding sequence. Promoter replacement cassettes were amplified by PCR fromthe CEN/ARS plasmid-based TEF1 promoter collection containing the loxP-K.l.LEU2-loxP selectable genetic marker (see Fig. 4). PCR was carried out usingthe forward primer P7 and the reverse primer 8a-e (Table 1) equipped withappropriate flanking sequences for integration of the PCR-amplified cassette byhomologous recombination. PCR conditions were as described above, exceptthat primer annealing was carried out at 52°C. The PCR products were ethanolprecipitated, and about 0.5 �g was transformed into 10 �l of competent yeast cellsof strain BY4741 (Frozen-EZ Yeast Transformation II; ZYMO RESEARCH,Orange, CA).

Correct genomic integration of TEF1 promoter versions was confirmed byPCR diagnosis using two pairs of primers, i.e., P9/P10 and P11/P10 (Table 1).PCR conditions were the same as for the amplification of promoter replacementcassettes (see above), except that the annealing temperature was 60°C and PCRwas performed for 30 cycles.

Determination of specific GPDH activity, glycerol, and glucose. Glycerol3-phosphate dehydrogenase (GPDH) activity was measured in cells growinglogarithmically in YSC Leu� as previously described (19). Measurements ofglycerol and glucose were performed with kits from r-Biopharm (Darmstadt,Germany). Cumulative yields of glycerol (and biomass) on glucose were deter-mined by dividing the net amount of glycerol (or biomass) formed by the netamount of glucose depleted.

RESULTS

Characterization of the TEF1 promoter mutant collectionafter retransformation of plasmids. Recently, an S. cerevisiaeTEF1 promoter mutant library comprising 14,000 clones wascreated by error-prone PCR and cloning of the randomly mu-tated TEF1 promoter variants upstream of a green fluorescentprotein (GFP) reporter within a CEN/ARS plasmid. Usingfluorescence-activated cell sorting, 11 yeast clones with graduallyincreasing fluorescences were sorted out of this library (2).

In order to confirm that the differences in specific fluores-cence were caused by the mutations in the plasmid-based TEF1promoter, the 11 plasmids were isolated and retransformedinto yeast. After retransformation, the specific fluorescence ofeach clone was measured in logarithmically growing cells. Thegrowth rates of the various mutants were all nearly identical,which is important for reliable comparison of specific GFPfluorescences. Two different synthetic media were used: onewith 2% glucose and one with 2% ethanol and 2% glycerol ascarbon sources. The carbon catabolism of S. cerevisiae aerobi-cally grown in medium containing 2% glucose is respirofer-mentative, whereas it is fully repiratory in ethanol-glycerolmedium. Figure 1 shows that the relative strengths of the TEF1promoter mutants were independent of the two growth mediatested. Promoter performance was also measured directly byquantitative RT-PCR for the reporter gene mRNA. Theseresults were well correlated with the reporter gene proteinlevels as measured by specific fluorescence (Fig. 1). Theseresults strongly support the hypothesis that the varying levelsof reporter gene expression in all clones were indeed due todifferences in basal transcriptional activity, i.e., caused by vary-ing strengths of the different TEF1 promoter mutants. Theselected promoters cover a range of activities from about 8 to120% in comparison to the unmutated TEF1 promoter.

All 11 selected TEF1 promoter mutants were sequenced(Fig. 2). Until this sequencing step, we acted on the assumptionthat we had been working with the TEF2 promoter, since theplasmid p416-TEF used as a template for error-prone PCR ofthe TEF promoter (2) was described as containing the TEF2promoter of S. cerevisiae (6, 17). However, it became obviousfrom the sequencing that the plasmid p416-TEF instead con-tains the promoter of the TEF1 gene, which does show a highdegree of homology to the TEF2 promoter (26).

The number of detected mutations within the mutated 401bp of the TEF1 promoter (corresponding to the region from�412 to �11 of the native S. cerevisiae TEF1 promoter) rangedfrom 4 (mutant 10, whose activity is reduced to 78% of that ofthe unmutated TEF1 promoter) to 71 (mutant 2, exhibitingvery low promoter activity). The mutations seemed to be fairlyrandomly distributed throughout the promoter sequence.However, a few positions were mutated in two or even threedistinct mutants, but other regions remained completely un-touched, especially the very 3� end of the promoter (Fig. 2).

5268 NEVOIGT ET AL. APPL. ENVIRON. MICROBIOL.


http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

Interestingly, mutant promoter 6, which showed a higher ac-tivity than the unmutated TEF1 promoter, had only a single-nucleotide deletion at position �251. By deleting the singleguanine at this position, a continuous string of 10 adenines wasgenerated, a fact which may contribute to the increased activitycompared to the unmutated TEF1 promoter.

Generation of promoter replacement cassettes. In our priorwork, promoter delivery into the genome was a prerequisite forobtaining robust and reliable quantitative genotype-phenotyperelationships (2). As such, we created a set of 12 promoterreplacement cassettes which can be used to integrate into theyeast genome and regulate the expression of a desired gene. Toobtain promoter replacement cassettes, it was necessary toclone a selectable genetic marker upstream of the TEF1 pro-moters in our plasmid library. We choose the markers loxP-K.l.LEU2-loxP (7), selectable by leucine prototrophy, and loxP-KanMX-loxP (8), selectable by resistance to the antibioticgeneticin G418. The markers are flanked by loxP sequencesand therefore allow multiple usage of these replacement cas-settes. Furthermore, the use of the KanMX selectable markerenables promoter replacement in industrial yeast strains thatdo not carry auxotrophic mutations. The 3� ends of our primersP5 and P6a-j (Table 1), used for the recombination-based cloning

of loxP-K.l.LEU2-loxP and loxP-KanMX-loxP into the region up-stream of the promoters in our CEN/ARS-based plasmids, cor-respond to the 3� ends of the unique primers previously designedto amplify a set of various selectable markers (7, 8) from thegenetic marker plasmids available at EUROSCARF (http://web.uni-frankfurt.de/fb15/mikro/euroscarf/). Therefore, if required,other genetic markers can be integrated into the CEN/ARS de-rivatives to create alternate knockout cassettes for promoter in-sertion by using the same strategy described here (see Materialsand Methods).

The integration of the selectable markers into the CEN/ARSplasmid-based TEF1 promoter collection had no significantimpact on the activities of the TEF1 promoter mutants. Onlyslight variations in relative specific fluorescence were seenwhen the CEN/ARS plasmid-bearing clones were compared tothe corresponding yeast clones containing a marker integrationsite. This slight variance is attributable to clonal variability.

Use of the TEF1 promoter collection to study control byGPD1 expression of glycerol production in S. cerevisiae. Previ-ous studies suggested that GPDH is the rate-limiting step inthe glycerol biosynthetic pathway of S. cerevisiae (1, 4, 5, 16, 18,19, 23). It has been shown that the deletion of GPD1, one ofthe two isogenes encoding GPDH, caused a reduction in glyc-erol production compared to the GPD1 wild type. Further-more, the strong overexpression of GPD1 led to a significantlyincreased rate and yield of glycerol production. However, it hasalso been found that the multicopy overexpression of GPD1had a negative impact on the growth rate and biomass yield(16, 18–20, 23, 24). Therefore, in order to obtain the simulta-neous phenotypes of high biomass and glycerol yields, it isnecessary to optimize the expression of GPD1. To do so, weused our TEF1 promoter collection to investigate graduallyincreasing levels of GPD1 expression. Five members of theTEF1 promoter library, including the weakest and strongestpromoter mutants, as well as the unmutated TEF1 promoter,were integrated into the yeast genome upstream of the GPD1gene (Fig. 3). For PCR amplification of the promoter replace-ment cassettes, the plasmid collection bearing loxP-K.l.LEU2-loxP as a selectable marker was chosen. The PCR diagnosticsof the LEU2 prototrophic transformants revealed that morethan 90% of the clones showed the correct integration, high-lighting the high efficiency of this method.

The specific GPDH activities of the verified transformantswere measured and normalized to the activity obtained withthe unmutated TEF1 promoter and plotted as a function of thepromoter strength (Fig. 3). The values for promoter strengthwere obtained by calculating the mean values of the two met-rics of relative reporter expression, i.e., fluorescence and tran-script level, as measured in the CEN/ARS plasmid-based pro-moter collection. This promoter strength metric is analogousto the metric used for our previously reported E. coli library(2). A linear correlation, with a similar dynamic range, wasobtained between relative promoter strength and relativeGPDH activity (Fig. 3).

Both glycerol and biomass yields were measured in thefive integrated transformants and plotted as a function ofspecific GPDH activity (Fig. 4). These results are contrastedwith the three data points we obtained from a GPD1 dele-tion mutant, from a strain overexpressing GPD1, from a2�m-based multicopy plasmid, and from the wild-type back-

FIG. 1. Characterization of promoter strengths in the differentmembers of the TEF1 promoter mutant collection after retransforma-tion of plasmids. Three different metrics are shown for each memberof the promoter collection: (i) reporter expression at the protein levelmeasured by specific fluorescence of GFP during growth in glucose asa carbon source, (ii) reporter expression at the mRNA level in cellsgrown in glucose medium measured by real-time PCR, and (iii) specificfluorescence of GFP during growth in ethanol-glycerol as carbonsources. All determinations were carried out using logarithmicallygrowing cells (OD600, about 1.5). The data shown are mean values ofat least two independent experiments and were normalized to thereference (unmutated TEF1 promoter; open symbol). The coefficientsof variance (CV) were 7%, 19%, and 11%. These numbers are thestandard deviation divided by the mean. The value for specific reporterfluorescence in ethanol-glycerol medium (19%) is highly biased bymutant 2, which has the lowest activity and a large standard deviationcompared with the mean. If this point is removed, the average CV is13.5% instead of 19%.



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

ground strain (Fig. 4). This illustrates the new informationobtained by the promoter collection versus the limited datapoints accessible without using it. The curve for the glycerolyield shows that the glycerol yield depends linearly on

GPDH activity in the range of activities obtained by themembers of our TEF1 promoter mutant collection. Theseresults support previous assertions that GPDH constitutes arate-limiting step in the synthesis of glycerol. However, the

FIG. 2. Multiple-sequence alignment of the unmutated TEF1 promoter (TEF1p; boldface) and the 11 selected TEF1 promoter mutants withgraded activities. The normalized promoter strength is shown for each mutant at the beginning of each mutant sequence. The stars mark thosenucleotides which remained unchanged in all promoter versions, whereas mutations are highlighted. The underlined sequences correspond to theproposed transcription factor binding sites according to the literature (3).



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

data also suggest that this linear dependence does not con-tinue indefinitely and saturates well before reaching thelevel obtained in multicopy overexpression of GPD1. More-over, biomass formation was significantly reduced in thetransformant bearing the overexpression plasmid. However,the moderate increase in GPDH activity obtained by usingmembers of our TEF1 promoter collection allowed an in-crease in glycerol yields without having an adverse effect ongrowth. Indeed, moderate GPDH overexpression may evenhave positively affected biomass formation.

DISCUSSION

A collection of TEF1 promoter mutants of various strengthsfor S. cerevisiae (2) was retransformed, rigorously character-

ized, and found to provide a robust tool for controlling geneexpression. Furthermore, the relative strengths of these pro-moters were unaffected by the carbon source. Measurement ofreporter mRNA confirmed that yECitrine fluorescence corre-lated with transcript levels and can be used as a measurementof promoter strength (assuming that the half-lives are the samefor all 12 yECitrine mRNAs generated by our TEF1 promotercollection).

The sequence analysis of the TEF1 promoter mutantsshowed high variance in both the locations and densities of themutated residues. A few mutations were overrepresented in anumber of mutants; however, the number of TEF1 promotermutants investigated here is too low to infer reliably whichmutations have positive, negative, or zero impact on promoter

FIG. 3. Genomic integration of five TEF1 promoter versions (the unmutated TEF1 promoter and four mutants) upstream of the GPD1 gene in S.cerevisiae. (A) Strategy of promoter replacement. (B) Activities of GPDH in the yeast transformants, depending on promoter strength. Promoter strengthwas calculated as the mean value of two different metrics shown in Fig. 1, i.e., relative reporter fluorescence and relative mRNA levels. Error bars indicatethe range between normalized mRNA and normalized fluorescence for each promoter.



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

activity, using the algorithm developed by Jensen et al. (un-published data). Mutations were seen to affect the known tran-scription factor binding sites UASrpg (Rap1p) and CT-Box(Gcr1p) (3), but many mutations were outside of known bind-ing sites. The highest number of mutations (including thosewithin the promoter motifs mentioned above) was observed inmutant 2, which had the lowest promoter activity. The TEF1promoter mutant 6 showed only six point mutations and ex-hibited the highest promoter activity, even higher than that ofthe unmutated TEF1 promoter. It was also interesting that asingle guanidine deletion at position �251 gave rise to a stringof 10 adenosines. The importance of this mutation over theremaining five mutations in this mutant could be ascertaineddefinitively only by site-directed mutagenesis and further anal-ysis of the resulting promoters.

By cloning two different selectable markers (loxP-K.l.LEU2-loxP and loxP-KanMX-loxP) upstream of the TEF1 promotermutants, we provided two additional plasmid collections thatcan be used as templates for PCR amplification of promoterreplacement cassettes. The integration of a promoter in frontof a gene in the genome can overcome problems associatedwith expression from promoters on a plasmid (32). Moreover,the integrative approach abolishes the promotion of targetgene expression from its native promoter and any associatedregulation modalities, which allows a more accurate assess-ment of genotype-phenotype characterization. The use ofKanMX as a selectable marker allows the application of this

promoter collection in industrial yeasts lacking auxotrophies.The loxP sequences flanking the genetic markers enable rescueof the marker and, hence, multiple use of the replacementcassettes. Such multiple replacements are necessary for mostmetabolic-engineering applications, in which the expressionlevels of many pathway enzymes have to be optimized to obtaina maximum flux (15).

The collection of promoter replacement cassettes offers theopportunity to analyze the level of control of any enzymeactivity on metabolic fluxes or phenotypes. Here, the promoterreplacement cassettes were used to vary levels of cytosolicGPDH activities by integrating the TEF1 promoter mutantcollection upstream of the GPD1 gene in the yeast genome.The linear correlation between promoter strength calculatedfrom plasmid-based reporter gene expression and the specificGPDH activity reflecting GPD1 gene expression highlights twoimportant characteristics of this integration system: (i) therelative strengths of our promoters were unaffected by thegenomic integration and (ii) an increase in the promoterstrength of a gene resulted in an equal increase in activity inthe enzyme encoded by that gene.

Moreover, our analysis indicated that GPD1 was transcrip-tionally controlled. The use of this promoter collection up-stream of other genes in the future will certainly reveal enzymeswhose activities vary nonlinearly with promoter strength, espe-cially if posttranslational control prevails.

Another aspect of promoter replacement analysis that willvary with the target gene under investigation is the expressionlevels of the library members relative to the wild-type level.This difference is due to differences in the endogenous levels ofgene transcription, which can vary by orders of magnitudebetween genes. In the case of GPD1 gene expression, somemembers of our promoter collection caused a specific GPDHactivity lower than the wild-type level, and some others mem-bers caused a higher activity. It is conceivable that for a genewith very low expression in the wild type, the weakest promoterof our collection could cause overexpression. We are currentlyworking on increasing the dynamic range of the TEF1 pro-moter collection to extend its utility. The challenge of isolatingvery weak promoters lies in the fact that a GFP reporter is notsensitive enough for this task.

The rate-limiting role of GPDH for glycerol production in S.cerevisiae was confirmed by this analysis, at least in the range ofGPDH activities up to twice the wild-type level (Fig. 4). How-ever, the result of our detailed analysis of GPD1 control on theglycerol pathway illustrates that the regime in which this lineardependence is observed does not extend to strong overexpres-sion. As a result, the correlation between glycerol productionand GPDH activity saturates at levels below the value of mul-ticopy strong overexpression. Moreover, this multicopy over-expression resulted in a decreased biomass yield. A negativeimpact on the growth rate associated with the strong overex-pression of GPD1 has been observed in several independentstudies (16, 18–20, 23, 24) and traced back to acetaldehydeaccumulation and net ATP loss. However, interpolation of ourdata (Fig. 4) suggests that a GPDH activity of about 0.5 to 0.6U/mg protein is the optimal enzyme activity level to obtainmaximal glycerol yield without adversely impacting growth. Itcan be postulated that this GPDH activity could be obtained byintegrating two copies of the strongest TEF1 mutant promoter

FIG. 4. Impacts of specific GPDH activity on glycerol and biomassyields in S. cerevisiae. The values obtained by genomic integration offive TEF1 promoter versions (open symbols) and the previously knownvalues of (i) GPD1 wild type, (i) deletion, and (iii) multicopy overex-pression (closed symbols) are shown. Samples for measuring glycerol,biomass, and glucose for the calculation of yields (YX/S) were taken atthe beginning of the fermentation and during logarithmic growth (op-tical densities were between 1.5 and 1.9). Experiments were carried outin shake flask cultures in a synthetic medium containing 2% glucose(YSC Leu�). YP/S, product yield. Error bars represent the standarddeviations for two or more independent experiments.



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

(mutant 6). This information could be useful if overexpressionof GPD1 has to be achieved by an integrative, instead of aplasmid-based, approach.

The promoter collection provided here should be a usefultool for both basic and applied research in metabolic-pathwayanalysis and functional genomics, as highlighted by our inves-tigation of glycerol yield. Finally, it has been noted that thepromoter sequences of the genes coding for the translation andelongation factors are highly conserved among yeasts and evenhyphal fungi (14, 29) and, therefore, are functional acrossspecies borders. In bacteria, a library created for Lactococcuslactis was seen to be functional in E. coli (10). In fungi, it hasbeen shown that the TEF2 promoter of the fungus Ashbyagossypii is functional in S. cerevisiae (33) and the TEF1 pro-moter of the nonconventional yeast Arxula adeninivorans isfunctional in S. cerevisiae and several non-Saccharomycesyeasts of biotechnological interest (31). Therefore, it may bepossible to use our TEF1 promoter collection in other yeastspecies or hyphal fungi as well, a hypothesis which must betested in the future.

ACKNOWLEDGMENTS

This work was supported by the Berliner Programm zur Forderung derChancengleichheit von Frauen in Forschung und Lehre, the DuPont-MITAlliance, and Department of Energy grant DE-FG02-99ER15015.

REFERENCES

1. Albertyn, J., S. Hohmann, J. M. Thevelein, and B. A. Prior. 1994. GPD1,which encodes glycerol-3-phosphate dehydrogenase, is essential for growthunder osmotic stress in Saccharomyces cerevisiae, and its expression is regu-lated by the high-osmolarity glycerol response pathway. Mol. Cell. Biol.14:4135–4144.

2. Alper, H., C. Fischer, E. Nevoigt, and G. Stephanopoulos. 2005. Tuninggenetic control through promoter engineering. Proc. Natl. Acad. Sci. USA102:12678–12683.

3. Bi, X., and J. R. Broach. 1999. UASrpg can function as a heterochromatinboundary element in yeast. Genes Dev. 13:1089–1101.

4. Bjorkqvist, S., R. Ansell, L. Adler, and G. Liden. 1997. Physiological re-sponse to anaerobicity of glycerol-3-phosphate dehydrogenase mutants ofSaccharomyces cerevisiae. Appl. Environ. Microbiol. 63:128–132.

4a.Brachmann, C. B., A. Davies, G. J. Cost, E. Caputo, J. Li, P. Hieter, and J.D. Boeke. 1998. Designer deletion strains derived from Saccharomyces cer-evisiae S288C: a useful set of strains and plasmids for PCR-mediated genedisruption and other applications. Yeast 14:115–132.

5. Cronwright, G. R., J. M. Rohwer, and B. A. Prior. 2002. Metabolic controlanalysis of glycerol synthesis in Saccharomyces cerevisiae. Appl. Environ.Microbiol. 68:4448–4456.

6. Funk, M., R. Niedenthal, D. Mumberg, K. Brinkmann, V. Ronicke, and T.Henkel. 2002. Vector systems for heterologous expression of proteins inSaccharomyces cerevisiae. Methods Enzymol. 350:248–257.

7. Gueldener, U., J. Heinisch, G. J. Koehler, D. Voss, and J. H. Hegemann.2002. A second set of loxP marker cassettes for Cre-mediated multiple geneknockouts in budding yeast. Nucleic Acids Res. 30:e23. [Online.]

8. Guldener, U., S. Heck, T. Fielder, J. Beinhauer, and J. H. Hegemann. 1996.A new efficient gene disruption cassette for repeated use in budding yeast.Nucleic Acids Res. 24:2519–2524.

9. Jensen, P. R., and K. Hammer. 1998. Artificial promoters for metabolicoptimization. Biotechnol. Bioeng. 58:191–195.

10. Jensen, P. R., and K. Hammer. 1998. The sequence of spacers between theconsensus sequences modulates the strength of prokaryotic promoters. Appl.Environ Microbiol. 64:82–87.

11. Jeppsson, M., B. Johansson, P. R. Jensen, B. Hahn-Hagerdal, and M. F.Gorwa-Grauslund. 2003. The level of glucose-6-phosphate dehydrogenaseactivity strongly influences xylose fermentation and inhibitor sensitivity inrecombinant Saccharomyces cerevisiae strains. Yeast 20:1263–1272.

12. Jin, Y. S., H. Ni, J. M. Laplaza, and T. W. Jeffries. 2003. Optimal growth andethanol production from xylose by recombinant Saccharomyces cerevisiaerequire moderate D-xylulokinase activity. Appl. Environ. Microbiol. 69:495–503.

13. Jones, K. L., S. W. Kim, and J. D. Keasling. 2000. Low-copy plasmids canperform as well as or better than high-copy plasmids for metabolic engineer-ing of bacteria. Metab. Eng. 2:328–338.

14. Kitamoto, N., J. Matsui, Y. Kawai, A. Kato, S. Yoshino, K. Ohmiya, and N.Tsukagoshi. 1998. Utilization of the TEF1-alpha gene (TEF1) promoter forexpression of polygalacturonase genes, pgaA and pgaB, in Aspergillus oryzae.Appl. Microbiol. Biotechnol. 50:85–92.

15. Koffas, M. A., G. Y. Jung, and G. Stephanopoulos. 2003. Engineering me-tabolism and product formation in Corynebacterium glutamicum by coordi-nated gene overexpression. Metab. Eng. 5:32–41.

16. Michnick, S., J. L. Roustan, F. Remize, P. Barre, and S. Dequin. 1997.Modulation of glycerol and ethanol yields during alcoholic fermentation inSaccharomyces cerevisiae strains overexpressed or disrupted for GPD1 en-coding glycerol 3-phosphate dehydrogenase. Yeast 13:783–793.

17. Mumberg, D., R. Muller, and M. Funk. 1994. Regulatable promoters ofSaccharomyces cerevisiae: comparison of transcriptional activity and their usefor heterologous expression. Nucleic Acids Res. 22:5767–5768.

18. Nevoigt, E., R. Pilger, E. Mast-Gerlach, U. Schmidt, S. Freihammer, M.Eschenbrenner, L. Garbe, and U. Stahl. 2002. Genetic engineering of brew-ing yeast to reduce the content of ethanol in beer. FEMS Yeast Res. 2:225–232.

19. Nevoigt, E., and U. Stahl. 1996. Reduced pyruvate decarboxylase and in-creased glycerol-3-phosphate dehydrogenase [NAD�] levels enhance glyc-erol production in Saccharomyces cerevisiae. Yeast 12:1331–1337.

20. Nguyen, H. T., A. Dieterich, K. Athenstaedt, N. H. Truong, U. Stahl, and E.Nevoigt. 2004. Engineering of Saccharomyces cerevisiae for the production ofL-glycerol 3-phosphate. Metab. Eng. 6:155–163.

21. Patkar, A., N. Vijayasankaran, D. W. Urry, and F. Srienc. 2002. Flowcytometry as a useful tool for process development: rapid evaluation ofexpression systems. J. Biotechnol. 93:217–229.

22. Raymond, C. K., T. A. Pownder, and S. L. Sexson. 1999. General method forplasmid construction using homologous recombination. BioTechniques 26:134–138, 140–141.

23. Remize, F., L. Barnavon, and S. Dequin. 2001. Glycerol export and glycerol-3-phosphate dehydrogenase, but not glycerol phosphatase, are rate limitingfor glycerol production in Saccharomyces cerevisiae. Metab. Eng. 3:301–312.

24. Remize, F., J. L. Roustan, J. M. Sablayrolles, P. Barre, and S. Dequin. 1999.Glycerol overproduction by engineered Saccharomyces cerevisiae wine yeaststrains leads to substantial changes in by-product formation and to a stim-ulation of fermentation rate in stationary phase. Appl. Environ. Microbiol.65:143–149.

25. Robinson, A. S., J. A. Bockhaus, A. C. Voegler, and K. D. Wittrup. 1996.Reduction of BiP levels decreases heterologous protein secretion in Saccha-romyces cerevisiae. J. Biol. Chem. 271:10017–10022.

26. Schirmaier, F., and P. Philippsen. 1984. Identification of two genes codingfor the translation elongation factor EF-1 alpha of S. cerevisiae. EMBO J.3:3311–3315.

27. Sheff, M. A., and K. S. Thorn. 2004. Optimized cassettes for fluorescentprotein tagging in Saccharomyces cerevisiae. Yeast 21:661–670.

28. Solem, C., and P. R. Jensen. 2002. Modulation of gene expression made easy.Appl. Environ. Microbiol. 68:2397–2403.

29. Steiner, S., and P. Philippsen. 1994. Sequence and promoter analysis of thehighly expressed TEF gene of the filamentous fungus Ashbya gossypii. Mol.Gen. Genet. 242:263–271.

30. Tao, L., R. E. Jackson, and Q. Cheng. 2005. Directed evolution of copynumber of a broad host range plasmid for metabolic engineering. Metab.Eng. 7:10–17.

31. Terentiev, Y., A. H. Pico, E. Boer, T. Wartmann, J. Klabunde, U. Breuer, W.Babel, M. Suckow, G. Gellissen, and G. Kunze. 2004. A wide-range integra-tive yeast expression vector system based on Arxula adeninivorans-derivedelements. J. Ind. Microbiol. Biotechnol. 31:223–228.

32. Verstrepen, K. J., and J. M. Thevelein. 2004. Controlled expression of ho-mologous genes by genomic promoter replacement in the yeast Saccharo-myces cerevisiae. Methods Mol. Biol. 267:259–266.

33. Wach, A., A. Brachat, R. Pohlmann, and P. Philippsen. 1994. New heterol-ogous modules for classical or PCR-based gene disruptions in Saccharomycescerevisiae. Yeast 10:1793–1808.

34. Yan, W., and E. A. Craig. 1999. The glycine-phenylalanine-rich region de-termines the specificity of the yeast Hsp40 Sis1. Mol. Cell. Biol. 19:7751–7758.



http://aem.asm

.org/D

ownloaded from

http://aem.asm.org/

engineering of promoter replacement cassettes for fine ... · efﬁciency dh5 chemically competent...

Documents