. 14: 783–790 (1998)

A Versatile Set of Vectors for Constitutive and

Regulated Gene Expression in Pichia pastoris

IRINA B. SEARS, JAMES O’CONNOR, OLIVIA W. ROSSANESE AND BENJAMIN S. GLICK*

Department of Molecular Genetics and Cell Biology, University of Chicago, 920 East 58th Street, Chicago,IL 60637, U.S.A.

Received 16 October 1997; Accepted 23 January 1998

The budding yeast Pichia pastoris is an attractive system for exploring certain questions in cell biology, butexperimental use of this organism has been limited by a lack of convenient expression vectors. Here we describe a setof compact vectors that should allow for the expression of a wide range of endogenous or foreign genes in P. pastoris.A gene of interest is inserted into a modified pUC19 polylinker; targeted integration into the genome then results instable and uniform expression of this gene. The utility of these vectors was illustrated by expressing the bacterialâ-glucuronidase (GUS) gene. Constitutive GUS expression was obtained with the strong GAP promoter or themoderate YPT1 promoter. The regulatable AOX1 promoter yielded very strong GUS expression in methanol-growncells, negligible expression in glucose-grown cells, and intermediate expression in mannitol-grown cells. GenBankAccession Numbers are: pIB1, AF027958; pIB2, AF027959; pIB3, AF027960; pIB4, AF027961. ? 1998 John Wiley& Sons, Ltd.

Yeast 14: 783–790, 1998.

— Pichia pastoris; expression vectors; gene regulation

INTRODUCTION (Romanos et al., 1992; Cregg et al., 1993; Scoreret al., 1994). Recently, the glyceraldehyde-3-

58th Street, Chicago, IL 60637, USA. Tel: (+1) 773 702 5315;fax: (+1) 773 702 3172; e-mail: bsglick@midway.uchicago.edu

During the past decade, Pichia pastoris has beenexploited by biotechnologists for the high-levelproduction of foreign proteins (Romanos et al.,1992; Cregg et al., 1993). P. pastoris can utilizemethanol as a carbon source, and growth onmethanol results in strong induction of the peroxi-somal enzyme alcohol oxidase (Gleeson andSudbery, 1988). The major alcohol oxidase iso-zyme is encoded by the AOX1 gene (Cregg et al.,1989). Vectors containing the AOX1 promoterhave been widely used for the regulated over-production of cytosolic and secreted proteins

*Correspondence to: B. S. Glick, Department of MolecularGenetics and Cell Biology, University of Chicago, 920 East

CCC 0749–503X/98/080783–08 $17.50? 1998 John Wiley & Sons, Ltd.

phosphate dehydrogenase (GAP) promoter hasalso been employed for the strong constitutiveexpression of foreign genes (Waterham et al.,1997).

As an experimental organism, P. pastoris sharesmany of the advantages of Saccharomyces cerevi-siae, including mating and sporulation, transfor-mation with integrating or replicating vectors, andgene replacement by homologous recombination(Cregg et al., 1985; Cregg and Madden, 1987;Gleeson and Sudbery, 1988; Gould et al., 1992;Crane and Gould, 1994). Cell biologists initiallybecame interested in P. pastoris because it providesa convenient system for studying peroxisomebiogenesis (Gould et al., 1992; Liu et al., 1992).The recent major advances in this field have

derived largely from genetic studies of P. pastorisand other budding yeasts (Erdmann et al., 1997).

PCR was carried out with a Perkin-Elmer 2400PCR machine using either Pfu or Taq2000

Plasmid construction

784 . . .

P. pastoris is also being used for a genetic analysisof autophagy (Tuttle and Dunn, 1995). We havechosen P. pastoris as a model system to studythe organization of the transitional endoplasmicreticulum and Golgi apparatus, because unlikeS. cerevisiae, P. pastoris has coherent Golgi stackslocated next to discrete sites of transitional ER(Gould et al., 1992; Glick, 1996; O.W.R andB.S.G., in preparation).

An essential tool for yeast cell biology is vectorsfor the expression of endogenous or foreign genes(Schneider and Guarente, 1991). Expression vec-tors have been developed for P. pastoris (Cregget al., 1993; Scorer et al., 1994; Waterham et al.,1997), but many of them are cumbersome to usebecause of large size and/or a limited number ofrestriction sites for subcloning. Here we describe aset of compact expression vectors that contain apUC19-derived polylinker for the insertion ofcloned genes. Because P. pastoris CEN sequencesare not yet available for generating stable episomalplasmids, we created vectors that can be integratedinto the chromosomal HIS4 locus (Romanos et al.,1992). The targeted integration of these vectors isvery efficient, and correct integrants can be ident-ified by a simple polymerase chain reaction (PCR)assay.

Different vectors were designed to allow foreither constitutive or regulated gene expression.Strong constitutive expression can be obtainedwith the GAP promoter, and moderate constitutiveexpression with the promoter from the YPT1 gene,which encodes a small GTPase involved in secre-tion (Segev et al., 1988). Regulated expression canbe obtained with the AOX1 promoter. In the past,use of the AOX1 promoter has yielded expressionlevels that were either very high (on methanol) orvery low (on glucose or other carbon sources)(Tschopp et al., 1987; Waterham et al., 1997).However, we have found that growing cells onmannitol results in an intermediate level of expres-sion from the AOX1 promoter. Together, theseexpression vectors provide a versatile tool for theexperimental manipulation of P. pastoris.

MATERIALS AND METHODS

General methods

Standard procedures were used for recombinant

DNA manipulations (Ausubel et al., 1995). The

? 1998 John Wiley & Sons, Ltd.

DNA polymerase (Stratagene, La Jolla, CA)according to the manufacturer’s instructions.Automated fluorescent DNA sequencing wasperformed by the dye terminator method using anApplied Biosystems Prism 377 machine. Unlessotherwise indicated, all chemicals were purchasedfrom Sigma (St Louis, MO) or Fisher Scientific(Pittsburgh, PA).

Yeast growth media were as follows. YPD: 1%yeast extract, 2% peptone, 2% glucose, 20 mg/ladenine sulfate, 20 mg/l uracil. SD"His: 0·67%yeast nitrogen base, 2% glucose, complete supple-ment mixture minus histidine (CSM"His; Bio101, Vista, CA). SYD: 0·67% yeast nitrogenbase, 0·05% yeast extract, 0·4 mg/l biotin, 40 mg/larginine hydrochloride, 2% glucose. SYG, SYMand SYN were similar to SYD except thatinstead of glucose they contained 1% glycerol,0·5% methanol or 2% mannitol, respectively.

A modified version of the P. pastoris HIS4 gene(Cregg et al., 1985; Crane and Gould, 1994) wascreated as follows. pYM22, a kind gift of JimCregg, contained a HIS4 gene in which the internalBamHI site had been removed (Cregg andMadden, 1987). This mutant gene was subclonedas an EcoRI-BamHI fragment into pUC19(Yanisch-Perron et al., 1985) to yield pUC19-HIS4. Next, a 612-bp internal fragment of HIS4was amplified by PCR from pYM22 with Pfupolymerase using the primers GCACCCTTACCCCAGAAGTCATCT and GTGCCGGAGGAGTAATCTCCACAA. The amplified fragment wascloned into pUC19-HIS4 that had been digestedwith KpnI and blunt-ended with T4 DNA polym-erase. The resulting plasmid, pUC19-HIS4*,contains two point mutations that eliminatethe KpnI sites in HIS4 without altering thecorresponding amino acids.

pIB1 was constructed as follows. First, thecomplementary oligonucleotides GATCCACTAGTCTCGAGCTGCA and GCTCGAGACTAGTG were annealed, and the resulting double-stranded fragment was inserted into pUC19 thathad been digested with BamHI and PstI. In theresulting plasmid, termed pUC19-SX, the XbaIand SalI sites in the polylinker were replaced withSpeI and XhoI sites, respectively. Next, a 260-bpfragment containing the transcription terminator

. 14: 783–790 (1998)

of P. pastoris AOX1 (Koutz et al., 1989) wasamplified by PCR from pPIC3K (Scorer et al.,

Transformation of P. pastoris by electroporation

Integration of pIB vector constructs into the HIS4locus

785 .

1994) with Pfu polymerase using the primersGCCGTCGCCAAGCTTCTTAGACATGACTGTTCCTCAGTTC and GGGACATGTGTGGGAAATACCAAGAAAAACATC. The PCR prod-uct was digested with HindIII and AflIII andinserted into pUC19-SX that had been cut with thesame enzymes, yielding pUC19-SX-AOXTT.Finally, the modified HIS4 gene from pUC19-HIS4* was excised with BglII, blunt-ended withKlenow enzyme, and subcloned into pUC19-SX-AOXTT that had been digested with SspI, yieldingpIB1. Sequencing confirmed that no undesiredmutations had been introduced into the polylinkeror AOX1 transcription terminator portions ofpIB1.

pIB2 was constructed as follows. The promoterof the P. pastoris GAP gene was excised frompHWO10 (Waterham et al., 1997) by digestingwith BstYI and blunt-ending with Klenow enzyme,then digesting with EcoRI. This fragment wassubcloned into pIB1 that had been digestedwith NdeI, blunt-ended with Klenow enzyme, anddigested with EcoRI.

pIB3 was constructed as follows. A 504-bpfragment containing the promoter of the P. pas-toris YPT1 gene (I.B.S. and B.S.G., unpublisheddata) was amplified by PCR from plasmid DNAwith Pfu polymerase using the primers CGAGGCATACATATGATGAGTCACAATCTGCTTCCA and GTGAATTCGACTGCTATTATCTCTGTGTGTA. The PCR product was digestedwith NdeI and EcoRI and cloned into pIB1 thathad been digested with the same enzymes.

pIB4 was constructed as follows. The promoterof the P. pastoris AOX1 gene was excised frompHIL-D2 (Despreaux and Manning, 1993) bydigesting with SacI and blunt-ending with T4DNA polymerase, then digesting with EcoRI. Thisfragment was subcloned into pIB1 that had beendigested with NdeI, blunt-ended with Klenowenzyme, and digested with EcoRI.

The â-glucuronidase (GUS) gene was excisedfrom pBI101 (Clontech, Palo Alto, CA) by digest-ing with SacI, blunt-ending with T4 DNApolymerase, and digesting with BamHI. This frag-ment was subcloned into pIB2, pIB3, and pIB4after digestion with XhoI, blunt-ending withKlenow enzyme, and digestion with BamHI. Theresulting constructs (pIB2-GUS, pIB3-GUS andpIB4-GUS) were linearized with SalI andintegrated into the HIS4 locus as described below.

? 1998 John Wiley & Sons, Ltd.

This procedure was adapted from the method ofBecker and Guarente (1991). A 50-ml culture ofPPY12, a his4 arg4 auxotrophic strain of P. pas-toris (Gould et al., 1992), was grown in YPD at30)C with good aeration to an OD600 of 1–2.The culture was then supplemented with 1 ml of1·0 -Na+-HEPES, pH 8·0 and 1 ml of 1·0 -DTT, and incubation was continued for anadditional 15 min at 30)C. The cells were trans-ferred to a chilled 50-ml Falcon tube, centrifugedfor 3 min at 2000 g (3000 rpm) in a tabletop cen-trifuge at 4)C, and resuspended in 50 ml ice-colddouble-distilled H2O. This H2O wash step wasrepeated, followed by a wash with 20 ml cold1·0 -sorbitol. Finally, the cells were resuspendedin 200 ìl cold 1·0 -sorbitol. A 40-ìl aliquot ofyeast cells was mixed with 0·1–1 ìg DNA in ¦5 ìlof a low-salt solution. This mixture was transferredto a chilled electroporation cuvette (0·2 cm gap)and pulsed with a Bio-Rad Gene Pulser set at1·5 kV, 25 ìF, 200 ohms (time constant 25 ms).The cell suspension was immediately diluted with1 ml of cold 1·0 -sorbitol and transferred to a1·5-ml centrifuge tube. After centrifugation for1 min at 2000 g (5000 rpm) in a microfuge at roomtemperature, the upper 800 ìl of liquid wasremoved and the cells were gently resuspended inthe remaining volume. Finally, the cells werespread on SD"His plates supplemented with1·0 -sorbitol, and the plates were incubated at30)C for 2–3 days until colonies appeared.

Two ìg of a pIB vector (with or without aninsert in the polylinker) was digested with eitherSalI or StuI, ethanol-precipitated, washed twicewith cold 70% ethanol, and resuspended in 10 ìlH2O. For transformation, 2·5 ìl (0·5 ìg) of linear-ized DNA was used in an electroporation. Sixseparate transformants were colony-purified byrestreaking on SD"His plates. Genomic DNAwas then isolated from each transformant using asmall-scale Easy DNA preparation (Invitrogen,Carlsbad, CA) according to the manufacturer’sinstructions. For each 50-ìl PCR reaction, 0·25 ìl(20·5 ìg) of genomic DNA was added to amixture containing PCR buffer, 200 ì of eachdNTP, 0·5 ì each of the primers GCTGTTAAGGTTCGTATGGAGAAAC (sense) and GTGTAGTCTTGAGAAATTCTGAAG (antisense),

. 14: 783–790 (1998)

and 0·25 ìl of Taq2000 DNA polymerase that hadbeen preincubated with 0·25 ìl of TaqStart anti-

tion buffer, depending upon the amount of GUSpresent. Enzyme activity was measured by a

786 . . .

body (Clontech, Palo Alto, CA). The PCR mix-tures were incubated for 2 min at 94)C, followedby 30 cycles of: denaturation, 10 s at 94)C; anneal-ing, 30 s at 55)C; elongation at 68)C. The elonga-tion time was 40 s per kilobase of expectedamplified product, with an increment of 20 s percycle during cycles 11–30. The reactions wereterminated by a 7-min incubation at 68)C followedby a hold of 4)C. Twenty ìl of each reactionmixture was analysed on an agarose gel. As apositive control, a parallel PCR amplification wasperformed using 10 ng of the starting pIB vector inplace of genomic DNA.

Measurement of promoter activities using GUS as

a reporter gene

RESULTS AND DISCUSSION

Yeast strains transformed with pIB2, pIB2-GUS, pIB3-GUS or pIB4-GUS were grown asprecultures in 5 ml of YPD in a 15-ml Falcon tubefor 24 h at 30)C. These precultures could be storedat 4)C for up to a month. Aliquots of the precul-tures were diluted into 50 ml of medium usingthe following dilutions: SYD, 1:50,000; SYG,1:15,000; SYM, 1:1000; SYN, 1:1000. These cul-tures were grown at 30)C with good aeration forapproximately 20 h to an OD600 of 0·4–0·6. Cellswere harvested by centrifugation for 3 min at1000 g (2000 rpm) in a tabletop centrifuge andresuspended in 2 ml of freshly-made extractionbuffer: 50 m-sodium phosphate, pH 7·0, 10 m-â-mercaptoethanol, 5 m-EDTA, 0·1% TritonX-100, 0·25 m-phenylmethylsulfonyl fluoride,1 ì-pepstatin. The suspension was transferred toa 15-ml Falcon tube and cooled on ice. Acid-washed glass beads (0·5 mm diameter) were addedto give a total volume of 3 ml. The mixture wasvortexed for 30 s at top speed, then cooled on icefor 30 s. This procedure was repeated five times intotal, followed by centrifugation for 3 min at1000 g at 4)C. The supernatant was then centri-fuged for 15 min at 100,000 g (50,000 rpm) in aBeckman Optima TLX ultracentrifuge at 4)C. Theclarified supernatant was divided into 200-ìl aliq-uots, which were frozen in liquid N2 and stored at"80)C. One aliquot was used to measure theprotein concentration using the Bio-Rad ProteinAssay (Bio-Rad, Hercules, CA) with bovine serumalbumin as a standard; protein concentrations inthe cell extracts ranged from 0·3 to 1·1 mg/ml.

For GUS assays, cell extracts were either leftundiluted or else diluted 5- to 300-fold in extrac-

? 1998 John Wiley & Sons, Ltd.

fluorometric assay (Gallagher, 1992). A substratesolution of 2·5 m-4-methylumbelliferyl â--glucuronide was prepared in extraction bufferlacking protease inhibitors. Forty ìl of diluted orundiluted extract was added to 160 ìl of substratesolution, and the tube was transferred to a37)C water bath. After 5, 10 and 15 min, 50-ìlaliquots were transferred to tubes containing 2 ml0·2 -Na2CO3 (stop buffer). A zero-minute timepoint was obtained by adding 10 ìl of extract to atube containing 2 ml of stop buffer and 40 ìl ofsubstrate solution. Fluorescence was measured ona Perkin-Elmer LS 50 luminescence spectrometerusing an excitation wavelength of 365 nm, anemission wavelength of 460 nm and a slit width of5·0 nm. The fluorescence values were translatedinto pmol of 4-methylumbelliferone productbased on the signal obtained with knownamounts of this compound. Enzymatic activitywas then calculated with the KaleidaGraphprogram (Synergy Software, Reading, PA) byfitting the time course results to a straight line anddetermining the slope.

Features of the pIB vectors

Figure 1 shows a diagram of pIB1. This plasmid

was derived from the common cloning vectorpUC19 by (a) inserting a modified P. pastorisHIS4 gene, and (b) placing the transcription ter-minator region from AOX1 immediately down-stream of the polylinker. Because HIS4 containsboth XbaI and SalI sites, these two sites in thepolylinker were replaced with compatible SpeI andXhoI sites, respectively. pIB1 is a useful vectorfor expressing genes containing endogenouspromoters that function in P. pastoris.

The expression vectors pIB2, pIB3 and pIB4were derived from pIB1 by placing the promotersfrom the P. pastoris genes for glyceraldehyde-3-phosphate dehydrogenase (GAP; Waterham et al.,1997), YPT1 (I.B.S. and B.S.G., unpublished data)or alcohol oxidase (AOX1; Koutz et al., 1989),respectively, immediately upstream of thepolylinker (Figure 1; Table 1). A gene inserted intothe polylinker of one of these vectors will beexpressed under control of the adjacent promoter.The presence of the AOX1 transcription termin-ator ensures that the 3*-end of the transcript will beprocessed correctly (Koutz et al., 1989).

. 14: 783–790 (1998)

787 .

Figure 1. Structures of the pIB vectors and sequence of the polylinker. Unique restriction sites are indicated, pIB2, pIB3 andpIB4 were derived from pIB1 by inserting the appropriate promoters between the NdeI and EcoRI sites. In all cases the EcoRIsite was preserved; the NdeI site was preserved in pIB3 but destroyed in pIB2 and pIB4. The SacI site is no longer unique in pIB3,and the HindIII site is no longer unique in pIB4. Note that the SphI recognition sequence contains an ATG triplet, so the 5*-endof an open reading frame must be inserted into pIB2, pIB3 or pIB4 upstream of the SphI site. AOX TT, transcription terminatorfrom P. pastoris AOX1; ORI, ColE1 origin of replication; Ap, â-lactamase gene conferring ampicillin resistance; HIS4, P. pastorisHIS4; P-GAP, promoter from P. pastoris GAP; P-YPT1, promoter from P. pastoris YPT1; P-AOX1, promoter from P. pastorisAOX1.

? 1998 John Wiley & Sons, Ltd. . 14: 783–790 (1998)

Table 1. Properties of the pIB vectors

788 . . .

For integration into the P. pastoris genome, aconstruct is linearized by digestion within theHIS4 locus (Rothstein, 1991; Romanos et al.,1992). The SalI and StuI sites are available for thispurpose. (If a gene of interest contains both SalIand StuI sites, it should still be possible to generatea linearized plasmid by partial digestion withone of these enzymes.) The linearized plasmid istransformed into a histidine auxotroph that con-tains point mutations in the chromosomal HIS4gene. Transformants are then selected on plateslacking histidine.

Assessment of correct integration

If a linearized pIB construct integrates into the

P. pastoris genome outside of the HIS4 locus, thestrain will remain unable to grow in the absence ofhistidine because the plasmid-borne HIS4 gene hasbeen interrupted. Only integrants that insert cor-rectly will regenerate a functional HIS4 gene.However, even in the absence of integration, afunctional chromosomal copy of HIS4 could begenerated by a gene conversion event in which theplasmid-borne sequences act as donor (Rothstein,1991). To determine whether a transformant hascorrectly integrated the pIB construct, we preparegenomic DNA and perform a PCR reation, using asense primer that hybridizes near the 3*-end ofHIS4 and an antisense primer that hybridizes nearthe 5*-end of HIS4 (see Materials and Methods).An amplified product will be obtained only if anintegration event has generated two nearby copiesof HIS4. The size of this product will be approxi-mately 3 kb (the size of the pIB vector minus thesize of HIS4) plus the size of the gene inserted intothe polylinker. Using this method, we have readilydetected integration with inserted genes as large as6·5 kb (not shown).

In our experience using different pIB vectors andvarious inserted genes, we consistently find that

? 1998 John Wiley & Sons, Ltd.

about 50% of the transformants contain correctintegrations (not shown). Thus, analysis of sixindependent transformants normally yields at leastone positive clone.

Because integration of a pIB construct generatestwo copies of HIS4, the integrated plasmid mightoccasionally be excised from the chromosome viahomologous recombination (Rothstein, 1991;Romanos et al., 1992). In practice, however, wehave not observed such instability. For example,when a strain expressing an epitope-tagged proteinwas examined by immunofluorescence, the cellsexhibited uniform staining, even if the culture hadbeen grown in non-selective YPD medium (notshown).

Measurement of promoter activities

Plasmidname

Size(bp)

GenBankaccession no. Promoter

Promotercharacteristics

pIB1 5270 AF027958 (None)pIB2 5546 AF027959 GAP Strong constitutivepIB3 5565 AF027960 YPT1 Moderate constitutivepIB4 5796 AF027961 AOX1 Very strong on methanol; strong on mannitol;

weak on glycerol; off on glucose

The â-glucuronidase (GUS) gene from Escher-ichia coli has proven to be a useful reporter formeasuring promoter activity in a variety of organ-isms (Gallagher, 1992), including the buddingyeasts S. cerevisae and Yarrowia lipolytica(Marathe and McEwen, 1995; Bauer et al., 1993).Neither of these yeasts shows significant levels ofendogenous GUS activity. Similarly, we observeda very low background signal in GUS assays withextracts from pIB2-transformed P. pastoris cells(Table 2 legend). However, when a pIB2-GUSconstruct was integrated into the P. pastorisgenome, enzyme activity was observed (Table 2).The specific activity of GUS expressed from theGAP promoter in P. pastoris was comparable tothe activity reported for galactose-grown S. cerevi-siae cells expressing GUS from the strong GAL1promoter (Marathe and McEwen, 1995). WhenGUS was expressed in P. pastoris from the YPT1promoter present in pIB3, a weaker but still sig-nificant enzyme activity was detected (Table 2).Expression levels from the YPT1 promoter wereabout 10- to 100-fold lower than those from the

. 14: 783–790 (1998)

biotechnology, such strong expression is often adisadvantage for cell biology experiments. In S.

ACKNOWLEDGEMENTS

Table 2. Activities of the various promoters in cells

789 .

GAP promoter. Therefore, pIB3 should be usefulfor expressing genes that would be toxic whenoverexpressed.

As previously reported (Waterham et al., 1997),the GAP promoter was constitutively active onvarious carbon sources (Table 2). The YPT1 pro-moter was also constitutively active (Table 2), aswould be expected given the important role ofYpt1p in the operation of the secretory pathway(Segev et al., 1988). In contrast, expression fromthe AOX1 promoter present in pIB4 was stronglyregulated by the carbon source. Very high levels ofGUS activity were observed in methanol-growncells, but virtually no activity was detectable inglucose-grown cells (Table 2). These data are con-sistent with previous reports that the AOX1 pro-moter is strongly induced by methanol andrepressed by glucose (Tschopp et al., 1987).In glycerol-grown cells, the AOX1 promoterdrove a low but measurable level of GUSexpression (Table 2). Low-level expression fromthe AOX1 promoter has also been observed onoleic acid-grown cells (Waterham et al., 1997).

Although massive overexpression from themethanol-induced AOX1 promoter is desirable for

? 1998 John Wiley & Sons, Ltd.

cerevisiae the GAL1 promoter is strongly inducedby galactose and repressed by glucose, but inter-mediate expression levels are observed in cellsgrown on raffinose (Johnston et al., 1994). Simi-larly, when P. pastoris cells were grown on manni-tol, GUS expression from the AOX1 promoterwas approximately 30-fold lower than inmethanol-grown cells (Table 2). Thus, mannitolappears to be a useful carbon source for obtainingan intermediate level of expression from the AOX1promoter.

It was recently reported that the GAP promoteris similar in strength to the methanol-inducedAOX1 promoter (Waterham et al., 1997). How-ever, we found that GUS expression from theAOX1 promoter in methanol-grown cells wasabout eight-fold higher than expression from theGAP promoter in glucose-grown cells (Table 2).This result fits with the observation that alcoholoxidase can comprise up to 30% of the totalprotein in methanol-grown P. pastoris cells(Couderc and Baratti, 1980). The differencebetween our data and those from the earlier studymay be due to the use of different reporter genes(GUS versus â-lactamase), or may reflect theuse of a different polylinker for inserting clonedgenes.

In conclusion, we have created a series ofintegrating vectors that allow for stable, uniformexpression of cloned genes in P. pastoris. By choos-ing the appropriate promoter and carbon source,one can obtain various levels of constitutive orregulated expression. Promoter activities can bequantified using the GUS reporter gene. Thesetechniques will facilitate the use of P. pastoris as amodel organism for cell biology.

The set of four pIB vectors is available free ofcharge upon request.

grown on different carbon sources

Integrated plasmidCarbonsource

GUSactivity

pIB2-GUS (GAP promoter) GlucoseGlycerolMethanolMannitol

70·448·811·318·8

pIB3-GUS (YPT1 promoter) GlucoseGlycerolMethanolMannitol

0·630·840·451·67

pIB4-GUS (AOX1 promoter) GlucoseGlycerolMethanolMannitol

0·050·34

587·620·3

As described in Materials and Methods, extracts from cellstransformed with the indicated plasmids were assayed for GUSactivity, which is given in units of pmol 4-methylumbelliferoneproduced per min per ìg cellular protein. Activities representthe means from two separate experiments. To measure back-ground GUS activity in the cells, extracts were prepared from astrain transformed with the empty pIB2 vector. The followingbackground values were subtracted from the numbers shown:glucose, 0·03; glycerol, 0·02; methanol, 0·08; mannitol, 0·02.

Thanks to Jim Cregg for plasmids and advice, andto members of the Glick lab for critical reading ofthe manuscript. B.S.G. was supported by anAmerican Cancer Society Institutional ResearchGrant and by grants from the Diabetes ResearchFoundation, the March of Dimes Birth DefectsFoundation, the Cancer Research Foundation andthe Pew Scholars Program. O.W.R. was supportedby NIH predoctoral training grant no. 5-20941.

. 14: 783–790 (1998)

REFERENCES yeast Pichia pastoris as a model organism for a geneticand molecular analysis of peroxisome assembly. Yeast

790 . . .

Ausubel, F. M., Brent, R., Kingston, R. E., et al. (Eds)(1995). Short Protocols in Molecular Biology, 3rd edn.Wiley, New York.

Bauer, R., Paltauf, F. and Kohlwein, S. D. (1993).Functional expression of bacterial â-glucuronidaseand its use as a reporter system in the yeast Yarrowialipolytica. Yeast 9, 71–75.

Becker, D. M. and Guarente, L. (1991). High-efficiencytransformation of yeast by electroporation. MethodsEnzymol. 194, 182–187.

Couderc, R. and Baratti, J. (1980). Oxidation of metha-nol by the yeast Pichia pastoris: purification andproperties of alcohol oxidase. Agric. Biol. Chem. 44,2279–2289.

Crane, D. I. and Gould, S. J. (1994). The Pichia pastorisHIS4 gene: nucleotide sequence, creation of a non-reverting his4 deletion mutant, and development ofHIS4-based replicating and integrating plasmids.Curr. Genet. 26, 443–450.

Cregg, J. M., Barringer, K. J., Hessler, A. Y. andMadden, K. R. (1985). Pichia pastoris as a host systemfor transformations. Mol. Cell. Biol. 5, 3376–3385.

Cregg, J. M. and Madden, K. R. (1987). Developmentof yeast transformation systems and construction ofmethanol-utilization-defective mutants of Pichia pas-toris by gene disruption. In Stewart, G. G., Russell, I.,Klein, R. D. and Hiebsch, R. R. (Eds), BiologicalResearch on Industrial Yeasts. CRC Press, New York,pp. 1–18.

Cregg, J. M., Madden, K. R., Barringer, K. J., Thill,G. P. and Stillman, C. A. (1989). Functional charac-terization of the two alcohol oxidase genes from theyeast Pichia pastoris. Mol. Cell. Biol. 9, 1316–1323.

Cregg, J. M., Vedvick, T. S. and Raschke, W. C. (1993).Recent advances in the expression of foreign genes inPichia pastoris. Bio/Technology 11, 905–910.

Despreaux, C. W. and Manning, R. F. (1993). The dacAgene of Bacillus stearothermophilus coding for-alanine carboxypeptidase: cloning, structure andexpression in Escherichia coli and Pichia pastoris.Gene 131, 35–41.

Erdmann, R., Veenhuis, M. and Kunau, W.-H. (1997).Peroxisomes: organelles at the crossroads. Trends CellBiol. 7, 400–407.

Gallagher, S. R. (Ed.) (1992). GUS Protocols: Usingthe GUS Gene as a Reporter of Gene Expression.Academic Press, New York.

Gleeson, M. A. and Sudbery, P. E. (1988). The methy-lotrophic yeasts. Yeast 4, 1–15.

Glick, B. S. (1996). Cell biology: alternatives to baker’syeast. Curr. Biol. 6, 1570–1572.

Gould, S. J., McCollum, D., Spong, A. P., Heyman,J. A. and Subramani, S. (1992). Development of the

? 1998 John Wiley & Sons, Ltd.

8, 613–628.Johnston, M., Flick, J. S. and Pexton, T. (1994). Mul-

tiple mechanisms provide rapid and stringent glucoserepression of GAL gene expression in Saccharomycescerevisiae. Mol. Cell Biol. 14, 3834–3841.

Koutz, P., Davis, G. R., Stillman, C., Barringer, K.,Cregg, J. and Thill, G. (1989). Structural comparisonof the Pichia pastoris alcohol oxidase genes. Yeast 5,167–177.

Liu, H., Tan, X., Veenhuis, M., McCollum, D. andCregg, J. M. (1992). An efficient screen forperoxisome-deficient mutants of Pichia pastoris. J.Bacteriol. 174, 4943–4951.

Marathe, S. V. and McEwen, J. E. (1995). Vectors withthe gus reporter gene for identifying and quantitatingpromoter regions in Saccharomyces cerevisiae. Gene154, 105–107.

Romanos, M. A., Scorer, C. A. and Clare, J. J. (1992).Foreign gene expression in yeast: a review. Yeast 8,423–488.

Rothstein, R. (1991). Targeting, disruption, replace-ment, and allele rescue: integrative DNA transforma-tion in yeast. Methods Enzymol. 194, 281–301.

Schneider, J. C. and Guarente, L. (1991). Vectors forexpression of cloned genes in yeast: regulation, over-production, and underproduction. Methods Enzymol.194, 373–388.

Scorer, C. A., Clare, J. J., McCombie, W. R., Romanos,M. A. and Sreekrishna, K. (1994). Rapid selectionusing G418 of high copy number transformants ofPichia pastoris for high-level foreign gene expression.Bio/Technology 12, 181–184.

Segev, N., Mulholland, J. and Botstein, D. (1988). Theyeast GTP-binding YPT1 protein and a mammaliancounterpart are associated with the secretion machin-ery. Cell 52, 915–924.

Tschopp, J. F., Brust, P. F., Cregg, J. M., Stillman,C. A. and Gingeras, T. R. (1987). Expression of thelacZ gene from two methanol-regulated promoters inPichia pastoris. Nucl. Acids Res. 15, 3859–3876.

Tuttle, D. L. and Dunn, W. A., 6Jr (1995). Divergentmodes of autophagy in the methylotrophic yeastPichia pastoris. J. Cell Sci. 108, 25–35.

Waterham, H. R., Digan, M. E., Koutz, P. J., Lair, S. V.and Cregg, J. M. (1997). Isolation of the Pichiapastoris glyceraldehyde-3-phosphate dehydrogenasegene and characterization of its promoter. Gene 186,37–44.

Yanisch-Perron, C., Vieira, J. and Messing, J. (1985).Improved M13 phage cloning vectors and hoststrains: nucleotide sequences of the M13mp18 andpUC19 vectors. Gene 33, 103–119.

. 14: 783–790 (1998)