Positive Selection of Novel Peroxisome Biogenesis-Defective Mutants of the Yeast Pichia pastoris

Corresponding author: James M. Cregg, Department of Biochemistry and Molecular Biology, Oregon Graduate Institute of Science and Technology, P.O. Box 91000, Portland, OR 97291-1000., cregg{at}bmb.ogi.edu (E-mail)

Communicating editor: A. G. HINNEBUSCH

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

We have developed two novel schemes for the direct selection of peroxisome-biogenesis-defective (pex) mutants of the methylotrophic yeast Pichia pastoris. Both schemes take advantage of our observation that methanol-induced pex mutants contain little or no alcohol oxidase (AOX) activity. AOX is a peroxisomal matrix enzyme that catalyzes the first step in the methanol-utilization pathway. One scheme utilizes allyl alcohol, a compound that is not toxic to cells but is oxidized by AOX to acrolein, a compound that is toxic. Exposure of mutagenized populations of AOX-induced cells to allyl alcohol selectively kills AOX-containing cells. However, pex mutants without AOX are able to grow. The second scheme utilizes a P. pastoris strain that is defective in formaldehyde dehydrogenase (FLD), a methanol pathway enzyme required to metabolize formaldehyde, the product of AOX. AOX-induced cells of fld1 strains are sensitive to methanol because of the accumulation of formaldehyde. However, fld1 pex mutants, with little active AOX, do not efficiently oxidize methanol to formaldehyde and therefore are not sensitive to methanol. Using these selections, new pex mutant alleles in previously identified PEX genes have been isolated along with mutants in three previously unidentified PEX groups.

PEROXISOMES are organelles found in virtually all eukaryotic cells and are characterized by the presence of catalase and at least one hydrogen peroxide-generating oxidase (WATERHAM and CREGG 1997 ). The organelles are enclosed by a single membrane and vary dramatically in size, abundance, and enzyme content depending upon the organism, cell type, and environmental conditions. In mammals, peroxisomes are involved in a variety of essential catabolic and anabolic pathways such as the oxidative degradation of long-chain fatty acids, purines, amino acids, and pipecolic acid as well as the biosynthesis of plasmalogens, cholesterol, and bile acid. As demonstrated by Zellweger syndrome and other lethal peroxisomal biogenesis disorders, peroxisomes are indispensable for human survival (SUBRAMANI 1997 ).

Because peroxisomes lack nucleic acids and ribosomes, all peroxisomal proteins must be nuclear encoded. Peroxisomal proteins are synthesized on free polysomes and post-translationally imported into the organelle via peroxisomal targeting signals (PTS) (WATERHAM and CREGG 1997 ). Two classes of matrix protein PTSs have been characterized. The first and most common, PTS1, is composed of a tripeptide sequence (SKL and conservative variants) present at the extreme carboxy terminus of many animal, plant, and yeast peroxisomal matrix proteins (DE HOOP and AB 1992 ; SUBRAMANI 1993 ). The second, less common PTS, PTS2, has a consensus sequence of RLX₅H/QL and is located near the amino terminus of matrix proteins such as 3-ketoacyl-coenzyme A thiolase of mammals (OSUMI et al. 1991 ; RACHUBINSKI and SUBRAMANI 1995 ) and yeast (GLOVER et al. 1994 ). Some matrix proteins appear to have neither targeting signal, suggesting that other matrix protein PTSs have yet to be discovered. In addition to PTSs, specific components of the matrix protein import pathway have been elucidated, including a PTS1 receptor (Pex5p; MCCOLLUM et al. 1993 ), a PTS2 receptor (Pex7p; REHLING et al. 1996 ; ZHANG and LAZAROW 1996 ), and putative PTS receptor docking proteins (Pex13p and Pex14p; ELGERSMA et al. 1996 ; ERDMANN and BLOBEL 1996 ; GOULD et al. 1996 ; ALBERTINI et al. 1997 ; KOMORI et al. 1997 ). Much less is known about the targeting signals of peroxisomal membrane proteins (PMPs). Evidence suggests that PMPs are targeted and inserted by a mechanism that is independent of that for matrix proteins. Recently, an internal hydrophilic 20-amino acid loop was demonstrated to be necessary and sufficient for targeting of a PMP from Candida boidinii (DYER et al. 1996 ).

Certain yeast species have served as productive model systems for investigations of peroxisome biogenesis. In yeasts, peroxisome proliferation is markedly induced by growth on carbon sources such as oleic acid and methanol, making them a convenient source of the organelles and their enzymes for biochemical studies. Furthermore, our lab and others have shown that yeasts require peroxisomes only for metabolism of these carbon sources and not others (e.g., glucose) and that yeast mutants with defects in peroxisome biogenesis (pex mutants) can be found among collections of strains that are specifically defective in growth on methanol and/or oleate (ERDMANN et al. 1997 ; WATERHAM and CREGG 1997 ). These observations have resulted in the isolation of numerous pex mutants in more than a dozen PEX genes in each of four yeast species: Saccharomyces cerevisiae, Hansenula polymorpha, Pichia pastoris, and Yarrowia lipolytica (ERDMANN et al. 1997 ; WATERHAM and CREGG 1997 ). The yeast PEX genes have been isolated and their products characterized to provide important insights into peroxisome biogenesis mechanisms (WATERHAM and CREGG 1997 ). Importantly, because the primary sequences of their products have been conserved through evolution, they have been used to identify their human PEX homologues in the databases, many of which have subsequently been shown to be genes affected in patients with peroxisome biogenesis disorders (SUBRAMANI 1997 ).

Among yeasts, P. pastoris is unique in that it is able to grow on either methanol or oleic acid. With P. pastoris, virtually all mutants specifically defective in growth on these two peroxisome-requiring substrates are pex mutants. Nevertheless, the need to screen ever larger collections of strains for mutants in new PEX genes is laborious and time consuming. More direct means to select for P. pastoris pex mutants have been needed. Here, we describe two highly efficient positive selection schemes for the isolation of P. pastoris pex mutants. Their utilization has resulted in the isolation of a large number of new alleles of previously identified PEX genes, novel mutants in two genes encoding potential transcription factors, and pex mutants in three novel complementation groups.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Strains, media, and microbial techniques:
P. pastoris strains used in this study are listed in Table 1. JC144 was constructed by integration of vector pHW011 into the alcohol oxidase 1 (AOX1) locus of GS200. Cultures (liquid and agar) were grown or induced at 30° in YPD medium [1% (wt/vol) yeast extract, 2% (wt/vol) peptone, 0.4% (wt/vol) glucose] or minimal media containing 0.17% (wt/vol) yeast nitrogen base without amino acids and with ammonium sulfate (Difco Laboratories Inc., Detroit), 0.5% (wt/vol) supplemented with one of the following: 0.5% (vol/vol) methanol (YNM medium); 0.1–0.4% (wt/vol) glucose (YND medium); 0.5% (vol/vol) ethanol (YNE medium); 0.2% (vol/vol) oleate plus 0.02% (vol/vol) Tween 40 and 0.05% (wt/vol) yeast extract (YNO medium for induction experiments); 0.1% (vol/vol) oleate and 0.05% Tween 40 (YNO medium for growth experiments); and 0.4% (wt/vol) glucose plus 0.5 mM allyl alcohol (AAD medium). Methanol-sorbitol medium was 12.0% (vol/vol) methanol, 0.2% (wt/vol) sorbitol, 0.2% (wt/vol) yeast extract, and 0.4% (wt/vol) peptone (MSY medium). Sporulation/mating medium was 0.5% (wt/vol) sodium acetate, 1% (wt/vol) potassium chloride, and 1% (wt/vol) glucose. Alcohol oxidase (AOX) activity assay medium was 50 mM Tris-HCl, pH 8.0, 0.1% (wt/vol) digitonin, 0.04% (wt/vol), 2,2'-azino-bis(3-ethylbenz-thiazoline-6-sulfonic acid (ABTS; Sigma, St. Louis), 0.02% (wt/vol) peroxidase, and 0.5% (vol/vol) methanol. For solid medium, agar was added to 2% (wt/vol). For growth of auxotrophic strains, requisite amino acids were added to a final concentration of 50 µg/ml. Cultivation of Escherichia coli strain DH5 and standard recombinant DNA techniques were performed essentially as described previously (SAMBROOK et al. 1989 ).

Plasmid constructions:
pHW011, an E. coli-P. pastoris shuttle vector capable of expressing the P. pastoris AOX1 gene under control of the P. pastoris glyceraldehyde-3-phosphate dehydrogenase gene promoter (P_GAP), was constructed as follows: (1) The AOX1 gene was amplified from plasmid pPG5.4 (CREGG et al. 1989 ; KOUTZ et al. 1989 ) by the polymerase chain reaction (PCR) using a forward primer composed of the first 18 bases of the AOX1 open reading frame preceded by an EcoRI site (5'-GAATTCATGGCTATCCCCGAAGAG-3') and a 19-base reverse primer complementary to a region located 247–266 bases downstream of the AOX1 stop codon, which includes a genomic HindIII site and AOX1 terminator sequence (5'-AAGCTTGCACAAACGAACG-3'). (2) After digestion with EcoRI and HindIII, the AOX1 gene fragment was subcloned in EcoRI-HindIII-digested pBS-SK^- (Stratagene, La Jolla, CA) and subsequently as an EcoRI-ClaI fragment into EcoRI-ClaI-digested vector pHW010 (WATERHAM et al. 1997 ) to produce vector pHW011. Prior to electrotransformation into P. pastoris strain GS200, pHW011 was linearized at a BglII site in the AOX1 gene.

pOPGP-1, an E. coli-P. pastoris shuttle vector capable of expressing a peroxisomal targeted red-shifted form of the green fluorescent protein (EGFP) under control of the P. pastoris PEX8 gene promoter (P_PEX8), was constructed. As a first step, a P_PEX8 expression vector was made by replacing the AOX1 gene promoter (P_AOX1) fragment in the P. pastoris expression vector pHIL-A1 (Phillips Petroleum, Bartelsville, OK) with a DNA fragment containing P_PEX8. The P_PEX8 fragment was obtained from vector pYT4 (LIU et al. 1995 ) by digestion with BglII. The termini of the BglII fragment were made blunt ended by treatment with the Klenow fragment of DNA polymerase I and the fragment was inserted into SmaI- and EcoRV-digested pBS-SKII^- (Stratagene). The resulting plasmid was then digested with EcoRI, Klenow treated, and ligated to remove an EcoRI site within the P_PEX8 fragment. This plasmid was then used in PCR to produce a P_PEX8 fragment that contained the BamHI site of pBS-SKII^- at its 5' terminus and introduced an EcoRI site on its 3' terminus using the oligonucleotides 5'-AACAGCTATGACCATG-3' and 5'-CAGGAATTCTAACAGGCACCTGAAGATAGGT-3'. The PCR product was digested with BamHI and EcoRI and inserted into BglII- and EcoRI-digested pHIL-A1 to produce the PEX8 promoter-driven expression vector pK312. To generate an EGFP gene whose product is targeted to peroxisomes, two complementary oligonucleotides were first synthesized, which formed an adapter fragment encoding the last nine amino acids and stop codon of Pex8p (including the PTS1 sequence AKL) with a HindIII site at the 3' terminus and a SalI site at the 5' terminus (5'-TCGACGCCCAATCAACCGCAAAGTTATAAACCGGTA-3' and 5'-AGCTTACCGGTTTATAACTTTGCGGTTGATTGGGCG-3'). The adapter was inserted into HindIII- and XhoI-digested pEGFP-C3 (Clontech Laboratories, Inc., Palo Alto, CA), which resulted in a chimeric gene encoding EGFP with the last nine amino acids of Pex8p fused in frame to its carboxyl terminus (EGFP-PTS1). A PCR was performed that amplified the EGFP-PTS1 fusion with an EcoRI site immediately 5' of EGFP and included an EcoRI site located 3' of the fusion gene using oligonucleotides 5'-CGGAATTCATGGTGAGCAAGGGCGAGGAG-3' and 5'-CAGAATTCGAAGCTTACCGGTTTATAACTTTGCGG-3'. The EGFP-PTS1 fragment was then digested with EcoRI and inserted into the unique EcoRI site of pK312 to produce pOPGP-1. The PTS2-EGFP vector, pTW65, expresses the chimeric protein under control of the P. pastoris acyl-CoA oxidase promoter; it was a gift from Dr. Suresh Subramani (University of California, San Diego, La Jolla, CA) and was described in ELGERSMA et al. 1998 . A vector that encoded a chimeric protein composed of the 460 (out of 461) N-terminal-most amino acids of P. pastoris Pex2p fused to the N terminus of EGFP and expressed under control of P_AOX1 was constructed as follows. First, a 750-bp NheI-HindIII fragment containing the EGFP gene from plasmid pEGFP-C3 (Clontech) was ligated into XbaI- and HindIII-digested pUC18 to create pUC18-EGFP. Second, a DNA fragment encoding amino acids 1–460 of Pex2p was generated by PCR using oligonucleotides 5'-TCAGGATCCATGCCCAATAGGCTCATACC-3' and 5'-GACCATGGCACCAACGAAAAAACCTGG-3' as primers. This resulted in a fragment with a BamHI site at the 5' terminus and an NcoI site at the 3' terminus. The PEX2 fragment was digested with these enzymes and inserted into the same sites in pUC18-EGFP to create pUC18-PEX2-EGFP. Third, this vector was digested with SmaI and XhoI, and an ~2-kb fragment containing the PEX2-EGFP fusion was ligated into PmlI- and XhoI-digested pPICZB (Invitrogen, Carlsbad, CA) to create pLC303. The vector was linearized within the AOX1 promoter region by digestion with PmeI prior to transformation into P. pastoris. The luciferase gene vector pHWO17 expresses the gene under control of P_AOX1 and was described in WATERHAM et al. 1996 .

Mutagenesis:
The procedure for mutagenesis with N-methyl-N'-nitro-N-nitrosoguanidine (NTG) was as described by CREGG et al. 1990 with the following modifications: (1) Cells were cultured in YND plus arginine; (2) NTG was dissolved in acetone (10 mg NTG/ml acetone) and used fresh; and (3) prior to mutagenesis, cultures were divided into three equal portions and aliquots treated with 0.02, 0.04, or 0.08 mg NTG/ml of cells. These treatments resulted in the killing of 67, 90, and 96% of cells, respectively. Glycerol (to 30%) was added to mutagenized cultures and cultures were frozen at -70° until use. Freezing killed an additional 90% of cells surviving mutagenesis.

For ultraviolet-light (UV) mutagenesis, the following procedure was followed: (1) Cells were pregrown on YPD medium to an OD₆₀₀ of ~1. (2) Cells were harvested and resuspended in sterile water at an OD₆₀₀ of 0.3. (3) Twenty milliliters of the culture in water were transferred to a petri dish and irradiated for 30–40 sec with gentle shaking (this UV treatment resulted in the death of 90–99% of the cells). (4) Under dim light to minimize photoreactivation repair, 100-µl aliquots of culture were spread on agar plates.

Preparation of cell-free and whole-cell extracts:
To prepare cell-free extracts, cells were precultured in YPD and harvested at ~1 OD₆₀₀. For mutant strains, 50 OD₆₀₀ units were resuspended in 50 ml of either YNM or YNO medium and induced for 6 hr at 30°. For wild-type strains, 25 OD₆₀₀ units were resuspended in 50 ml and processed as described for mutant strains. Cells were harvested and washed twice with ice-cold 50 mM potassium phosphate buffer, pH 7.0, and then frozen at -20°. Cells were thawed and resuspended in 400 µl of the same buffer along with 0.5 µl of 1 M phenylmethanesulfonyl fluoride (PMSF). Cold 0.5-mm-diameter acid-washed glass beads were then added to ~1/3 volume of buffer, and the mixture vortexed at 4° at high speed for 10 min, followed by a 20-min centrifugation in a minicentrifuge at 14,000 rpm and 4°. The supernatant was removed and stored on ice. For whole-cell extracts, the postdisruption centrifugation step was omitted, and the supernatant was removed after the beads had settled.

Subcellular fractionation:
Cells were pregrown in YPD medium and transferred during logarithmic growth phase (1–1.5 OD₆₀₀) by centrifugation into YNO or YNM medium and induced for 6 hr at 30°. Subcellular fractionations were performed as described by LIU et al. 1992 with the following modifications: (1) 10 mM Na₂SO₃ was used in place of dithiothreitol (DTT). (2) Incubation with Zymolyase was for 15–45 min. (3) Immediately following treatment with Zymolyase, 10 µl of 1 M PMSF was added. (4) Following the collection of protoplasts by centrifugation, a wash using 5 mM 3-[N-morpholino]propanesulfonic acid, 0.5 M KCl, 10 mM Na₂SO₃, pH 7.2 buffer was done to remove Zymolyase. (5) MES buffer was composed of 5 mM MES, pH 6.0, 1.2 M sorbitol, 0.5 mM EDTA, 0.1% ethanol, 0.21 mg/ml NaF, 1 mM PMSF, 1 mM leupeptin, and 1 mM aprotinin. The final centrifugation was performed at 30,000 x g for 30 min.

Biochemical methods:
Peroxisomal AOX (VERDUYN et al. 1984 ; VAN DER KLEI et al. 1990 ), catalase (CAT; UEDA et al. 1990 ), acyl-CoA oxidase (DOMMES et al. 1981 ), mitochondrial cytochrome c oxidase (DOUMA et al. 1985 ), and ß-lactamase (WATERHAM et al. 1997 ) activities were assayed at 30° according to published procedures. Activities were expressed as micromoles of product/min/mg of protein for AOX, cytochrome c oxidase, acyl-CoA oxidase, and ß-lactamase. Activities for catalase were expressed as E₂₄₀/min/mg of protein. Protein concentrations were determined using the Pierce (Rockford, IL) Bicinchoninic acid protein assay kit with bovine serum albumin as a standard. Transfer of protein to nitrocellulose filters after sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) was performed using a Mini trans-blot electophoretic transfer cell (Bio-Rad Laboratories, Richmond, CA) as indicated by the manufacturer. Immunoblotting experiments were performed with specific polyclonal antibodies to AOX, catalase, or thiolase (a gift from W. H. Kunau, Ruhr University, Bochum, Germany) using either the Western Light Kit (Tropix, Bedford, MA) or the color development assay using 5-bromo-4-chloro-3-indolyl phosphate (BCIP) and p-nitroblue tetrazolium chloride (NBT; Bio-Rad Laboratories).

Fluorescence microscopy:
Strains transformed with either pOPGP-1 or pLC303 were grown overnight in YND (0.1% glucose) plus 0.5% glycerol to 1 OD₆₀₀, inoculated into YNM to a starting OD₆₀₀ of 1.0, and grown at 30° with shaking for 4–6 hr. Strains transformed with pTW65 were precultured in YPD and grown to 1 OD₆₀₀, inoculated into YNO for induction at a starting OD₆₀₀ of 1, and grown for 12 hr at 30° with shaking. Slides were prepared by adding 10 µl of culture to a slide and affixing the coverslip with rubber cement.

Genetic methods:
Complementation testing was done as described by CREGG et al. 1998 . Procedures for backcrossing and random spore analysis were as described in LIU et al. 1992 . To verify that spore products of the first backcrossed formaldehyde dehydrogenase (fld1)-derived strains were wild type with respect to the FLD1 gene, the progeny were crossed with an fld1 strain and replica plated to YNM plates. If complementation occurred, i.e., growth on methanol, the backcrossed strain did acquire the FLD1 gene during the first backcross.

Miscellaneous methods:
Adobe Photoshop (Adobe Systems Inc., Mountain View, CA) was used for scanning photographs of plates and X-ray films developed from immunoblots. The scans were then imported into Freehand (Macromedia Inc., San Francisco) to add text. Adobe Photoshop was also used for scanning EGFP expression negatives. These scans were then imported into Adobe PageMaker to arrange and add text. Electron microscopy was performed as described in WATERHAM et al. 1996 .

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

Selection for pex mutants using allyl alcohol:
Allyl alcohol is a nontoxic substrate that is oxidized by AOX, a peroxisomal matrix enzyme and the first enzyme in the yeast methanol metabolic pathway, to acrolein, a substance that is toxic (Figure 1; CIRIACY 1975 ; SIBIRNY et al. 1988 , SIBIRNY et al. 1989 ). Thus, methanol-grown wild-type P. pastoris cells with high levels of AOX are sensitive to allyl alcohol because of its conversion to acrolein. In previous studies, we showed that, in methanol-induced P. pastoris pex mutants that lack functional peroxisomes, AOX is synthesized but does not assemble and remains in the cytoplasm as misfolded inactive aggregates (LIU et al. 1992 ; WATERHAM et al. 1997 ). Thus, methanol-induced pex mutants should be relatively insensitive to allyl alcohol, providing a potential means of selectively growing pex mutants in mutagenized cultures. However, AOX is typically present only in P. pastoris cells grown or induced on methanol, and pex mutants cannot grow on methanol. To circumvent this problem, a P. pastoris strain was constructed that expresses AOX on glucose medium. This strain, JC144, was made by transforming P. pastoris GS200 with pHW011, a vector that expresses the P. pastoris AOX1 gene under the control of the constitutive P_GAP. In glucose-grown cells of both wild-type and pex strains transformed with this plasmid, AOX protein was present. However, in JC144, AOX was active, while in pex mutants it was not. The levels of AOX activity in JC144 were ~100-fold lower than in methanol-grown wild-type P. pastoris but more than 100-fold higher than those typically seen in glucose-grown cells of this yeast; subcellular fractionation studies showed that the AOX activity was localized to peroxisomes (data not shown). Importantly, JC144 was sensitive to 0.5 mM allyl alcohol on allyl alcohol dextrose (AAD) plates, while P. pastoris pex mutants transformed with pHW011 were not (Figure 2, A and B). Thus, using JC144 as the parent strain for mutagenesis and AAD plates, it was possible to select for pex mutants and against the parent strain and other types of mutants that continued to express and assemble active AOX.

View larger version (10K): In this window In a new window Download PPT slide   Figure 1. Metabolism of methanol and allyl alcohol in yeast. AOX, alcohol oxidase; CAT, catalase; FLD, formaldehyde dehydrogenase; FMD, formate dehydrogenase.

View larger version (102K): In this window In a new window Download PPT slide   Figure 2. Allyl alcohol selection scheme for the isolation of P. pastoris pex mutants. (A) The parent strain constitutively expresses AOX1 from PGAP. Wild-type cells die as a result of the active AOX converting allyl alcohol to acrolein. Pex mutants grow due to the absence of active AOX. (B) Growth of strains expressing AOX on allyl alcohol-glucose plates. All strains were transformed with pHWO11 and then spread on an allyl alcohol-glucose plate.

To isolate P. pastoris pex mutants, cultures of strain JC144 were subjected to mutagenesis with NTG and spread on AAD plates. The optimal concentration of mutagenized cells was empirically determined to be ~2 x 10⁶ cells/ml. Higher concentrations resulted in fewer colonies, possibly due to a "neighbor effect" (i.e., killing of cells by diffusion of acrolein from active AOX-containing cells to those without active AOX). The resulting colonies were further screened as described in LIU et al. 1992 . Briefly, 2951 allyl alcohol-resistant colonies were tested first for methanol utilization (Mut) by replica plating. The resulting Mut^- were then tested for oleic acid utilization (Out) in liquid oleic acid medium. Thirty-eight colonies were identified that were defective in growth on both of these peroxisome-requiring substrates. Complementation tests demonstrated that 33 of the Mut^- Out^- belonged to previously identified PEX gene groups (Table 2). As described below, the remaining 5 Mut^- Out^- mutants represented two non-PEX complementation groups that we have tentatively named MXR1 and MXR2 and that appear to encode transcription factors required for expression of certain peroxisomal enzymes.

Efficiency of the allyl alcohol selection scheme:
To determine the efficiency of the allyl alcohol selection scheme, samples of the same NTG-mutagenized cultures of JC144 were subjected to selection on AAD plates and to the random negative screening method (GOULD et al. 1992 ; LIU et al. 1992 ). Under optimal conditions, 1.2% of the colony population on the AAD plates were pex mutants, whereas only ~0.09% of the cells surviving mutagenesis were pex mutants in the population prior to selection. Thus, the allyl alcohol selection provided an ~13-fold enrichment in pex mutants.

Selection of pex mutants using an fld1 strain and high methanol:
In wild-type P. pastoris cells growing on methanol, formaldehyde, the product of AOX oxidation of methanol, is further oxidized to formate and then carbon dioxide by two nicotinamide adenine dinucleotide-dependent dehydrogenases (VEENHUIS and HARDER 1987 ; SHEN et al. 1998 ; Figure 1). However, in strains defective in FLD, toxic formaldehyde is not further metabolized and accumulates, killing these cells. The build-up of formaldehyde does not occur in methanol-induced fld1 pex mutants, because these strains have little active AOX and, therefore, cannot convert methanol to formaldehyde. Thus, fld1 pex mutants survive exposure to methanol. However, neither fld1 nor pex mutants grow on methanol. As described above, most carbon sources other than methanol repress methanol metabolic pathway enzyme synthesis. An exception is sorbitol. In medium containing a mixture of 12% methanol and 0.2% sorbitol (MSY), wild-type and fld1 strains contained methanol pathway enzymes at levels similar to those observed when methanol is the sole carbon source. Thus sorbitol could be used as a carbon source that would allow for the selective growth of fld1 pex strains (Figure 3A). To demonstrate this pex mutant selection scheme, Figure 3B shows an MSY plate on which selected strains had been streaked. As expected, wild-type (not shown) and fld1 pex strains grew on this medium while the fld1 strain did not.

View larger version (121K): In this window In a new window Download PPT slide   Figure 3. Methanol selection scheme for the isolation of P. pastoris pex mutants using an fld1 mutant parent strain. (A) fld1 nonpex cells die due to accumulation of formaldehyde while fld1 pex mutants, which lack AOX, accumulate little or no formaldehyde and grow. (B) Growth of fld1 and pex fld1 strains on high methanol-sorbitol plates.

To select for pex mutants using this scheme, UV- or NTG-mutagenized cultures of the fld1 strain were spread on MSY plates. After incubation, the resulting large colonies were streaked onto YPD plates, incubated at 30° for 24 hr, and then replica plated onto a second MSY plate to confirm their ability to grow on the methanol-sorbitol medium. Strains able to grow on MSY plates were further tested for their ability to grow on oleate, and strains that were Out^- were collected for further study.

Efficiency of the fld1/high-methanol selection scheme:
A total of 582 methanol-resistant strains were examined for Out phenotype and 293 (50%) were Out^-. A sample of 149 of the methanol-resistant and Out^- mutants was subjected to complementation analysis and found to represent new alleles of each of the 10 previously identified PEX gene groups plus 3 new complementation groups. As described below, these new groups were determined to be defective in previously unidentified PEX genes (Table 2). Thus, at least 25% of strains arising from the selection procedure were pex mutants. This represented an ~278-fold enrichment over the percentage of pex mutants in the nonenriched control cell population.

To determine whether the fld1 mutant strain was required for the high-methanol scheme to work, we also performed the selection using the same conditions except with P. pastoris strains that were wild type with respect to their ability to use methanol [GS115 (his4) and GS190 (arg4)]. Out of 340 mutants derived from these two strains that were resistant to the high methanol concentration, we identified 11 pex mutants representing six different PEX gene complementation groups (Table 2). Thus, at 3.2%, the high-methanol selection scheme generated a strong (~35-fold) enrichment for pex mutants even without the use of the fld1 mutant strain. However, the fld1 strain resulted in a further 8-fold enrichment for pex mutants.

Genetic analysis of mutants in new complementation groups:
Isolated mutants (described above) that did not fall into prevously identified PEX gene groups—including eight reported by our group in LIU et al. 1992 and two additional groups unique to a collection described by GOULD et al. 1992 —were first backcrossed at least three times. Complementation studies with the backcrossed derivatives revealed that the strains represented recessive alleles in five new complementation groups.

Biochemical and morphological characterization of the new mutant groups:
Representatives of each new group, along with wild-type and pex1 control strains, were examined for the presence, activity, and subcellular location of selected peroxisomal enzymes. In total-cell extracts prepared from methanol-induced cells of three of the mutants (temporarily named pexA, pexB, and pexC), near wild-type levels of catalase (CAT) activity were observed (Table 3). Conversely, AOX activity, which is typically low or absent in pex mutants (LIU et al. 1992 ; WATERHAM and CREGG 1997 ), was also low in these strains. In immunoblots prepared from these strains, AOX protein was also low in two of the strains (pexA and pexB) but nearly normal in pexC (data not shown). When induced on oleate medium, CAT activity and thiolase protein were near wild-type levels, while acyl-CoA oxidase activity was present at low but significant levels. All of these results were characteristic of previously isolated pex strains (GOULD et al. 1992 ; LIU et al. 1992 ).

Results of subcellular fractionations with each of the three strains were also typical of those observed for other pex mutants. Methanol- and oleate-induced cells of each of the three strains were subjected to subcellular fractionation into a pellet fraction containing mostly mitochondria and peroxisomes, and a cytosolic supernatant fraction. Whereas ~50% of CAT activity was present in the organellar pellet from wild-type cells, most CAT activity from the pexA, pexB, pexC, and pex1 strains was present in the supernatant fraction (Table 4).

Methanol- and oleate-induced cells of pexA, pexB, and pexC were also examined by electron microscopy for peroxisomes and no normal peroxisomes were observed (Figure 4; pexB not shown), confirming that each of these strains represents mutants in new PEX genes.

View larger version (127K): In this window In a new window Download PPT slide   Figure 4. Electron micrographs showing subcellular morphology of selected P. pastoris strains. (A) Proliferation of small peroxisomes in oleate-induced wild-type cells. (B) Proliferation of large peroxisomes in methanol-induced wild-type cells. (C) Lack of recognizable peroxisomes in pexA oleate-induced cells. (D) Lack of recognizable peroxisomes in pexC methanol-induced cells (peroxisomal remnants indicated by arrow). (E) Peroxisomes (indicated by arrows) in oleate-induced mxr1. (F) Lack of recognizable peroxisomes in methanol-induced mxr1. P, peroxisome; N, nucleus; V, vacuole; L, lipid body. Bar, 1.0 µM.

The three new pex mutant groups were further characterized with regard to the function of their peroxisomal targeting signal (PTS) pathways. Three PTS pathways have been defined in yeasts and other eukaryotes: two, PTS1 and PTS2, are specific to peroxisomal matrix proteins, and the third, mPTS, is specific to peroxisomal integral membrane proteins (WATERHAM and CREGG 1997 ). To examine each pathway, the fate of selected proteins with well-characterized targeting signals was determined. For the PTS1 pathway, we expressed two heterologous PTS1 proteins, luciferase and an EGFP-PTS1 chimeric protein (GOULD et al. 1987 ; KALISH et al. 1996 ; MONOSOV et al. 1996 ). After subcellular fractionation of oleate-induced cells of each of the new pex mutants, luciferase activity was found predominantly in the supernatant fractions (Table 4), whereas in wild-type cells ~50% of luciferase activity was present in the pellet fraction. These results were similar to those seen for CAT activity and indicated that the PTS1 pathway was defective in all three pex mutants. Further support for this conclusion was obtained by fluorescence microscopy of EGFP-PTS1-expressing methanol-induced cells of the mutants. In wild-type cells, a tight cluster of strongly fluorescing peroxisomes was observed (Figure 5A). However, in each of the pex mutants the entire cytoplasm fluoresced (Figure 5D, G, J, and M). To investigate PTS2 pathway function, the fates of two PTS2 proteins, P. pastoris thiolase and a PTS2-EGFP chimeric protein, were examined in each mutant (GLOVER et al. 1994 ; ELGERSMA et al. 1998 ). Subcellular fractions from oleate-induced cells of each mutant showed thiolase protein predominantly in the supernatant fractions, whereas the majority of thiolase protein was present in the pellet fraction from wild-type cells (Figure 6). Furthermore, fluorescence microscopy of PTS2-EGFP-expressing oleate-induced cells of each pex mutant showed a diffuse cytoplasmic fluorescence pattern typical of PTS2 pathway-defective pex mutants (Figure 5E, Figure H, Figure K, and N; ELGERSMA et al. 1998 ). Thus, both PTS1- and PTS2-matrix protein import pathways are defective in all three new pex mutant groups.

View larger version (59K): In this window In a new window Download PPT slide   Figure 5. Subcellular location of EGFP-PTS1, PTS2-EGFP, and mPTS-EGFP in new pex mutants. (A, D, G, J, M) Cells expressing PTS1-EGFP on methanol. Note the clusters of strongly fluorescing peroxisomes in wild type (A) vs. the cytosolic fluorescence in the mutants pexA (D), pexB (G), pexC (J), pex1 (M). (B, E, H, K, N) Cells expressing PTS2-EGFP on oleate. As in EGFP-PTS1, the wild-type cells (B) exhibit a punctate pattern—although less pronounced than A—vs. the diffuse pattern seen in the mutants pexA (E), pexB (H), pexC (K), and pex1 (N). (C, F, I, L, O) Cells expressing mPTS-EGFP on methanol. (C) Wild type exhibits localization of fluorescence to the peroxisomal membranes seen as rings. Arrow denotes a cell in which four such rings are visible. Mutants (F) pexA, (L) pexC, and (O) pex1 show fluorescence localized to what appears to be peroxisomal remnants while (I) pexB shows both cytosolic and peroxisomal localization.

View larger version (27K): In this window In a new window Download PPT slide   Figure 6. Subcellular localization of thiolase and catalase protein. Organelle pellet and cytosolic supernatant fractions obtained after subcellular fractionation of wild type, pexA, pexB, pexC, and pex1, induced on oleate, were analyzed by immunoblots with antibodies against thiolase and catalase. For thiolase, 20 µg of total protein was loaded in the wild-type lane while 40 µg of total protein was loaded in each of the other lanes. For catalase, 20 µg of total protein was loaded in each lane.

To examine mPTS pathway function, we expressed a Pex2p-EGFP chimeric protein in each of the new pex mutants. Pex2p is an integral membrane protein in P. pastoris and other organisms (WATERHAM et al. 1996 ). Evidence for proper targeting of the Pex2p-EGFP chimera was provided by the observation that the PEX2-EGFP gene fully complemented a P. pastoris pex2-deletion strain with regard to methanol- and oleate-growth phenotypes and restoration of peroxisomal matrix protein targeting (not shown). In fluorescence microscopy of wild-type methanol-induced cells, the Pex2p-EGFP protein localized to the peroxisomal membranes, which appeared as bright rings surrounding the organelles (Figure 5C). In cells of the pexA, pexC, and control pex1 strains, this normal peroxisomal membrane fluorescence pattern was not observed. Instead, most cells contained one or two brightly fluorescing punctate structures (Figure 5F, Figure L, and Figure O). This pattern is typical of peroxisomal remnants or ghosts, residual peroxisome-related matrix protein import-incompetent vesicular structures that are common to most pex mutants (GOULD et al. 1996 ) and indicated that these mutants have peroxisomal remnants and that the Pex2p-EGFP protein was properly targeted to the remnants. Thus, unlike the matrix protein import pathways, the mPTS pathway is functional in pex A and pexC mutants. In pexB cells, fluorescence also localized to punctate structures, but the intensity of the signal was significantly lower than in the other pex mutants and a low intensity cytoplasmic fluorescence was apparent (Figure 5I). This result was not specific to one pexB transformant because each of 10 pexB strains transformed with the PEX2-EGFP expression vector showed the same pattern. One possible explanation for this result is that the pexB mutant contains peroxisomal remnants but is partially defective in membrane protein targeting. A second possibility is that the membrane protein-targeting system may be fully functional but the remnants are too small to accommodate all the Pex2p-EGFP protein present in pexB mutant cells.

The remaining two complementation group representatives (tentatively named mxr1 and mxr2) were examined as described above but with different results. Both had peroxisomal enzyme levels that were uncharacteristic of pex mutants (Table 3). In extracts prepared from oleate-induced cells of mxr1, levels of acyl-CoA oxidase activity, CAT activity, and thiolase protein were similar to those in wild-type cells. However, in methanol-induced cells of the strain, levels of AOX and CAT activity (Table 3) and protein (not shown) were low (Table 3). In mxr2 cells, oleate- and methanol-induced peroxisomal enzyme levels were low. These included activity levels for acyl-CoA oxidase and CAT (Table 3) and protein levels for thiolase and CAT (not shown) in oleate-induced cells, and activity (Table 3) and protein (not shown) for AOX and CAT in methanol-induced cells of the mutant. Subcellular fraction studies showed that CAT activity in homogenates prepared from methanol- or oleate-induced cells of mxr1 and mxr2 was primarily located in pellet fractions (Table 4). Taken together, these results suggested that mxr1 and mxr2 mutants contained functional peroxisomes. However, except for oleate-induced cells of mxr1, the organelles were likely to be much smaller than normal due to the low levels of matrix proteins. These conclusions were supported by electron microscopic examination of mxr1 and mxr2. In methanol-induced cells of mxr1 and mxr2 and oleate-induced cells of mxr2, the large, numerous peroxisomes typical of wild-type P. pastoris were absent. However, in oleate-induced cells of mxr1, peroxisomes were readily apparent but somewhat smaller and fewer in number than in wild-type cells (Figure 4).

We considered the possibility that MXR1 and MXR2 were genes required for expression of methanol pathway (MXR1 and MXR2) and oleate pathway enzymes (MXR2 only). Preliminary evidence for this hypothesis was obtained by introducing a vector that expresses E. coli ß-lactamase under control of the AOX1 promoter of P. pastoris into each mutant (WATERHAM et al. 1997 ). The ability of the mutants to induce expression of ß-lactamase from this promoter in response to methanol was then assessed. Both mxr1 and mxr2 induced only very low levels of ß-lactamase relative to either wild-type or pex1 strains (Table 3), suggesting that MXR1 and MXR2 may be involved in transcription of AOX1 and perhaps other methanol pathway genes.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

The P. pastoris pex mutant selection schemes described here compare favorably with three schemes previously reported for selection of S. cerevisiae pex mutants. VAN DER LEIJ et al. 1992 reported the use of the catalase inhibitor 3-amino-1,2,4-triazol (3-AT) as a pex mutant selective agent. The basic concept for the selection is that, without catalase, oleate-induced S. cerevisiae cells (containing peroxisomal ß-oxidation pathway enzymes) should accumulate lethal levels of H₂O₂. However, mutants that lacked a functional ß-oxidation system, including pex mutants, should survive. A scheme reported by ZHANG et al. 1993 is based on the same H₂O₂ toxicity principle except that a parent strain defective in catalase (both cytoplasmic catalase T and peroxisomal catalase A) is employed. Using this same general strategy, we attempted to select for pex mutants using a P. pastoris mutant defective in catalase activity as the parental strain and selection plates containing methanol and sorbitol. However, for reasons that are unknown, the procedure did not result in a significant enrichment for pex mutants.

A third selection scheme described by ELGERSMA et al. 1993 makes use of a parental strain that expresses a chimeric polypeptide composed of the bleomycin resistance protein fused to peroxisomal luciferase (Ble-Luc). In the parental Ble-Luc strain, the fusion protein is targeted to peroxisomes via the luciferase PTS1. As a result, the strain is sensitive to phleomycin, a toxic compound and Ble ligand. However, pex mutants derived from the Ble-Luc strain are resistant to the drug because, without peroxisomes, the Ble-Luc protein is no longer sequestered and is free to bind the drug. The most effective of the three methods appears to be the phleomycin resistance scheme where ~10% of colonies surviving the drug are pex mutants (ELGERSMA et al. 1993 ). This number is comparable to the frequency of pex mutants from our selection methods (~1.2% of colonies arising after the allyl alcohol selection and ~25% of colonies surviving the methanol selection).

Comparison of the types and frequencies of mutants generated by our P. pastoris enrichment schemes reveals interesting similarities and differences. As expected, most of the pex mutants isolated by the selection schemes represent new alleles in PEX genes previously identified by the negative screening method (Table 2; GOULD et al. 1992 ; LIU et al. 1992 ). Even the relative frequencies with which specific pex alleles are found are generally similar. However, there are some striking differences as well. The allyl alcohol selection resulted in the isolation of five mutants in two genes (tentatively named MXR1 and MXR2), which were not seen with either the negative screen or the methanol selection methods. Although the mxr mutants grow on neither methanol nor oleic acid, they are clearly not pex mutants, because catalase fractionates to the crude organellar pellet in differential centrifugation experiments with these strains. Thus, they represent the only mutants that are specifically defective in growth on these peroxisome-requiring substrates that are not pex mutants. The low levels of peroxisomal matrix proteins in mxr1 (methanol-utilization enzymes only) and mxr2 (both methanol- and oleate-utilization enzymes) suggest these mutants may be defective in a factor required for the transcription of matrix enzymes. Further studies on these mutants and the affected genes are in progress. The fld1/high methanol selection method preferentially enriches for mutations in some PEX genes such as PEX3, PEX13, and PEXA (each at ~11% of all pex mutants isolated) relative to either the negative screening or allyl alcohol methods (see Table 2). Importantly, the high methanol selection procedure succeeded in yielding numerous alleles in each of three new P. pastoris PEX gene groups. [Because we do not know whether these new groups are homologues of PEX genes previously identified in other organisms, we have temporarily named them PEXA, PEXB, and PEXC. Once their genes have been cloned, the sequence of their products will be used to assign them to a new or existing PEX homologue group according to the guidelines described in DISTEL et al. 1996 .] Although the reason for the differences in frequencies and types of mutants resulting from the different pex mutant selection schemes is unclear, it is apparent that the schemes have subtle but distinct biases that influence the frequency of pex mutants that are found. Thus, to maximize our chances of identifying mutants in novel PEX genes, we believe the best strategy is the continued application of both schemes.

The identification of genes required for peroxisome biogenesis (PEX genes) through the isolation of yeast pex mutants has revolutionized our understanding of this intriguing organelle (WATERHAM and CREGG 1997 ). Furthermore, the amino acid sequences predicted by yeast PEX genes have been utilized to identify their human counterparts, several of which have recently been shown to be defective in patients with the lethal peroxisomal biogenesis disorder Zellweger syndrome (SUBRAMANI 1997 ). The isolation of yeast mutants in novel PEX genes will undoubtedly provide further insights into peroxisomal biogenesis mechanisms as well as the identification of additional Zellweger genes.

	FOOTNOTES

¹ Present address: Department of Pediatrics, Academic Medical Centre, University of Amsterdam, 11065 AZ Amsterdam, The Netherlands.

	ACKNOWLEDGMENTS

We thank Dr. S. Subramani of the University of California at San Diego for the PTS2-EGFP expression vector pTW65 and Dr. W.-H. Kunau for the antithiolase antibodies. We also thank T. Hadfield (Oregon Graduate Institute, Portland, OR) for assistance in preparing the manuscript. This research was supported by National Institutes of Health grants DK-43698 and TW-00547 to J.M.C. J.L.C. is the recipient of a postdoctoral fellowship from the American Heart Association, Oregon Affiliate.

Manuscript received April 13, 1998; Accepted for publication January 6, 1999.

	LITERATURE CITED

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION LITERATURE CITED

ALBERTINI, M., P. REHLING, R. ERDMANN, W. GIRZALSKY, and J. A. K. W. KIEL et al., 1997  Pex14p, a peroxisomal membrane protein binding both receptors of the two PTS-dependent import pathways. Cell 89:83-92[Medline].

CIRIACY, M., 1975  Genetics of alcohol dehydrogenase in Saccharomyces cerevisiae. Isolation and genetic analysis of adh mutants. Mutat. Res. 29:315-326.

CREGG, J. M., K. R. MADDEN, K. J. BARRINGER, G. P. THILL, and C. A. STILLMAN, 1989  Functional characterization of the two alcohol oxidase genes from the yeast Pichia pastoris.. Mol. Cell. Biol. 9:1316-1323[Abstract/Free Full Text].

CREGG, J. M., I. J. VAN DER KLEI, G. J. SULTER, M. VEENHUIS, and W. HARDER, 1990  Peroxisome-deficient mutants of Hansenula polymorpha.. Yeast 6:87-97.

CREGG, J. M., S. SHEN, M. JOHNSON and H. WATERHAM, 1998 Classical genetic manipulation, Chapter 2, pp. 17–26 in Methods in Molecular Biology, Pichia Protocols 1998, edited by D. R. HIGGINS and J. M. CREGG. Humana Press Inc., Totowa, NJ.

DE HOOP, M. J. and G. AB, 1992  Import of proteins into peroxisomes and other microbodies. Biochem. J. 286:657-669.

DISTEL, B., R. ERDMANN, S. J. GOULD, G. BLOBEL, and D. I. CRANE et al., 1996  A unified nomenclature for peroxisome biogenesis factors. J. Cell Biol. 135:1-3[Free Full Text].

DOMMES, V., C. BAUMGART, and W.-H. KUNAU, 1981  Degradation of unsaturated fatty acids in peroxisomes: existence of a 2,4-dieonyl-CoA reductase pathway. J. Biol. Chem. 256:8259-8262[Abstract/Free Full Text].

DOUMA, A. C., M. VEENHUIS, W. DE KONING, M. EVERS, and W. HARDER, 1985  Dihydroxyacetone synthase is localized in the peroxisomal matrix of methanol-grown Hansenula polymorpha.. FEBS Lett. 143:237-243.

DYER, L. M., J. A. MCNEW, and J. M. GOODMAN, 1996  The sorting sequence of the peroxisomal integral membrane protein PMP47 is contained within a short hydrophilic loop. J. Cell Biol. 133:269-280[Abstract/Free Full Text].

ELGERSMA, Y., M. VAN DEN BERG, H. F. TABAK, and B. DISTEL, 1993  An efficient positive selection procedure for the isolation of peroxisomal import and peroxisome assembly mutants of Saccharomyces cerevisiae.. Genetics 135:731-740[Abstract].

ELGERSMA, Y., L. KWAST, A. KLEIN, T. VOORN-BROUWER, and M. VAN DEN BERG et al., 1996  The SH3 domain of the Saccharomyces cerevisiae peroxisomal membrane protein Pex13p functions as a docking site for Pex5p, a mobile receptor for the import of PTS1-containing proteins. J. Cell Biol. 135:97-109[Abstract/Free Full Text].

ELGERSMA, Y., M. ELGERSMA-HOOISMA, T. WENZEL, J. M. MCCAFFERY, and M. G. FARQUHAR et al., 1998  A mobile PTS2 receptor for peroxisomal protein important in Pichia pastoris.. J. Cell Biol. 140:807-820[Abstract/Free Full Text].

ERDMANN, R. and G. BLOBEL, 1996  Identification of Pex13p, a peroxisomal membrane receptor for the PTS1 recognition factor. J. Cell Biol. 135:111-121[Abstract/Free Full Text].

ERDMANN, R., M. VEENHUIS, and W.-H. KUNAU, 1997  Peroxisomes: organelles at the crossroads. Trends Cell Biol. 7:400-406[Medline].

GLOVER, J. R., D. W. ANDREWS, S. SUBRAMANI, and R. A. RACHUBINSKI, 1994  Mutagenesis of the amino targeting signal of the Saccharomyces cerevisiae 3-ketoacyl-CoA thiolase reveals conserved amino acids required for the import into peroxisomes in vivo.. J. Biol. Chem. 269:7558-7563[Abstract/Free Full Text].

GOULD, S. J., G.-A. KELLER, and S. SUBRAMANI, 1987  Identification of a peroxisomal targeting signal at the carboxy terminus of firefly luciferase. J. Cell Biol. 105:2923-2931[Abstract/Free Full Text].

GOULD, S. J., D. MCCOLLUM, A. P. SPONG, J. A. HEYMAN, and S. SUBRAMANI, 1992  Development of the yeast Pichia pastoris as a model organism for a genetic and molecular analysis of peroxisome assembly. Yeast 8:613-628[Medline].

GOULD, S. J., J. E. KALISH, J. C. MORRELL, J. BJORKMAN, and A. J. URQUHART et al., 1996  Pex13p is an SH3 protein of the peroxisome membrane and a docking factor for the predominantly cytoplasmic PTS1 receptor. J. Cell Biol. 135:85-95[Abstract/Free Full Text].

KALISH, J. E., G.-A. KELLER, J. C. MORRELL, S. J. MIHALIK, and B. SMITH et al., 1996  Characterization of a novel component of the peroxisomal protein import apparatus using fluorescent peroxisomal proteins. EMBO J. 15:3275-3285[Medline].

KOMORI, M., S. W. RASMUSSEN, J. A. K. W. KIEL, R. J. S. BAERENDS, and J. M. CREGG et al., 1997  The Hansenula polymorpha PEX14 gene encodes a novel peroxisomal membrane protein essential for peroxisome biogenesis. EMBO J. 16:44-53[Medline].

KOUTZ, P., G. R. DAVIS, C. STILLMAN, K. BARRINGER, and J. CREGG et al., 1989  Structural comparison of the Pichia pastoris alcohol oxidase genes. Yeast 5:167-177[Medline].

LIU, H., X. TAN, M. VEENHUIS, B. MCCOLLUM, and J. M. CREGG, 1992  An efficient screen for peroxisome-deficient mutants of Pichia pastoris.. J. Bacteriol. 174:4943-4951[Abstract/Free Full Text].

LIU, H., X. TAN, K. RUSSELL, M. VEENHUIS, and J. M. CREGG, 1995  PER3, a gene required for peroxisome biogenesis in Pichia pastoris, encodes a peroxisomal membrane protein involved in protein import. J. Biol. Chem. 270:10940-10951[Abstract/Free Full Text].

MCCOLLUM, D., E. MONOSOV, and S. SUBRAMANI, 1993  The pas8 mutant of Pichia pastoris exhibits the peroxisomal import deficiencies of Zellweger syndrome cells—the PAS8 protein binds to the COOH-terminal tripeptide peroxisomal targeting signal, and is a member of the TPR protein family. J. Cell Biol. 121:761-774[Abstract/Free Full Text].

MONOSOV, E. A., T. J. WENZEL, G. H. LUERS, J. A. HEYMAN, and S. SUBRAMANI, 1996  Labeling of peroxisomes with green fluorescent protein in living P. pastoris cells. J. Histochem. Cytochem. 44:581-589[Abstract].

OSUMI, T., T. TSUKAMOTO, S. HATA, S. YOKOTA, and S. MIURA et al., 1991  Amino-terminal presequence of the precursor of peroxisomal 3-ketoacyl-CoA thiolase is a cleavable signal peptide for peroxisomal targeting. Biochem. Biophys. Res. Commun. 181:947-954[Medline].

RACHUBINSKI, R. A. and S. SUBRAMANI, 1995  How proteins penetrate peroxisomes. Cell 83:525-528[Medline].

REHLING, P., M. MARZIOCH, F. NIESEN, E. WITTKE, and M. VEENHUIS et al., 1996  The import receptor for the peroxisomal targeting signal 2 (PTS2) in Saccharomyces cerevisiae is encoded by the PAS7 gene. EMBO J. 15:2901-2913[Medline].

SAMBROOK, J., E. F. FRITSCH and T. MANIATIS, 1989 Molecular Cloning: A Laboratory Manual, Ed. 2. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.

SHEN, S., G. SULTER, T. W. JEFFRIES, and J. M. CREGG, 1998  A strong nitrogen source-regulated promoter for controlled expression of foreign genes in the yeast Pichia pastoris.. Gene 216:93-102[Medline].

SIBIRNY, A. A., V. I. TITORENKO, M. V. GONCHAR, V. M. UBIYVOVK, and G. P. KSHEMINSKAYA et al., 1988  Genetic control of methanol metabolism in yeast. J. Basic Microbiol. 28:293-319[Medline].

SIBIRNY, A. A., O. P. VITVITSKAYA, A. R. KULACHKOVSKY, and V. M. UBIYVOVK, 1989  Selection and properties of the mutants of methylotrophic yeast Hansenula polymorpha defective in alcohol oxidase. Mikrobiologiya (Moscow) 58:751-759.

SUBRAMANI, S., 1993  Protein import into peroxisomes and biogenesis of the organelle. Annu. Rev. Cell Biol. 9:445-478.

SUBRAMANI, S., 1997  PEX genes on the rise. Nat. Genet. 15:331-333[Medline].

UEDA, M., S. MOZAFFAR, and A. TANAKA, 1990  Catalase from Candida boidinii 2201. Methods Enzymol. 188:463-465[Medline].

VAN DER KLEI, I. J., L. V. BYSTRYKH, and W. HARDER, 1990  Alcohol oxidase from Hansenula polymorpha CBS 4732. Methods Enzymol. 188:420-422[Medline].

VAN DER LEIJ, I., M. VAN DEN BERG, R. BOOT, M. FRANSE, and B. DISTEL et al., 1992  Isolation of peroxisome assembly mutants from Saccharomyces cerevisiae with different morphologies using a novel positive selection procedure. J. Cell Biol. 119:153-162[Abstract/Free Full Text].

VEENHUIS, M., and W. HARDER, 1987 Metabolic significance and biogenesis of microbodies in yeasts, pp. 436–458 in Peroxisomes in Biology and Medicine, edited by H. D. FAHIMI and H. SIES. Springer-Verlag, Berlin.

VERDUYN, C., J. P. VAN DIJKEN, and W. A. SCHEFFERS, 1984  Colorometric alcohol assays with alcohol oxidase. J. Microbiol. Methods 2:15-25.

WATERHAM, H. R. and J. M. CREGG, 1997  Peroxisome biogenesis. BioEssays 19:57-66[Medline].

WATERHAM, H. R., Y. DE VRIES, K. A. RUSSELL, W. XIE, and M. VEENHUIS et al., 1996  The Pichia pastoris PER6 gene product is a peroxisomal integral membrane protein essential for peroxisome biogenesis and has sequence similarity to the Zellweger syndrome protein PAF-1. Mol. Cell. Biol. 16:2527-2536[Abstract].

WATERHAM, H. R., M. E. DIGAN, P. J. KOUTZ, S. V. LAIR, and J. M. CREGG, 1997  Isolation of the Pichia pastoris glyceraldehyde-3-phosphate dehydrogenase gene and regulation and use of its promoter. Gene 186:37-44[Medline].

WATERHAM, H. R., K. A. RUSSELL, Y. DE VRIES, and J. M. CREGG, 1997  Peroxisomal targeting, import, and assembly of alcohol oxidase in Pichia pastoris.. J. Cell Biol. 139:1419-1431[Abstract/Free Full Text].

ZHANG, J. W. and P. B. LAZAROW, 1996  Peb1p (Pas7p) is an intraperoxisomal receptor for the NH₂-terminal, Type 2, peroxisomal targeting sequence of thiolase: Peb1p itself is targeted to peroxisomes by an NH₂-terminal peptide. J. Cell Biol. 132:323-334.

ZHANG, J. W., Y. HAN, and P. B. LAZAROW, 1993  Novel peroxisome clustering mutants and peroxisome biogenesis mutants of Saccharomyces cerevisiae.. J. Cell Biol. 123:1133-1147[Abstract/Free Full Text].