Torsten Schöneberg and Anke Edelmann Rothemund, Manja Kamprad, Henning Otto, Bär, Wolfgang Schellenberger, Sven Katrin Tanneberger, Jürgen Kirchberger, Jörg SUBUNIT IN PICHIA PASTORIS RELEVANCE OF A THIRD TYPE OF IDENTIFICATION AND FUNCTIONAL A Novel Form of 6-Phosphofructokinase: Enzyme Catalysis and Regulation: doi: 10.1074/jbc.M611547200 originally published online May 23, 2007 2007, 282:23687-23697. J. Biol. Chem. 10.1074/jbc.M611547200 Access the most updated version of this article at doi: . JBC Affinity Sites Find articles, minireviews, Reflections and Classics on similar topics on the Alerts: When a correction for this article is posted • When this article is cited • to choose from all of JBC's e-mail alerts Click here Supplemental material: http://www.jbc.org/content/suppl/2007/05/24/M611547200.DC1.html http://www.jbc.org/content/282/32/23687.full.html#ref-list-1 This article cites 65 references, 19 of which can be accessed free at at Charité - Med. Bibliothek on February 10, 2015 http://www.jbc.org/ Downloaded from at Charité - Med. Bibliothek on February 10, 2015 http://www.jbc.org/ Downloaded from
Transcript
Torsten Schöneberg and Anke EdelmannRothemund, Manja Kamprad, Henning Otto,Bär, Wolfgang Schellenberger, Sven Katrin Tanneberger, Jürgen Kirchberger, Jörg SUBUNIT IN PICHIA PASTORISRELEVANCE OF A THIRD TYPE OFIDENTIFICATION AND FUNCTIONAL A Novel Form of 6-Phosphofructokinase:Enzyme Catalysis and Regulation:
doi: 10.1074/jbc.M611547200 originally published online May 23, 20072007, 282:23687-23697.J. Biol. Chem.
10.1074/jbc.M611547200Access the most updated version of this article at doi:
.JBC Affinity SitesFind articles, minireviews, Reflections and Classics on similar topics on the
Alerts:
When a correction for this article is posted•
When this article is cited•
to choose from all of JBC's e-mail alertsClick here
A Novel Form of 6-PhosphofructokinaseIDENTIFICATION AND FUNCTIONAL RELEVANCE OF A THIRD TYPE OFSUBUNIT IN PICHIA PASTORIS*□S
Received for publication, December 18, 2006, and in revised form, May 23, 2007 Published, JBC Papers in Press, May 23, 2007, DOI 10.1074/jbc.M611547200
Katrin Tanneberger‡, Jurgen Kirchberger‡, Jorg Bar‡, Wolfgang Schellenberger‡, Sven Rothemund§,Manja Kamprad¶, Henning Otto�, Torsten Schoneberg‡1, and Anke Edelmann‡
From the ‡Institute of Biochemistry, Molecular Biochemistry, Medical Faculty, University of Leipzig,Johannisallee 30, 04103 Leipzig, Germany, the §Interdisziplinares Zentrum fur Klinische Forschung (IZKF) core unit PeptideTechnologies, Medical Faculty, University of Leipzig, Inselstrasse 22, 04103 Leipzig, Germany, the ¶Institute of ClinicalImmunology and Transfusion Medicine, Medical Faculty, University of Leipzig, Johannisallee 30, 04103 Leipzig, Germany,and the �Institute of Chemistry and Biochemistry, Free University Berlin, Thielallee 63, 14195 Berlin, Germany
Classically, 6-phosphofructokinases are homo- and hetero-oligomeric enzymes consisting of � subunits and �/� subunits,respectively. Herein, we describe a new form of 6-phosphofruc-tokinase (Pfk) present in several Pichia species, which is com-posed of three different types of subunit, �, �, and �. Thesequence of the � subunit shows no similarity to classic Pfk sub-units or to other known protein sequences. In-depth structuraland functional studies revealed that the � subunit is a constitu-tive component of Pfk from Pichia pastoris (PpPfk). Analyses ofthe purified PpPfk suggest a heterododecameric assembly fromthe three different subunits. Accordingly, it is the largest andmost complex Pfk identified yet. Although, the � subunit is notrequired for enzymatic activity, the � subunit-deficient mutantdisplays a decreased growth on nutrient limitation and reducedcell flocculation when compared with the P. pastoris wild-typestrain. Subsequent characterization of purified Pfks from wild-type and � subunit-deficient strains revealed that the allostericregulation of the PpPfk by ATP, fructose 2,6-bisphosphate, andAMP is fine-tuned by the � subunit. Therefore, we suggest thatthe � subunit contributes to adaptation of P. pastoris to energyresources.
The ATP-dependent 6-phosphofructokinase (EC 2.7.1.11,phosphofructokinase-1, ATP:D-fructose-6-phosphate 1-phos-photransferase (Pfk))2 catalyzes in many organisms the phos-phorylation of fructose 6-phosphate (Fru 6-P) at position 1. The
Pfk activity is generally being sensitive to a number of allostericregulators, e.g.ATP,AMP,NH4
�, and fructose 2,6-bisphosphate(Fru 2,6-P2). Therefore, this irreversible reaction is consideredto be one of the rate-limiting steps of glycolysis (1–3). Mosteukaryotic Pfks are heteromeric enzymes consisting of sub-units, which evolved from a single ancestor gene by gene dupli-cation and mutational events (4, 5). Specific amino acid resi-dues involved in catalytic and regulatory functions of Pfk fromEscherichia coli (6, 7) are conserved in yeast and mammalianPfk genes. In eukaryotes the N-terminal half of a Pfk subunitobviously retained the catalytic function, whereas in the C-ter-minal half allosteric ligand binding sites have evolved from for-mer catalytic and regulatory sites (4, 8, 9). This assumption issupported by studies with mutants of Saccharomyces cerevisiaeexpressing only the � or the � subunit of Pfk. It was demon-strated that one subunit type alone is able to form an enzymat-ically active Pfk entity in vivo (10, 11). Crystallographic analysisshowed that an active bacterial Pfk consists of four identicalsubunits (12, 13). No high resolution structure of a eukaryoticPfk is available yet. But electron microscopic studies with S.cerevisiae Pfk (ScPfk) at 10.8-Å resolution suggested anoctameric enzyme assembly (14).Recently we co-purified a protein component together with
the known Pfk � and � subunits (15–17) from the methylotro-phic yeast Pichia pastoris. This unknown protein could only beseparated under denaturating conditions and with loss of Pfkactivity. Herein, we present the sequence of the co-purifiedcomponent and describe this new protein as a constitutivelybound and regulatory relevant subunit of Pfk from P. pastoris(PpPfk) and other Pichia sp. Based on themolecularmass of thenative PpPfk and the molar ratio and the molecular mass of theindividual subunits we propose an enzyme complex formed offour �, �, and � subunits.
EXPERIMENTAL PROCEDURES
Yeast Strains and Growth Conditions—Strains used for iso-lation of nucleic acids and for analysis of Pfk proteins are sum-marized in supplemental Table S1. Yeast cellswere cultivated inYP medium (1% yeast extract and 2% BactoPepton) containing2% glucose or 0.5% methanol at 30 °C under rotation up to thegrowth phase as indicated. Minimal medium containing 0.67%yeast nitrogen base, 0.5% ammonium sulfate, and 2% glucose
* This work was supported by the IZKF-Leipzig and the Bundesministeriumfur Bildung und Forschung. The costs of publication of this article weredefrayed in part by the payment of page charges. This article must there-fore be hereby marked “advertisement” in accordance with 18 U.S.C. Sec-tion 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to theGenBankTM/EBI Data Bank with accession number(s) AY686600.
□S The on-line version of this article (available at http://www.jbc.org) containssupplemental Tables S1 and S2 and Figs. S1–S12 and additional references.
1 To whom correspondence should be addressed. Tel.: 49-341-972-2150; Fax:49-341-972-2159; E-mail: [email protected].
2 The abbreviations used are: Pfk, 6-phosphofructokinase; DIG, digoxigenin;FACS, fluorescence-activated cell sorting; MALDI-TOF, matrix-assistedlaser desorption ionization-time of flight; MS, mass spectrometer; PpPfk,6-phosphofructokinase from P. pastoris; RACE, rapid amplification of cDNAends; ScPfk, 6-phosphofructokinase from S. cerevisiae; Fru 6-P, fructose6-phosphate; Fru 2,6-P2, fructose 2,6-bisphosphate; PBS, phosphate-buff-ered saline; FACS, fluorescence-activated cell sorting.
supplemented with adenine and amino acids but lacking uracilwas used as selective medium. All biochemicals for cell cultiva-tion were purchased from Difco (BD Biosciences) andInvitrogen.Preparation of Cell-free Extract and Assays—Cell-free
extractwas prepared according to Schwock et al. (18). Pfk activ-ity measurement followed basically procedures described else-where (16). For kinetic studies a Pfk assay with simultaneousATP and Fru 6-P regeneration (100mM imidazole/HCl, pH 6.6,100 mM KCl, 10 mM MgCl2, 20 mM potassium phosphate, 0.2mM NADH�, 0.6 mM phosphoenolpyruvate, 8.5 units of pyru-vate kinase/ml, 7 units of lactate dehydrogenase/ml, 1 unit offructose-1,6-bisphosphatase/ml; ATP, Fru 6-P, AMP, and Fru2,6-P2 as indicated) was used (16). A two-state Monod-Wyman-Changeux model was applied to describe the ATPvelocity curves under the assumptions: 1) an octameric alloster-ic mode, 2) AMP and Fru 2,6-P2 binding to the R-state enzymeonly, and 3) ATP serves as substrate (KS
ATP) in a hyperbolicmanner, but acts also as allosteric inhibitor (KT
ATP),
v � V ��ATP�
�KSATP � �ATP��
�1
1 � L(Eq. 1)
L � �m0 ��1 � �ATP�/KT
ATP�
�1 � �AMP�/KRAMP� � �1 � �Fru 2,6-P2�/KR
Fru 2,6-P2��8
(Eq. 2)
where V is maximum activity, m0 is the allosteric constant,KSATP is the ATP Michaelis constant, KT
ATP is the ATP-bindingconstant of the T-state enzyme, and KR
AMP and KRFru 2,6-P2 are
AMP and Fru 2,6-P2 binding constants of the R-state enzyme.For description of the Fru 6-P velocity curves and the
dependence of Pfk activity on AMP and Fru 2,6-P2 concentra-tions, a generalized Hill equation was used,
v � V0 � �Vmax � V0� ���X�/KA
X�nH
�1 � ��X�/KAX�nH
(Eq. 3)
where X is Fru 6-P, AMP or Fru 2,6-P2, and KAX is the half-
activity constant.The kinetic data were fitted to Equations 1–3 by non-linear
regression analysis applying SigmaPlot 9.0 (Systat Software,Inc., San Jose, CA) that uses the Marquardt-Levenberg algo-rithm for minimization. Alcohol oxidase activity was measuredin a reaction coupled to horseradish peroxidase and the oxida-tion of 2,2�-azinobis(3-ethylbenzthiazoline-6-sulfonic acid)(Sigma-Aldrich) in100mMpotassiumphosphatebuffer, pH7.5, at25 °C according to the company’s technical information. Fordetermination of protein concentrations, the procedure of Brad-ford (19) was applied using bovine serum albumin as standard.Purification of Pfk from P. pastoris and Protein Sequencing—
Pfk was isolated from cell-free extract of P. pastoris strainMH458 as described previously (16). N-terminal sequences ofpolypeptides were determined according to the Edman proce-dure using the Protein Sequencer 473A (Applied Biosystems,Foster City, CA). Tryptic in-gel digestion and MALDI massspectrometry measurements of the generated tryptic peptideswere carried out as described previously (20). The mass spec-
trometric measurements were performed on a Bruker ReflexMALDI-TOF mass spectrometer (Bruker Daltonik, Bremen,Germany) equipped with an ion gate and pulsed ion extraction.Post source decay fragment ion spectra were obtained by usingthe FAST method (Bruker Daltonik).Generation of a Polyclonal Antibody against the � Subunit of
PpPfk—Subunits of purified Pfk from P. pastoris strain MH458were separated under reducing conditions by SDS-PAGE. The� subunit was cut out, destained in 10% acetic acid containing40%methanol at 4 °C, and extracted by electroelution (Electro-Eluter Model 422, Bio-Rad). Then, the protein was dialyzedagainst phosphate-buffered saline (PBS; 50 mM sodium phos-phate, 150 mM NaCl, pH 7.0). 200 �g of antigen in completeFreund’s adjuvant (0.5-ml final volume) was used for rabbitimmunization. After 5 weeks the animal was boosted in thesame way. Antiserum was fractionated by 50% ammonium sul-fate saturation. The precipitated protein was dialyzed against20 mM sodium phosphate buffer, pH 7.0, and loaded onto aprotein-A-Sepharose CL-4B column (Amersham Biosciences).The antibody was eluted with 100 mM citrate buffer, pH 3.0,neutralizedwith 1MTris/HCl, pH9.0, precipitatedwith ammo-nium sulfate, and dissolved in PBS. For affinity purification,purified PpPfk was covalently coupled with bromocyan-acti-vated Sepharose 4B as recommended by the manufacturer(AmershamBiosciences). After washingwith PBS, the antibodywas eluted with 3 M MgCl2, dialyzed alternate against 155 mMNaCl and PBS, and stored at�20 °C. Protein concentrationwascalculated according to A1cm,1mg/ml
279nm � 1.35 (21).Cloning of PpPFK3 Encoding the � Subunit—Touchdown
PCR was carried out with genomic DNA as template andHotStarTaqTM DNA Polymerase (Qiagen). Furthermore,degenerate primers 4 and 14 were used, which correspondedto the identified amino acid sequences of the � subunit (allprimers are listed in supplemental Table S2). PCR was per-formed under the following conditions: Predenaturation at95 °C for 15 min was followed by cycles of denaturation at94 °C for 30 s, annealing beginning at 72 °C for 30 s, andelongation at 72 °C for 90 s. The annealing temperature waslowered 1 °C per cycle to 50 °C, which then was applied forannealing in the next 20 cycles. To identify the 5�- and3�-ends, rapid amplification of cDNA ends (RACE)-PCR wasperformed as described previously (17) and according to themanufacturer’s protocol (Gene-RacerTM kit with clonedavian myeloblastosis virus reverse transcriptase, Invitrogen).PCR fragments were subcloned into pCR2.1 (the TOPOT-
MTA Cloning� kit for sequencing, Invitrogen) andsequenced in both directions using the ABI PRISM� Big-DyeTM Terminators version 2.0 Cycle Sequencing Kit(Applied Biosystems).Generation of Individual Pfk Subunit-deficient P. pastoris
Strains—Pfk subunit-deficient P. pastoris strains were gen-erated by homologous recombination. Each plasmid (pAE27,pAE28, and pAE34) harbored one of the PpPfk genes inter-rupted by URA3 from P. pastoris (plasmids are depicted insupplemental Fig. S1). Transformation of the P. pastorisstrain JC307 his4 ura3 was performed by electroporation(P. pastoris adjustment, GenePulser Xcell, Bio-Rad). Toscreen mutants and to verify homologous recombination,
Phosphofructokinase from P. pastoris
23688 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 32 • AUGUST 10, 2007
Southern blot analyses were performed as described previ-ously (22).SDS-PAGE/Western Blot Analysis—Western blot analysis
followed the description of Bar et al. (23) with the exception ofthe use of 10%polyacrylamide gels. Polyclonal rabbit antibodiesagainst the � subunit of PpPfk (this work) and against the puri-fied ScPfk (24) were applied. The anti-ScPfk antibody showedstrong cross-reactivity to the � and � subunits of PpPfk. Immu-nological detection was performed with anti-rabbit-IgG perox-idase conjugate (Dianova, Germany) and a chemiluminescentdetection ECLTM Western blotting system (Amersham Bio-sciences-GE Healthcare).FACS Analysis and Immunofluorescence Microscopy—Cells
were characterized by FACS analysis using forward light scat-tering and side light scattering, reflecting cell size and cell com-plexity, respectively. Thesewere recorded on linear scales. Flowcytometric analysis was performed using the FACSCaliburTMscanner equipped with CellQuestTM software (both BD Bio-sciences). Thus, cells were grown in YPmedium containing 2%glucose to an optical density of A580 nm � 10 and diluted 1:51(v/v) in the respective medium.For analysis of the subcellular distribution of Pfk protein by
immunofluorescence microscopy, P. pastoris cells (A580 nm 1) were processed according to Pringle et al. (25). Polyclonalantibodies against Pfk subunits (see SDS-PAGE/Westernblot analysis) were used for specific protein detection. Cy3-labeled goat anti-rabbit-IgG antibody (Dianova, Hamburg,Germany) 200-fold (v/v) diluted with PBS containing 0.1%bovine serum albumin was applied as secondary antibody.Nucleus staining was performed with 4�,6-diamidino-2-phe-nylindol (1 �g/ml in PBS, Serva, Heidelberg, Germany) atroom temperature for 5 min. Fluorescence images wereobtained with a fluorescence microscope (Leica DM 5000B,Leica Microsystems CMS GmbH, Wetzlar, Germany)
equipped with a 63�/1.4-0.6 oil immersion objective, aDFC350FX camera, and FW4000 software.Immunoprecipitation of PpPfk—The polyclonal antibody
against the � subunit of PpPfk (44 �l of affinity-purified IgGfraction; 0.6 �g/�l) was mixed with 200 �l of cell-free extractand stored at 4 °C for 30min. Then, 20mg of protein-A-agarose(wet weight, Roche Applied Science) washed with PBS wereadded. Following incubation at 4 °C for 3 h and centrifugationat 14,000 � g for 1 min, the gel was washed twice with ice-coldPBS. Immunoprecipitated proteinswere released by incubationwith 20 �l of 65 mM Tris/HCl buffer, pH 6.8, containing 20%Bromphenol Blue, 20% glycerol, 5% 2-mercaptoethanol, and 2%SDS in a boiling water bath for 5 min. Protein samples wereanalyzed by SDS-PAGE and Western blotting.
RESULTS
Molecular Identification of a Pfk � Subunit in P. pastoris—The protein band corresponding to the unknown polypeptidechain, whichwas co-purified togetherwith the� and� subunitsof PpPfk, was isolated from SDS-PAGE gel. Then, the N-termi-nal amino acid sequences of this protein and of several frag-ments obtained by chymotrypsin or trypsin degradation weredetermined by Edmanprocedure andMALDI-TOFpost sourcedecay analysis. Amino acid sequences identified are summa-rized in Table 1. Based on these results, degenerate primerswere designed (supplemental Table S2). PCR and cloning tech-niques (see “Experimental Procedures”) revealed a genomicDNA fragment of 3113 bp containing a complete codingsequence of 1056 bp (GenBankTM accession numberAY686600). The transcription start was found at �43 bp fromstart ATG by 5�-RACE-PCR from mRNA. The codingsequence, further referred to as PpPFK3 (according to thenomenclature of other Pfk genes), encodes a polypeptide with apredicted molecular mass of 40.8 kDa (Fig. 1). N-terminal and
TABLE 1Protein sequence analysis of the � subunit of PpPfk
Fragment Molecular mass, mono�M � H�� Putative sequencea
DaEDMAN sequencingb(M)VTKDSIIRDLERENVGPEFGEFLNTLQTDLNS N terminus � subunit (1–33)RSSR�PW�EDKVKGPALA N terminus internal sequence
a Numbers in parentheses represent amino acid position.b Purified PpPfk from P. pastoris strainMH458was subjected to SDS-PAGE. The � subunit, co-purified with the� and� subunits, was excised and partially sequenced by Edmanand MALDI-TOF post source decay techniques.
c In-gel tryptic digestion and MALDI-TOF analysis were used to verify the integrity of the C-terminal part of the � subunit.
Phosphofructokinase from P. pastoris
AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23689
internal amino acid sequences of the co-purified componentdetermined by protein sequencing were identical to the respec-tive sequence regions of the translated open reading frame ofPpPFK3. However, the potential � subunit appears at 34kDa when purified PpPfk is analyzed by SDS-PAGE (see Fig.5). The integrity of the N- and C-terminal ends of the �subunit isolated by SDS-PAGE (Table 1) was verified byEdman and MALDI-TOF analyses, respectively. Therefore,proteolytic modifications of this protein can be excluded.Accordingly, the discrepancy between the sequence-basedcalculated mass and the apparent molecular mass found bySDS-PAGE is caused by the specific migration property ofthe � subunit in this electrophoresis.
Extensive sequence analyses with various bioinformatictools (NCBI Blast analyses, Predict-Protein, ELM) revealedno significant sequence similarity to any sequence deposit inGenBankTM (last analysis February, 27, 2007). Within thecloned genomic 5� non-coding region (2000 bp) we identifiedputative consensus sequences forGCR1.This transcription fac-tor is involved in specific regulatory mechanisms for glycolyticgene expression (26, 27). In addition, an incomplete open read-ing frame of a hypothetical protein was detected. It showssequence homology to the hypothetical proteins CAHO2429.1
and DEHAOCO3872.g fromKluyveromyces lactis (8/324 � 27%)and from Debaryomyces hanseniiCBS767 (111/311 � 35%),respectively.Screening Other Yeasts for the
Presence of the Pfk � Subunit—Toaddress the question whether the �subunit is unique to P. pastoris, sev-eral other yeasts (supplementalTable S1) were initially screened forimmuno-cross-reactivity by West-ern blot analysis. For this purpose,we used the polyclonal antibodyagainst the PpPfk � subunit.Whereas differentP. pastoris strainsand several other Pichia species dis-played an immunoreactive bandbetween 34 and 42 kDa, all extractsof distantly related yeasts (see sup-plemental Table S1) showed nospecific immunoreactivity (datanot shown). Next, PCR wasapplied to amplify ortholog se-quences using degenerate primersets designed on the basis of the �subunit sequence from P. pastorisstrain MH458. So far, the presenceof a � subunit was verified in P.pastoris strains JC307 and GS115,and in P. pseudopastoris (Gen-BankTM accession numbersDQ352840, DQ374390, andDQ386148). The � subunits fromP. pastoris strains JC307 and
GS115 are identical and show 95.4% amino acid identity tothe � subunit of P. pastoris strainMH458. The � subunit of P.pseudopastoris, a species closely related to P. pastoris (28),displays 79.8% identity at the amino acid level to the � sub-units from P. pastoris strains.Generation and Functional Characterization of � Subunit-
deficient P. pastoris Strains—To analyze the relevance of the �subunit, threeP. pastorismutants were generated by deletion ofPpPFK3. Correct recombination and gene deletion were con-firmed by Southern and Western blot analyses (supplementalFig. S2). The transformed recipient P. pastoris strain, whichcontained the Ura3 marker gene homologously integrated atthe endogenous ura3 locus but still maintained an intactPpPFK3 locus, served as proper control (further referred to aswild-type strain).First, basic cell functions of the three � subunit-deficient
strains were studied. Since these strains behaved identically inall experiments, data are exemplarily shown for JC307–22�PpPFK3 (further referred to as � subunit-deficient strain).Growth of the � subunit-deficient strain was significantlyreduced by 20% after 22 h of cultivation (Fig. 2A). This effectwas found also under cultivation in a continuously oxygenizedatmosphere. Deletion of the � subunit did not interfere with the
FIGURE 1. Molecular identification of Pfk � subunits in different Pichia species. The coding sequence of the� subunit of Pfk, PFK3, was cloned from P. pastoris strains MH458, JC307, and GS115 and from P. pseudopastorisstrain Y01541 (GenBankTM accession numbers AY686600, DQ352840, DQ374390, and DQ386148). The isolatedsequences from P. pastoris strains GS115 and JC307 are identical. Amino acid sequences of the � subunitorthologs of P. pastoris strains JC307 and MH458 and from P. pseudopastoris strain Y01541 are shown andconserved positions are indicated (asterisks). Fragments identified by direct protein sequencing of the purified� subunit from P. pastoris strain MH458 (Table 1) are highlighted. A chymotrypsin cleavage site identified by alimited proteolysis experiment is arrowed.
Phosphofructokinase from P. pastoris
23690 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 32 • AUGUST 10, 2007
specific Pfk activity in cell-free extracts (Fig. 2B).Wild-type and� subunit-deficient strains were able to grow on medium con-taining glycerol, rhamnose, or cycloheximide. This was deter-mined by cultivation on pre-made culture plates (ID32CbioMerieux system) at 30 °C for 4 days (data not shown). Tem-perature sensitivity andNaCl tolerancewere also indistinguish-able between the strains (supplemental Fig. S3).Further, we analyzed the adaptation ability of themutant and
wild-type strains to glucose-containing medium after cultiva-tion on methanol. In both cases a 50% reduction of alcoholoxidase activity over 3 h was observed (supplemental Fig. S4A)as a result of induced degradation of peroxisomes (microauto-phagy) (15, 29). This was accompanied by a 50% increase in Pfkactivity (supplemental Fig. S4B). Further, the strains weretested for endocytosis and vacuolar morphology by in vivo
labeling (30). For this purpose weused the fluorescent lipophilic dyeFM1-43 (Molecular probes, Brus-sels, Belgium). FM1-43 internaliza-tion from the plasma membraneinto endosomes and vacuolar mem-branes was indistinguishablebetween mutant and wild-typestrains (supplemental Fig. S5).Althoughmost of the phenotypes
of the wild-type and the � subunit-deficient strains were very similar,cells of the wild-type strain showremarkable flocculation with in-creasing cell density. The cellsstarted to adhere at the middle log
growth phase. The formed macroscopic flocs rapidly sedi-mented when continuous shaking was stopped. This pheno-type was greatly reduced in � subunit-deficient strains (Fig.3). We observed that cell-cell adhesion in the wild-typestrain was abolished in the presence of 2 mM EDTA butoccurred again after the addition of 5 mM Ca2�. Adhesion ofcells lacking the � subunit was also induced by addition ofCa2� but to a lesser extent (data not shown). Differences inthe cellular structure of the � subunit-deficient and the wild-type strains were also reflected by FACS analysis (supple-mental Fig. S6).Kinetic Properties of the Wild-type and � Subunit-deficient
PpPfks—Because the � subunit is not essential for the catalyticfunction of PpPfk per se (see above), we initiated an in-depthkinetic analysis of PpPfks purified from wild-type and � sub-unit-deficient strains (Tables 2 and 3).The purity of all enzyme preparations was verified by SDS-
PAGE (supplemental Fig. S7). In contrast to ScPfk (for compar-ison see Ref. 31) the sensitivity of bothPpPfks toAMP (KA
AMP) ishigher, whereas that to Fru 2,6-P2 (KA
Fru 2,6-P2) was about 10-foldlower. Furthermore, the twoPpPfk formswere characterized bya more potent ATP inhibition (residual activity of 0.53 0.05for ScPfk). Comparing the kinetic data of the wild-type and the� subunit-deficient PpPfks, no remarkable differences in thehalf-saturating Fru 6-P concentration (K0.5
Fru6-P) and the Fru 6-Pcooperativity (nH) in the presence and absence of AMP and Fru2,6-P2were found. Further,Michaelis constants (KS
ATP, Table 3)for ATP as substrate are equal. However, a lower affinity of theallosteric ATP-binding site (KT
ATP, Table 3) was determined forthe � subunit-deficient PpPfk. Consequently, this enzyme formshowed a less efficient inhibition by ATP (supplemental Fig.S8). Further, a remarkable lower sensitivity of the mutatedPpPfk to AMP was observed. As shown in Fig. 4 (A and C), thereduced AMP activation becomes apparent especially at phys-iological ATP levels (�1 mM (32)). The � subunit-deficientPpPfk is also less sensitive to Fru 2,6-P2 at intermediate ATPconcentrations (Fig. 4,B andD). However, with increasingATPlevels the Fru 2,6-P2 activation ratio of the two enzymes con-verged (Fig. 4B). At very high concentrations (5mMATP) theactivation ratio of the � subunit-deficient PpPfk exceeded thatof the wild-type enzyme (data not shown). In terms of theMonod-Wyman-Changeux model (Equation 1), differences in
FIGURE 2. Growth curves and Pfk activity of wild-type and � subunit-deficient P. pastoris strains. Wild-type (open square) and � subunit-deficient (closed square) strains were cultivated in YP medium containing 2%glucose at 30 °C and 250 rpm. Cell density (A580 nm; A) and specific Pfk activity (B) were determined spectro-photometrically (see “Experimental Procedures”). Data are given as mean S.D. of at least three independentexperiments each performed in duplicate.
FIGURE 3. Reduced flocculation phenotype of P. pastoris cells lacking the� subunit of Pfk. Wild-type and � subunit-deficient strains were cultivated inYP medium containing 2% glucose at 30 °C and 250 rpm up to A580 nm 10.Sedimentation by gravity is depicted at 1 (left) and at 2 (right) min after takingthe samples from the shaker.
Phosphofructokinase from P. pastoris
AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23691
kinetics of the two PpPfk forms could not be described bychanges of the allosteric constant (m0). However, the data fit-ting revealed differences in the constants characterizing ATPbinding to the T-state enzyme andAMP and Fru 2,6-P2 bindingto the R-state enzyme (Table 3).Interactions andAssembly of PpPfk Subunits—Recent studies
have shown that the extent of proteolytic degradation dependson both the specificity of the protease and the presence of pro-tective substrates or allosteric effectors such as ATP (33).Therefore, limited proteolysis of the purified PpPfk was per-formed with chymotrypsin in presence of saturating ATP con-centration to analyze the topology of the subunit assembly(detailed conditions are given in the legend of Fig. 5). Afterproteolysis, enzyme activity was reduced by only 10% in com-parison to the non-degraded PpPfk. Although the three sub-units possess multiple cleavage sites to chymotrypsin, only the� subunit (70 kDa, ��) and � subunit (20 kDa, ��) weretruncated as demonstrated by SDS-PAGE (Fig. 5, lane 3). Theidentity of the �� fragment was verified by Edman protein
sequencing (Table 1). To analyze whether the chymotrypsin-treated enzyme still forms a high molecular complex, the sam-ple was subjected to high-performance liquid chromatographygel filtration. A main protein fraction was obtained, which cor-responded to 900 kDa (data not shown). The analysis of this900 kDa-protein fraction by SDS-PAGE (Fig. 5, lane 4) revealeda band pattern identical to the non-fractionated chymotrypsin-treated PpPfk (Fig. 5, lane 3).Immunoprecipitation was carried out to further prove inter-
actions between the three differentPpPfk subunits. As shown inFig. 6, � and � subunits were only precipitated with the � sub-unit-specific antibody in presence of the � subunit (wild-typestrain). Although precipitation was not complete, the relativesignal intensities of � and � subunits were always equal in thesupernatant and in the protein-A-precipitated fraction. Thisresult is indicative for a defined stoichiometry between thesubunits.Further, wild-type and � subunit-deficient strains were sub-
jected to immunofluorescencemicroscopy to analyze subcellu-lar distribution of PpPfk. The three different subunits displayeda similar cytosolic distribution pattern (supplemental Fig. S9,Aand B). The subcellular distribution of the �/� subunitsremained unchanged in the � subunit-deficient strain (supple-mental Fig. S9C).To analyze whether the � subunit can exist independently
from the classic Pfk subunits, � and � subunit-deficient strainswere generated by homologous recombination. The � subunitencoding gene, PpPFK2, was cloned recently (17), but sequenceinformation of PpPFK1, encoding the PpPfk � subunit, waslacking. We isolated a 6862-bp genomic fragment containingthe complete coding sequence ofPpPFK1 (2970 bp) and parts ofthe 5� and 3� non-coding regions (3809 and 83 bp, respectively)(GenBankTM accession number AF508861; supplemental Fig.S10) from P. pastoris strain MH458. Homologous recombina-tion of the constructs pAE27 (PpPFK1) and pAE28 (PpPFK2)(supplemental Fig. S1) was confirmed by Southern blotting(supplemental Fig. S11). As shown in Fig. 7A (lane 2), deletionof the� subunit resulted nearly in loss of� and � subunits in thecytosolic fraction. Analyzing the cell-free extract of the � sub-unit-deficient strain, we found the� subunit but only a very lowamount of the � subunit (Fig. 7A, lane 3). The individual dele-tion of both, the � subunit and the � subunit, significantlyretarded yeast growth on glucose and abolished Pfk activitymeasurable in cell-free extract (Fig. 7C, lanes 2 and 3).
Reconstitution of the wild-type PpPfk from the individualsubunits should provide further evidence for association of the� subunit with the �/� complex. As stated above, the � subunitcan only be separated from purified PpPfk under denaturatingconditions (16). Therefore, we initially attempted to reconsti-tute the enzyme from the individual subunits following com-plete denaturation by urea or guanidine hydrochloride. In con-trast to Pfk from S. cerevisiae (34), all efforts failed to refold andassemble the individual subunits.Next, we tried to express the�subunit in S. cerevisiae to reconstitute the purified polypeptidewith the isolated �/� complex from PpPfk. Heterologousexpression attempts were carried out under the control of theoriginal PpPFK3 promoter region and of the promoter of theScPfk � subunit. The wild-type and His-tagged � subunit con-
TABLE 2Comparison of kinetic properties of purified Pfks from P. pastoriswild type and � subunit-deficient strainsKinetic properties of the wild-type and the � subunit-deficient PpPfks were deter-mined. All experiments were performed with purified enzymes (see supplementalFig. S7). The constants refer to Equation 3 (see “Experimental Procedures”). Resid-ual Pfk activity was defined as quotient of enzyme activity at 5.0 mM ATP and ofenzyme activity at 0.3 mMATP in absence of any activator. The values are means S.D. of two independent enzyme preparations.
Parameter PpPfk PpPfkWild type �-Deficient
Without effectoraK0.5Fru 6-P (mM) 2.17 0.14 2.29 0.18
nHAMP (mM) 1.60 0.10 1.20 0.10Inhibition by 5 mM ATPc
Residual Pfk activity �0.05 �0.05a Measurement was carried out at 3 mM ATP.b Measurement was carried out at 0.3 mM Fru 6-P and 3 mM ATP.c Data were obtained by measuring at 0.3 mM Fru 6-P without allosteric activator.
TABLE 3Kinetic properties of purified Pfks from P. pastoris wild-type and �subunit-deficient strains calculated with a Monod-Wyman-ChangeuxmodelThe constants refer to Equations 1 and 2 (see “Experimental Procedures”).KS
ATP wasdetermined from kinetic data measured at 6 mM Fru 6-P. All other data wereobtained by measuring at 0.3 mM Fru 6-P. Kinetic constants were calculated fromdata shown in supplemental Fig. S8. Note thatVmax values refer to different enzymepreparations, which could not be compared with each other.
structs yielded only very low quantities of protein. The expres-sion product was also sensitive to proteolytic degradation andaggregated during purification (data not shown).Finally, we performed reconstitution experiments with the
PpPfk subunit deletion mutants according to Klinder et al.(11). Thus, two mutants individually lacking � and �, � and�, or � and � subunits were mixed equally prior to cell dis-
ruption. PpPfk activity was meas-ured in cell-free extracts of thesethree strain combinations (Fig.7C). The Fru 2,6-P2 activation wasalso monitored at 5 mM and 0.3mM ATP (Fig. 7D), because at ATPconcentrations �3 mM the activa-tion rate of the � subunit-deficientPpPfk was higher compared withthe wild-type PpPfk. As expected,individual deletion of either the �subunit or the � subunit was asso-ciated with loss of measurablePpPfk activity (Fig. 7C). However,after mixing the cells of thesemutants followed by combinedcell disruption, residual Pfk activ-ity was measured in the resultingextract. Moreover, the restoredactivity was slightly increased byaddition of Fru 2,6-P2 (Fig. 7D). Asmentioned above, the � subunit isnot essential for the catalytic func-tion of PpPfk. But its deletion altersthe PpPfk sensitivity, e.g. to Fru2,6-P2 (Fig. 7, C and D). The nativePpPfk phenotype with respect toFru 2,6-P2 activationwas restored inthe extract of mixed � subunit-defi-cient cells and cells lacking the � orthe � subunit (both expressing the �subunit) (Fig. 7D). To confirm aPpPfk assembly from the three sub-unit types, the cell-free extracts ofthe mutant combinations were sub-jected to immunoprecipitation withthe � subunit-specific antibody.Immunoprecipitates were analyzedby Western blotting using the anti-�/� subunit antibody (Fig. 7B). Asshown in Fig. 7B (lane 6), an effi-cient �/�/� complex assembly wasobtained in the extract of mixed �subunit- and � subunit-deficientcells. The other two mutant combi-nations revealed only faint signalsfor the precipitated �/� subunits.This is probably the result of thealready low amount of the remain-ing PpPfk subunits in the respectivemutant strains (Fig. 7A). Further,
one can speculate that the � subunit is stabilized by the � sub-unit but not by the � subunit. Consequently, the stabilized �subunit can more efficiently assemble with the high amount ofthe functional �/� complex preformed in the � subunit-defi-cient cells. To verify the complexation of �, �, and � subunits,cell-free extracts of themutant combinations were fractionatedby high-performance liquid chromatography gel filtration.
FIGURE 4. Effect of ATP, AMP, and Fru 2,6-P2 on the activity of Pfks from P. pastoris wild-type and �subunit-deficient strains. ATP-dependent activation of PpPfks from wild-type (closed circles) and � subunit-deficient (open circles) strains was determined at 1 mM AMP (A) and at 20 �M Fru 2,6-P2 (B). In both experiments,Fru 6-P concentration was kept constant at a physiological level of 0.3 mM. At ATP levels � 1 mM, no decreaseof the Fru 2,6-P2 activation ratio was observed for the � subunit-deficient PpPfk due to its less efficient ATPinhibition when compared with the wild-type PpPfk. The figures are based on the data shown in supplementalFig. S8. The activation ratio for AMP (C) and Fru 2,6-P2 (D) was calculated from the enzyme activity in presence(Vx) and absence (V0) of the respective activator (x) at constant Fru 6-P (0.3 mM) and ATP (3 mM) levels. Themutated PpPfk is less sensitive to both AMP and Fru 2,6-P2. Data representing one enzyme preparation areshown.
Phosphofructokinase from P. pastoris
AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23693
Pooled fractions corresponding to the molecular mass of thewild-type PpPfk were concentrated and analyzed by Westernblotting. Similar to the wild-type enzyme (supplemental Fig.S12A), the extract of mixed � subunit- and � subunit-deficientcells showed �/� and � subunit-specific immunoreactivity(supplemental Fig. S12B).
DISCUSSION
Pfk from Pichia Consists of Three Different Subunits—Allknown pro- and eukaryotic ATP-dependent Pfks are, so far,homo- or hetero-multimeric enzymes, which assemble from �subunits and �/� subunits, respectively. Herein, we demon-strate that Pfk from Pichia species displays an even more com-plex structure.We have identified and sequenced a new proteincomponent, which was co-purified together with the � and �subunits of PpPfk (Fig. 1). Further, we provided strong struc-tural and functional evidence that this 40.8-kDa protein isindeed a constitutive third subunit type of PpPfk. First, using anantibody against the � subunit, the � and � subunits of PpPfkwere co-precipitated in a stoichiometric manner (Fig. 6). Sec-ond, partial digestion of the purified native enzyme with chy-
motrypsin led to truncated � and � subunits, which still form astoichiometric complex with the � subunit (Fig. 5). Third,immunofluorescence studies showed identical subcellular dis-tributions for the�/� subunits and the � subunit (supplementalFig. S9). Fourth, clear evidence of association came from thegene deletion studies, which demonstrated that the presence ofthe � subunit in the cytosolic fraction critically depends on theexpression of the � and � subunits (Fig. 7A). Interestingly, thelack of the � subunit diminishes the cytosolic presence of boththe � and � subunits, indicating a subsequent dependence oneach other. Finally, further evidence that the � subunit is a rel-evant part of PpPfk came from reconstitution experiments (Fig.7) and functional analyses showingmodulator properties of the� subunit on Pfk activity (see below). In all our investigations,we found no indication that the � subunit can exist as a freesoluble protein not bound to �/� subunits.The � subunit displays no sequence similarity to known Pfk
subunits and to any other proteins identified yet. Although thisprotein was not found in the sequenced genomes of S. cerevi-siae,K. lactis, S. pombe, andC. albicans, we also identified the �subunit in P. pseudopastoris (Fig. 1). This is indicative of itsdistinct presence in at least some Pichia species. Usually, geneduplication events are accounted for the evolutionary occur-rence of � and � subunits in eukaryotic organisms (5, 8). Asimilar mechanism can be excluded for the introduction of the� subunit in Pichia. Therefore, other mechanisms such as lat-eral gene transfer events (35–37) have to be considered but theorigin of the PFK3 sequence remains open.Next, we addressed the central question of the stoichiometric
enzyme assembly. The molecular masses of the � subunit(108.8 kDa), the � subunit (103.7 kDa), and the � subunit (40.8kDa)were calculated from the respective amino acid sequences.
FIGURE 5. Limited proteolysis of purified PpPfk. Purified PpPfk (150 �g) wasdissolved in 50 mM sodium phosphate buffer containing 5 mM ATP (200 �land pH 7.0). 2 �l of �-chymotrypsin (0.2 mg/ml) were added, and the samplewas incubated at 25 °C for 2 h. Proteolytic activity was stopped by addition ofphenylmethylsulfonyl fluoride (1 mM final concentration), and the samplewas divided into two aliquots. One aliquot was kept for SDS-PAGE, and theother aliquot was subjected to gel filtration on an SE-HPLC BioSelect SEC400column (Bio-Rad) using 50 mM sodium phosphate buffer, pH 7.0, containing150 mM NaCl. Gel filtration yielded a main protein fraction of 900 kDa. Thisfraction was concentrated and analyzed by SDS-PAGE (lane 4) together withthe non-size-fractionated chymotrypsin-modified PpPfk (lane 3) and theundigested PpPfk (lane 2). Precision Plus Protein Unstained Standard (Bio-Rad) was used as molecular mass standard. Proteins were stained with Coo-massie Blue R250.
FIGURE 6. Immunoprecipitation of Pfks from cell-free extracts of P. pas-toris wild-type and the � subunit-deficient strains. Cell-free extracts fromthe P. pastoris wild-type and the � subunit-deficient strains were incubatedwith the anti-� subunit antibody. The resulting immunocomplexes were pre-cipitated with protein-A Sepharose. Samples of purified PpPfk (lane 1, positivecontrol), the cell-free extracts of the wild-type (lane 2), and the � subunit-deficient (lane 3) P. pastoris stains, the supernatants of both extracts afterimmunoprecipitation (lanes 4 and 5), and the respective protein A immuno-precipitates (lanes 6 and 7) were subjected to SDS-PAGE using a 10% polyac-rylamide gel. Separated proteins were blotted, and PpPfk subunits weredetected using polyclonal antibodies against the �/� subunits and againstthe � subunit.
Phosphofructokinase from P. pastoris
23694 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 32 • AUGUST 10, 2007
For the purified PpPfk a molecular mass of 975 28 kDa wasdetermined by sedimentation equilibrium measurements (16).Amolecularmass of850 kDawas estimated for the� subunit-deficient PpPfk by analytical gel filtration. To resolve the ratioof the subunits within the enzyme complex, subunits of thepurified wild-type PpPfk were separated by SDS-PAGE. Coo-massie Blue (Fig. 5; supplemental Fig. S7) and Amido Black gelstaining (data not shown) followed by densitometric evaluationof the bands revealed equal amounts of � and � subunits (each42% of total densitometric signal). Approximately 16% of theprotein corresponded to the � subunit. Sypro Ruby proteinstaining revealed equal amounts for � and � subunits, but the �subunit-specific signal contributed 25% to the total proteinamount (data not shown). The equal density of the � and �subunits suggests an �4�4 complex. Considering all resultslisted above a heterododecameric structure (�4�4�4) can beassumed for PpPfk. This is consistent with the fact that Pfksfrom other yeasts always are symmetric tetra- (�4) andoctameric (�8 or�4�4) enzymes (23, 38–41). However, we can-not exclude a higher number of � subunits within the enzymecomplex (e.g. �4�4�6) because of experimental uncertainties indetermining the subunit ratio by SDS-PAGE. Mammalian Pfkstend to self-associate and form large oligomeric complexes.However, the minimum functional structure was a homotet-
ramer (3, 42). Therefore, Pfk fromPichia is the most complex andprobably the largest Pfk identifiedyet. Our analyses also provided evi-dence of how the individual Pfk sub-units are arranged in the enzymecomplex. In contrast to the � sub-unit only residual amounts of the �subunit were found in the cytosolicfraction when the � subunit wasdeleted (Fig. 7A). Based on thisresult, stabilizing interactionsbetween � and � subunits areunlikely. However, � and � subunitsassemble to an enzymatically activecomplex even in the absence of the�subunit. Cross-linking experimentswith purified PpPfk followed bySDS-PAGE and sequencing ofcross-linked fragments indicatedclose interactions between � and �subunits.3 Recent data for Pfk fromS. cerevisiae suggested that � sub-units form a homotetrameric core,where two pairs of � subunits areassociated peripherally (14, 43, 44).In keepingwith thismodel, Pfk fromP. pastoris appears to be similarlyorganized but four additional � sub-units are attached to the outside ofthis Pfk complex. This assumptionwas supported by our limited prote-olysis experiments of purifiedPpPfk, where only the � subunit
remained unmodified (Fig. 5). Because it is assumed that partialproteolysis of non-denatured proteinswill affect the outer com-ponents of protein complexes first, one can speculate that �subunits are probably located inside the native PpPfk.
� Subunit Is Involved in Adaptation to EnvironmentalChanges—The dependence of the glycolytic flux on carbonsource and on species-specific requirements was analyzed invarious yeasts (45–47). In addition to the regulation of tran-scription and post-translational modification, the allostericmodulation of Pfk and of several other enzymes appears toplay the most important role in controlling the glycolyticflux (48–52). However, cells of the methylotrophic actino-mycete Amycolatopsis methanolica possess a single inor-ganic pyrophosphate-dependent Pfk when grown on glu-cose. However, an ATP-Pfk is induced during cultivation onone-carbon compounds, e.g. methanol (53). Based on theunique PpPfk structure suggested above, one can speculatethat the existence of a particular Pfk species may be of advan-tage for methanol-assimilating yeasts such as Pichia. Inmany organisms Fru 2,6-P2 has been proposed to be the pre-dominant effector of glycolysis (54–56). However, its effect
3 J. Kirchberger and J. Bar, unpublished data.
FIGURE 7. Reconstitution of PpPfk in cell-free extract of mixed subunit deletion mutants. For in vitroreconstitution of PpPfk, two P. pastoris strains individually lacking one of the three subunit types were mixedequally prior to cell disruption. A, cell-free extracts of the indicated individual and mixed P. pastoris strains wereanalyzed by Western blotting using polyclonal antibodies against �/� subunits and against the � subunit ofPfk. B, to demonstrate proper assembly of the PpPfk complex, cell-free extracts of all combinations of mutantstrains were subjected to immunoprecipitation with the � subunit-specific antibody. Pellets from protein-A-based immunoprecipitation were analyzed by Western blotting using polyclonal antibodies against the �/�subunits. C, specific PpPfk activity was monitored at 0.3 mM ATP. D, the activation at 20 �M Fru 2,6-P2 (given as-fold activity over the activity without Fru 2,6-P2) was measured at 0.3 mM ATP (black bars) and 5 mM ATP (whitebars). In both experiments, Fru 6-P concentration was kept constant at a physiological level of 0.3 mM. Data arerepresented as mean S.D. of at least three independent experiments each performed in duplicate.
Phosphofructokinase from P. pastoris
AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23695
on PpPfk is less pronounced and requires higher metaboliteconcentrations in comparison to ScPfk (31).It has been suggested that Pfk and other enzymes of the gly-
colysis can form a complex with aldolase (EC 4.1.2.13) to allowfor efficient substrate channeling (57–59). Although � subunithas no sequence similarities with known aldolases we testedwhether the PpPfk containing the � subunit displays aldolaseactivity. We found no aldolase activity by using the coupledenzyme assay (16) without aldolase as auxiliary enzyme whereeither Fru 6-P or Fru 1,6-P2 served as substrates (data notshown).To evaluate the functional relevance of the � subunit, the
respective coding sequencewas deleted inP. pastoris.The com-parison of the wild-type and � subunit-deficient PpPfksrevealed that the � subunit is not essential for catalytic functionbut significantly modulates enzyme kinetics. The � subunitdeficiency is associated with a lower apparent affinity of theregulatoryATP-binding site. Consequently, theATP inhibitionof the mutated PpPfk is less efficient. Further, the � subunit-deficient PpPfk is characterized by a lower sensitivity to AMPand Fru 2,6-P2. These results suggest that the sensitivity ofPpPfk to ATP inhibition and its reverse by Fru 2,6-P2 and AMPis fine-tuned by the� subunit. Our data support the assumptionthat cellular glucose metabolism in P. pastoris is mainly con-trolled byATP andAMPvia regulation ofPpPfk activity. Atkin-son and co-workers (60) proposed in the “energy charge”hypothesis, that most branch points between anabolism andcatabolismmight be controlled by AMP, ADP, andATP. Refer-ring to PpPfk, glycolysis can be throttled by reducing PpPfkactivity in the situation of plenty of ATP even in the presence ofFru 2,6-P2. The higher ATP sensitivity of the � subunit-con-taining PpPfk may be of advantage under competitive condi-tions. In a situation of ATP depletion, PpPfk is efficient acti-vated by the accompanied increased level of AMP resulting inan enhanced glycolytic flux. Precise regulation of Pfk activityduring cellular adaptation to changes in natural carbon sourcesis particularly important in methylotrophic yeasts like P. pasto-ris. They are often found in pectin-rich environments such asfruit surfaces containing methyl ester compounds (61).Adaptation to environmental changes, in its extreme to
stress, can lead to yeast cell adhesion and formation of macro-scopic flocs protecting cells in the center (62). Flocculationoften occurs upon nutrient limitation during late-exponentialor stationary phase of growth and depends on pH, ethanol lev-els, or the carbon source available in the growth medium (63–65). Likewise, P. pastoris displayed flocculation in exponentialand stationary growth phase (Fig. 3). Interestingly, we foundthat � subunit deficiency diminished cell adhesion. The linkbetween the disturbed flocculation phenotype and the � sub-unit is not solved, yet. Flocculation is conferred by adhesins thatbind sugar residues and specific peptides or increase the cellsurface hydrophobicity. Expression of these special cell wallproteins is under tight control by several interacting regulatorypathways (62). One can speculate that suboptimal Pfk functiondue to � subunit deficiency somehow interferes with properfunction of some of these cell surface components. However, itis very unlikely that the � subunit directly participates in medi-ating flocculation as described for glyceraldehyde-3-phosphate
dehydrogenase from the yeast Kluyveromyces marxianus (66).Adhesion of � subunit-deficient cells can still be induced byaddition of Ca2� but to a lesser extent compared with the wild-type strain (data not shown).In sum, the heterododecameric PpPfk is the most complex
Pfk described yet. The newly identified � subunit is involved infine regulation of the enzymatic activity and yeast flocculation.Deletion of the � subunit reduces yeast growth, as a key markerof cellular fitness. Both sensitive tuning of Pfk activity and floc-culation appear to be relevant for fitness of Pichia cells. The �subunit improves these properties and probably provides anadvantage for adaptation to environmental changes.
Acknowledgments—We thank Pietro Nenoff (Laboratory of MedicalMicrobiology, Moelbis) for biochemical differentiation of P. pastoriswild-type and � subunit-deficient strains as well as Mike Francke(Paul Flechsig Institute, Leipzig) for supplying lipophilic dye FM1-43.We are grateful to the anonymous reviewers for the very constructivecomments and for many suggestions. We thank Klaus Huse (Fritz-Lipmann Institute, Jena), and Michael McLeish (University of Mich-igan, Ann Arbor) for critical reading of the manuscript.
REFERENCES1. Hofmann, E. (1976) Rev. Physiol. Biochem. Pharmacol. 75, 1–682. Hofmann, E., andKopperschlager,G. (1982)Methods Enzymol.90, 49–603. Uyeda, K. (1979) Adv. Enzymol. Relat. Areas Mol. Biol. 48, 193–2444. Poorman, R. A., Randolph, A., Kemp, R. G., and Heinrikson, R. L. (1984)
Nature 309, 467–4695. Heinisch, J., Ritzel, R. G., von Borstel, R. C., Aguilera, A., Rodicio, R., and
Zimmermann, F. K. (1989) Gene 78, 309–3216. Evans, P. R., and Hudson, P. J. (1979) Nature 279, 500–5047. Schirmer, T., and Evans, P. R. (1990) Nature 343, 140–1458. Kemp, R. G., and Gunasekera, D. (2002) Biochemistry 41, 9426–94309. Arvanitidis, A., and Heinisch, J. J. (1994) J. Biol. Chem. 269, 8911–891810. Heinisch, J. (1986) Curr. Genet. 11, 227–23411. Klinder, A., Kirchberger, J., Edelmann, A., and Kopperschlager, G. (1998)
Yeast 14, 323–33412. Evans, P. R., Farrants, G.W., andHudson, P. J. (1981) Philos. Trans. R. Soc.
Lond. B. Biol. Sci. 293, 53–6213. Shirakihara, Y., and Evans, P. R. (1988) J. Mol. Biol. 204, 973–99414. Ruiz, T., Mechin, I., Bar, J., Rypniewski, W., Kopperschlager, G., and
Radermacher, M. (2003) J. Struct. Biol. 143, 124–13415. Yuan, W., Tuttle, D. L., Shi, Y. J., Ralph, G. S., and Dunn, W. A. (1997)
J. Cell Sci. 110, 1935–194516. Kirchberger, J., Bar, J., Schellenberger,W.,Dihazi, H., andKopperschlager,
G. (2002) Yeast 19, 933–94717. Edelmann, A., and Bar, J. (2002) Yeast 19, 949–95618. Schwock, J., Kirchberger, J., Edelmann, A., Kriegel, T. M., and Kopper-
schlager, G. (2004) Yeast 21, 483–49419. Bradford, M. M. (1976) Anal. Biochem. 72, 248–25420. Shevchenko, A., Wilm, M., Vorm, O., and Mann, M. (1996) Anal. Chem.
68, 850–85821. Kirschenbaum, D. M. (1973) Anal. Biochem. 56, 237–26322. Edelmann, A., Kruger, M., and Schmid, J. (2005) J. Clin. Microbiol. 43,
6164–616623. Bar, J., Schellenberger, J., and Kopperschlager, G. (1997) Yeast 13,
1309–131724. Huse, K., and Kopperschlager, G. (1983) FEBS Lett. 155, 50–5425. Pringle, J. R., Adams,A. E., Drubin,D.G., andHaarer, B. K. (1991)Methods
Enzymol. 194, 565–60226. Clifton, D., Weinstock, S. B., and Fraenkel, D. G. (1978)Genetics 88, 1–1127. Holland, M. J., Yokoi, T., Holland, J. P., Myambo, K., and Innis, M. A.
(1987)Mol. Cell. Biol. 7, 813–82028. Dlauchy, D., Tornai-Lehoczki, J., Fulop, L., and Peter, G. (2003) Antonie
Phosphofructokinase from P. pastoris
23696 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 282 • NUMBER 32 • AUGUST 10, 2007
Van. Leeuwenhoek 83, 327–33229. Sakai, Y., Koller, A., Rangell, L. K., Keller, G. A., and Subramani, S. (1998)
J. Cell Biol. 141, 625–63630. Vida, T. A., and Emr, S. D. (1995) J. Cell Biol. 128, 779–79231. Nissler, K., Otto, A., Schellenberger,W., andHofmann, E. (1983)Biochem.
Biophys. Res. Commun. 111, 294–30032. Gancedo, J. M., and Gancedo, C. (1973) Biochimie (Paris) 55, 205–21133. Kopperschlager, G., Kriegel, T. M., and Birkenmeier, G. (1995) FEBS Lett.
363, 17–2134. Bar, J., Golbik, R., Hubner, G., andKopperschlager, G. (2000)Biochemistry
39, 6960–696835. Bapteste, E., Moreira, D., and Philippe, H. (2003) Gene (Amst.) 318,
185–19136. Muller, M., Lee, J. A., Gordon, P., Gaasterland, T., and Sensen, C. W.
(2001) J. Bacteriol. 183, 6714–671637. Slamovits, C. H., and Keeling, P. J. (2006) Eukaryot. Cell 5, 148–15438. Kopperschlager, G., Bar, J., Nissler, K., and Hofmann, E. (1977) Eur. J. Bio-
chem. 81, 317–32539. Reuter, R., Naumann, M., Bar, J., Haferburg, D., and Kopperschlager, G.
(2000) Yeast 16, 1273–128540. Schroter, A., and Kopperschlager, G. (1996) FEMS Microbiol. Lett. 142,
247–25241. Flores, C. L., Martinez-Costa, O. H., Sanchez, V., Gancedo, C., and Ara-
gon, J. J. (2005)Microbiology 151, 1465–147442. Foe, L. G., and Kemp, R. G. (1985) J. Biol. Chem. 260, 726–73043. Kopperschlager, G., Kriegel, T., and Bar, J. (1992) in Protein Interactions
(Visser, H., ed) pp. 345–358, VCH-Verlagsgesellschaft mbH, Weinheim44. Kricke, J., Mayer, F., Kopperschlager, G., and Kriegel, T. (1999) Int. J. Biol.
Macromol. 24, 27–3545. Flores, C. L., Rodriguez, C., Petit, T., and Gancedo, C. (2000) FEMS Mi-
crobiol. Rev. 24, 507–52946. Fraenkel, D. G. (2003) Curr. Opin. Microbiol. 6, 198–20147. Barelle, C. J., Priest, C. L., Maccallum, D. M., Gow, N. A., Odds, F., and
Brown, A. J. (2006) Cell. Microbiol. 8, 961–971
48. Larsson, C., Påhlman, I. I., and Gustafsson, L. (2000) Yeast 16, 797–80949. Moore, P. A., Sagliocco, F.,Wood, R.M., and Brown, A. J. (1991)Mol. Cell.
Biol. 11, 5330–533750. Huse, K., Jergil, B., Schwidop, W. D., and Kopperschlager, G. (1988) FEBS
Lett. 234, 185–18851. Pearce, A. K., Crimmins, K., Groussac, E., Hewlins, M. J. E., Dickinson,
J. R., Francois, J., Booth, I. R., and Brown, A. J. P. (2001)Microbiology 147,391–401
52. Sierkstra, L. N., Verbakel, J. M., and Verrips, C. T. (1992) J. Gen.Microbiol.138, 2559–2566
53. Alves, A.M. C. R., Euverink, G. J.W., Santos, H., and Dijkhuizen, L. (2001)J. Bacteriol. 183, 7231–7240
54. Bartrons, R., Van Schaftingen, E., Vissers, S., and Hers, H. G. (1982) FEBSLett. 143, 137–140
55. Okar, D. A., and Lange, A. J. (1999) BioFactors 10, 1–1456. Hers, H. G., and Van Schaftingen, E. (1982) Biochem. J. 206, 1–1257. Rais, B., Ortega, F., Puigjaner, J., Comin, B., Orosz, F., Ovadi, J., and Cas-
cante, M. (2000) Biochim. Biophys. Acta 1479, 303–31458. Rakus, D., Pasek, M., Krotkiewski, H., and Dzugaj, A. (2004) Biochemistry
43, 14948–1495759. Spivey, H. O., and Ovadi, J. (1999)Methods 19, 306–32160. Shen, L. C., F. L., Walton, G. M., and Atkinson, D. E. (1968) Biochemistry
N. (2000) Appl. Environ. Microbiol. 66, 4253–425762. Verstrepen, K. J., and Klis, F. M. (2006)Mol. Microbiol. 60, 5–1563. Mbawala, A., al Mahmood, S., Loppinet, V., and Bonaly, R. (1990) J. Gen.
Microbiol. 136, 1279–128464. Sampermans, S.,Mortier, J., and Soares, E. V. (2005) J. Appl.Microbiol. 98,
525–53165. Bony, M., Barre, P., and Blondin, B. (1998) Yeast 14, 25–3566. Fernandes, P. A., Sousa, M., and Moradas-Ferreira, P. (1993) Yeast 9,
859–866
Phosphofructokinase from P. pastoris
AUGUST 10, 2007 • VOLUME 282 • NUMBER 32 JOURNAL OF BIOLOGICAL CHEMISTRY 23697
