Investigation of Heterologously Expressed Glucose-6-Phosphate Dehydrogenase Genes in a Yeast zwf1 Deletion.

Glucose-6-phosphate dehydrogenase (G6PD) is a key enzyme of the oxidative part of the pentose phosphate pathway and serves as the major source of NADPH for metabolic reactions and oxidative stress response in pro- and eukaryotic cells. We here report on a strain of the model yeast Saccharomyces cerevisiae which lacks the G6PD-encoding ZWF1 gene and displays distinct growth retardation on rich and synthetic media, as well as a strongly reduced chronological lifespan. This strain was used as a recipient to introduce plasmid-encoded heterologous G6PD genes, synthesized in the yeast codon usage and expressed under the control of the native PFK2 promotor. Complementation of the hypersensitivity of the zwf1 mutant towards hydrogen peroxide to different degrees was observed for the genes from humans (HsG6PD1), the milk yeast Kluyveromyces lactis (KlZWF1), the bacteria Escherichia coli (EcZWF1) and Leuconostoc mesenteroides (LmZWF1), as well as the genes encoding three different plant G6PD isoforms from Arabidopsis thaliana (AtG6PD1, AtG6PD5, AtG6PD6). The plastidic AtG6PD1 isoform retained its redox-sensitive activity when produced in the yeast as a cytosolic enzyme, demonstrating the suitability of this host for determination of its physiological properties. Mutations precluding the formation of a disulfide bridge in AtG6PD1 abolished its redox-sensitivity but improved its capacity to complement the yeast zwf1 deletion. Given the importance of G6PD in human diseases and plant growth, this heterologous expression system offers a broad range of applications.


Introduction
All cells from bacteria to humans have developed mechanisms to cope with oxidative stress caused by reactive oxygen species (ROS), which are a by-product of respiratory metabolism and cause damage to membranes and DNA [1]. The importance of corresponding signaling pathways that elicit a proper cellular response to ROS accumulation, especially in tumor cells, has been underlined by studies ultimately earning last year's award of the Nobel prize for physiology and medicine (see [2], and references therein). Glucose-6-phosphate dehydrogenase (G6PD), a key enzyme of the pentose phosphate pathway (PPP) generates NADPH required to detoxify ROS, but also for methionine biosynthesis [3,4]). In human erythrocytes, G6PD is the exclusive source of NADPH and related to the most common human enzymopathic condition [5,6]. G6PD also plays a pivotal role in cancer cell proliferation, cellular signaling, embryogenesis, apoptosis, neurodegeneration [7][8][9], and cardiac dysfunction in mouse models [10]. The human G6PD gene resides on the X-chromosome and by NADPH and ATP, which reduces the efficiency of the G6PD reaction to less than 1% of its maximal capacity in human erythrocytes [40], would be alleviated by the reduced concentration of NADPH as it is channeled towards the reduction of glutathione and the scavenging of ROS [38]. Within this regulatory system, glycolytic GAPDH appears to be most sensitive towards oxidation at its catalytic cysteine residue, thus creating a highly responsive thiol-switch to reroute metabolism as an additional function [41,42].
We here expressed G6PD genes from several organisms in a zwf1 deletion strain of S. cerevisiae. Besides the opportunity to study the biochemical properties of the encoded enzyme in the context of a heterologous expression system, this provides the basis for the screening of genes for special purposes to be used in synthetic biology and biotechnology. Moreover, it will allow high-throughput screens for the discovery of drugs targeting G6PD. Importantly, in the genetic background of the strain used herein, a zwf1 deletion impairs vegetative growth under standard growth conditions, increases the sensitivity towards oxidative stress, and affects chronological lifespan.
Rich media were based on yeast extract (1% w/v), peptone (2% w/v) from Difco Laboratories Inc., Detroit, MI, USA, with glucose (2% w/v) as a carbon source (YEPD). Synthetic media were prepared with Difco yeast nitrogen base with ammonium sulfate as described in [46], with the addition of amino acids and bases using a mixture provided by MP Biomedicals (Eschwege, Germany; CSM-His-Leu-Trp-Ura) supplemented as required for selection of plasmids or deletion markers, and 2% glucose (w/v) as a carbon source (SCD). A quantity of 200 mg/L of G418 was used for selection of the kanMX marker. Hydrogen peroxide was added at the concentrations indicated to induce oxidative stress. For the preparation of solid media, 1.5% agar was added prior to sterilization. Hydrogen peroxide was added to provoke oxidative stress at a concentration of 1.75 mM. Yeasts were incubated at 30 • C, with constant agitation at 180 rpm for liquid cultures. Tetrad analyses were performed on YEPD plates using zymolyase 20T for digestion of the ascus walls and a Singer Instruments micromanipulator as described in [47]. Plates were incubated for 3 days at 30 • C and colony formation was documented by scanning. Brightness and contrast were adjusted for entire plates using Corel Photo Paint, prior to selection of exemplary four tetrads, each. A minimum of 18 tetrads were examined for each cross.

Serial Drop Dilution Assays
To examine phenotypes of yeast strains in serial drop dilution assays, cells were grown overnight in selective medium. The cultures were diluted to an OD 600 of 0.25 and again grown for 3-5 h. Exponentially growing cells were adjusted to an OD 600 of 0.1 with fresh medium and dilutions from 10 0 to 10 −3 were prepared. A quantity of 3 µL of each dilution was spotted onto plates with the indicated compositions and incubated for 2-3 days at 30 • C. Growth was documented by scanning the plates, adjustment of brightness and contrast of the entire images, and trimming using Corel Photo Paint.

Construction of Deletion Mutants, Cloning, and Tagging of Genes
A zwf1 deletion in S. cerevisiae was obtained by one-step gene replacement [48] by directly introducing a PCR product obtained with the primer pair 16.235/16.236 (Table 1) and pUG6-hh -(a derivative of pUG6 described in [49], in which the two HindIII sites were eliminated each by single nucleotide exchanges) as a template into the diploid strain DHD5, selecting for G418 resistance. The correct substitution of one of the ZWF1 alleles in the heterozygous diploid was verified by PCR with the flanking oligonucleotides 16.232/16.233 (Table 1), and the respective transformant was sporulated. The haploid segregant HOD269-1C (MATa ura3-52 leu2-3,112 his3-11,15 zwf1::kanMX) was obtained from a standard tetrad analysis performed on a YEPD plate [50]. Likewise, a rpe1 deletion, supposed to be synthetically lethal with a zwf1 defect, was constructed by PCR-mediated homologous recombination, using pJJH1287 as a template for amplification of the KlLEU2 marker with the oligonucleotide pair 19.139/19.140 (Table 1). pJJH1287 is a derivative of pUG6 mentioned above, in which the KlLEU2 open reading frame was placed between the TEF2 promoter and terminator sequences, substituting the kanMX resistance marker. Transformants were selected for growth on synthetic medium lacking leucine, verified by PCR with the flanking oligonucleotides 19.137/19.138 (Table 1) and subjected to tetrad analysis to yield haploid segregants carrying the deletion. In order to assess the biological function of heterologously expressed ZWF1 homologs, HOD408 (MATa/MATα ura3-52/ura3-52 leu2-3,112/leu2-3,112 his3-11,15/his3-11,15 RPE1/rpe1::KlLEU2 zwf1::kanMX/ZWF1) was then constructed by crossing. This diploid strain, which is heterozygous for both the zwf1 and the rpe1 deletion, was used as a recipient for transformation with different plasmids carrying URA3 as a selection marker. Transformants were sporulated and subjected to tetrad analysis. Growth of segregants carrying both deletion markers, kanMX and KlLEU2 together with the plasmid-borne URA3 gene, was taken as evidence that the respective heterologous gene can complement the zwf1 defect. In a similar fashion as described for RPE1, the yeast NQM1 gene was substituted by the KlLEU2 marker by a PCR product obtained from pJJH1287 with the primers 19.143/19.144 in the diploid strain DHD5. Transformants verified with the flanking primers 19.141/19.142 were sporulated and subjected to tetrad analysis to obtain a haploid segregant carrying the nqm1 deletion for crossing with the zwf1 deletion.
Plasmids expressing ZWF1 from S. cerevisiae or its homologs from other organisms were obtained by cloning either PCR products (for S. cerevisiae and K. lactis) or synthetic DNA-fragments under the control of a modified constitutive PFK2 promoter into the centromeric plasmid pJJH2064 [51]. Plasmids employed with their pJJH numbers are listed in Table 2, and their complete sequences are available upon request. Specifically, to clone the ZWF1 genes encoding the two yeast enzymes, the one from S. cerevisiae was amplified by PCR with the primer pair 16.234/17.036, and the one from the milk yeast Kluyveromyces lactis with the primer pair 17.309/18.002 (see Table 1 for oligonucleotide sequences), both using genomic DNA prepaired from the respective strains as template. For cloning of heterologous G6PD genes, their coding cDNA sequences were obtained (accession numbers given in Table 2) and synthesized in the yeast codon usage (Thermo Scientific GeneArt synthesis), introducing a unique BamHI site prior to the ATG start and a unique HindIII site following the translation stop codon. Mutations leading to exchanges of specific amino acid residues in the encoded proteins for AtG6PD1 were also introduced by synthesis of DNA fragments with convenient restriction sites for cloning into the original plasmids carrying the respective wild-type genes. All open reading frames were verified by Sanger sequencing (Seqlab, Göttingen, Germany) prior to their introduction into the yeast recipient cells.
For detection in Western blot analyses, 3xHA-tags were introduced by in vivo-recombination into recipient plasmids linearized with HindIII, dephosphorylated with alkaline phosphatase and co-transformed with PCR products obtained from pFA6a-3HA-SkHIS3 [52] as a template for amplification with 17.046 as a primer with homologies to the vector sequences located 3 to the genes and the appropriate primers generating in-frame fusions to the respective ORFs, as listed in Table 1. Nucleotide sequences complementary to the template for PCR reactions are given in capital letters, others in small letters. If the latter comprise more than 35 nucleotides, they provide the region for homologous recombination of the PCR product with the target sequences either in the vector (for 3HA tags) or the yeast genome (for construction of deletions). Restriction sites introduced with the primers into the PCR product are underlined. "for" and "rev" in the name, designating the respective forward and reverse primers for each pair. All forward primers used for tagging with 3HA were employed in conjunction with 2064revHA (17.046) to obtain PCR products suitable for recombination with the target vectors listed in Table 2. AtG6PD1for3HA (18.205) was used to tag the wild-type gene as well as the three mutant variants constructed. For cloning of KlZWF1, the reverse primer was designed to elimate an internal BamHI site of the gene by silent nucleotide exchanges (indicated by the small letters ct in the run of capital letters). "del5" and "del3" designate the forward and reverse primers for deletion of the respective genes.

Enzyme Assays and Western Blot Analysis
Glucose-6-phosphate dehydrogenase activities were determined routinely from yeast crude extracts prepared by breaking of cells with glass beads and centrifugation, as described previously [53]. Alternatively to the preparation of crude extracts from fresh overnight cultures, we verified that specific G6PD activities did not suffer from freezing the washed harvested cells at −20 • C for several days. Glass beads were then added to the frozen pellet and preparation of crude extracts followed the standard protocol from then on. Tris-HCl buffer, pH 7.5, was employed for preparation of crude extracts and enzymatic determinations. G6PD activities were determined by following the kinetics of NADP reduction at a final concentration of 0.4 mM, added from a 40 mM stock solution in extraction buffer, in a DU800 spectrophotometer (Beckman-Coulter, Krefeld, Germany) at 340 nm and 30 • C. Glucose-6-phosphate was added as a substrate at a final concentration of 2 mM, after determination of the extinction difference obtained by crude extract in the absence of the substrate. For treatment with dithiothreitol (DTT), 2 µL of a 100 mM stock solution was added to 200 µL of crude extract and incubated for 20 min at room temperature prior to enzymatic determinations. Likewise, diamide was added to a 200 µL aliquot of crude extract at a final concentration of 0.5 mM for 20 min at room temperature to provide an oxidative environment. All chemicals used for enzyme tests were obtained from Sigma-Aldrich Chemie GmbH, Munich, Germany.
For Western blot analyses, crude extracts prepared as above were mixed with SDS sample buffer (final concentrations: 60 mM Tris/HCl, pH 6.8, 10% glycerol, 2% SDS, 0.005% Bromphenol Blue) and heated to 98 • C for 5 min, prior to separation on precast 10% SDS-polyacrylamide gels (LTF-Labortechnik GmbH and Co KG Wasserburg, Germany). Blotting and preparation for immunological detection was done as explained in [54]: after completion of electrophoresis, gels were transferred to a nitrocellulose membrane (Whatman GmbH, Dassel) with the "Trans-Blot SD Semi-Dry Electrophoretic Transfer Cell" (Bio-Rad). Membranes were blocked for 1 to 3 h at room temperature with TBST (20 mM Tris, 150 mM NaCl, pH adjusted to 7.6 with HCl, 0.05% Tween 20) containing 3% bovine serum albumin (BSA, Roth, Karlsruhe, Germany). 3xHA-labelled G6PD enzymes were detected with an anti-HA mouse antibody (kindly provided by Anja Lorberg, Osnabrück), which was diluted 1:10000 in TBST with 3% BSA. An anti-mouse antibody from goat coupled to IRDye 800CW (Li-Cor Biosciences, Lincoln, NE, USA; diluted 1:5000 in TBST with 3% BSA) was used as a secondary antibody. As an internal loading control, the phosphofructokinase subunits were detected using an anti-PFK polyclonal antiserum from rabbit at a dilution of 1:10.000 [55], and an anti-rabbit IgG antibody from donkey coupled to IRDye 700DX (Rockland Immunochemicals, Gilbertsville, PA, USA, diluted 1:5000 with 3% BSA in TBST) as a secondary antibody. The Odyssey infrared imaging system was used for detection and quantification of signals (LI-COR Biosciences GmbH, Bad Homburg, Germany). For quantification, G6PD signals were determined relative to the control PFK signals. ScZwf1 (G6PD from S. cerevisiae) was set at 100% and normalized signals from the heterologous enzymes were calculated as percentages of this value.

Determination of Chronological Life Span
Precultures of cells of strain HD269-1C carrying CEN/ARS plasmids with the genes indicated were grown overnight at 30 • C in 2.5 mL of SCD medium lacking uracil for plasmid selection to reach an early stationary phase. They were diluted 200-fold in 25 mL of fresh medium in 100 mL Erlenmeyer flasks and further incubated with agitation (180 rpm/min) for 2-3 days until they reached the stationary phase. Samples were then taken for a total of two weeks at the times indicated, appropriately diluted to yield 50-500 viable cells per 100 µL and plated onto YEPD. The number of viable cells per mL of culture was then calculated from colony counts (colony forming units, CFU) after 3 days of incubation at 30 • C. CFUs determined at day 0 were set at 100% for each culture. CFUs obtained at each time point were then related to this value. Each clone was inoculated in triplicate and mean percentages of survival and standard deviations were calculated from these biological replicates.

A Yeast System for Heterologous Expression of G6PD Genes
In order to establish a versatile expression system for G6PD genes from organisms spanning the kingdoms of bacteria, fungi, plants, and animals, we obtained a deletion mutant in the sole S. cerevisiae ZWF1 gene, substituting its open reading frame (ORF) for a kanMX deletion cassette in one allele of the wild-type diploid strain DHD5 ( Figure 1A). To assess the growth phenotypes of haploid segregants carrying the zwf1 deletion, the diploid was subjected to tetrad analysis and spores were allowed to germinate and grow on rich medium with 2% glucose as a carbon source. In contrast to previous reports, we noted a distinct growth retardation associated with the zwf1 deletion in this strain background. Thus, a 2:2 segregation for large and small colonies was observed, with the smaller colonies invariably carrying the zwf1 deletion allele, as confirmed by their resistance towards G418 ( Figure 1B). This phenotype was not rescued by the addition of methionine to the rich medium, for which zwf1 deletions have been reported to be auxotrophic ( Figure 1B). One segregant from such an analysis, HOD269-1C, was then chosen for further investigations described below.  show growth after replica-plating onto rich medium supplemented with 200 mg/L of G418 (geneticin). Six representative tetrads are shown for each analysis, out of a total of at least 18 tetrads analyzed.
(C) Western blot analysis of crude extracts from strain HOD269-1C (zwf1::kanMX) carrying the genes indicated on a CEN/ARS vector. The upper picture shows the results of the blot analyzed with the Odyssey scanner with anti-PFK shown in red and anti-HA detecting the tagged G6PD enzymes in green. Columns in the lower picture show the quantification of G6PD signals normalized to the amount of PFK detected for each lane in the Western blot, setting ScZwf1 at 100%.

Heterologous G6PD Genes from Different Biological Kingdoms Complement a Yeast zwf1 Deletion
The zwf1 mutant thus constructed was then used as a recipient for heterologous expression of genes encoding G6PD isozymes throughout the biological kingdoms, as listed in Table 2. After the preparation of crude extracts from strains carrying the genes on the CEN/ARS vector pJJH2064 under the control of the constitutive PFK2 promoter, specific G6PD activities were determined and demonstrated the functional expression of genes from the closely related yeast Kluyveromyces lactis, the bacteria Escherichia coli and Leuconostoc mesenteroides, different isozymes of the plant Arabidopsis thaliana, and the human G6PD gene ( Table 2). The considerable variations in the specific activities of the heterologous enzymes raised the questions of whether i) the genes were properly expressed, and ii) the encoded proteins were stably produced in the yeast zwf1 deletion. Therefore, all encoded proteins were equipped with a C-terminal 3xHA-tag and subjected to a Western blot analysis, using the heterooctameric yeast glycolytic enzyme phosphofructokinase as a loading control and reference for quantification [55]. It should be noted that specific G6PD activities were not notably affected by the C-terminal tags as compared to those reported in Table 2, whereas the introduction of the same tags at the N-terminus caused a complete loss of enzymatic activity in all cases. As shown in Figure 1C, the C-terminally tagged heterologous enzymes of all but the plant enzymes were produced at approximately half the levels of the native ScZwf1-3HA, with the exception of the one from L. mesenteroides, which reached only 20% (protein levels of strains expressing the A. thaliana enzymes are discussed below in Section 2.3).
We proceeded by determining the physiological consequences of heterologous G6PD gene expression in the yeast zwf1 deletion. For this purpose, the hypersensitivity of the deletion towards oxidative stress exerted by the addition of hydrogen peroxide to the medium was employed [30,31,56]. Drop dilution assays confirmed that viability of the deletion carrying the expression vector lacking any G6PD sequence was strongly impaired in the presence of hydrogen peroxide as compared to the control plate without the stressor or the strain carrying the plasmid with the wild-type ZWF1 gene (Figure 2). Complementation ability of the heterologous genes largely correlated with the specific activities measured in the crude extracts of the respective strains, with KlZWF1 and HsG6PD1 displaying growth like the wild-type, and the AtG6PD genes and EcZWF1 conferring intermediate viability.
As an exception, the gene from L. mesenteroides appears to lack complementation despite the detectable specific activity in crude extracts.
Microorganisms 2020, 8,546 of the native ScZwf1-3HA, with the exception of the one from L. mesenteroides, which reached 20% (protein levels of strains expressing the A. thaliana enzymes are discussed below in Sectio We proceeded by determining the physiological consequences of heterologous G6PD expression in the yeast zwf1 deletion. For this purpose, the hypersensitivity of the deletion to oxidative stress exerted by the addition of hydrogen peroxide to the medium was emp [30,31,56]. Drop dilution assays confirmed that viability of the deletion carrying the expression lacking any G6PD sequence was strongly impaired in the presence of hydrogen perox compared to the control plate without the stressor or the strain carrying the plasmid with the type ZWF1 gene (Figure 2). Complementation ability of the heterologous genes largely corr with the specific activities measured in the crude extracts of the respective strains, with KlZWF HsG6PD1 displaying growth like the wild-type, and the AtG6PD genes and EcZWF1 conf intermediate viability. As an exception, the gene from L. mesenteroides appears to complementation despite the detectable specific activity in crude extracts. Drop dilution assays of a yeast zwf1 deletion strain expressing heterologous G6PD gene Logarithmically growing cultures of strain HOD269-1C (zwf1::kanMX) carrying the genes indicate on a CEN/ARS vector were diluted in fresh selective medium to an OD600 = 0.1 and subjected to ten fold serial dilutions as indicated below.A quantity of 3 µL of each dilution was dropped onto th plates indicated and growth was documented after 3 days incubation at 30°C.
In order to obtain a more stringent phenotype for the assessment of heterologously exp genes, we further confirmed that the zwf1 deletion showed synthetic lethality with an rpe1 de in the genetic background of the strain used herein. Therefore, HOD269-1C (zwf1::kanMX crossed to an isogenic HD56-5A derivative, in which RPE1 was substituted by a heterologous m from Kluyveromyces lactis (rpe1::KlLEU2), and subjected to tetrad analysis. While segregants resulting diploid HOD408 showing resistance to G418 and segregants prototrophic for leucine readily obtained, double deletions displaying both markers did not produce any viable pr ( Figure 3A). In contrast, a combination of zwf1 with a nqm1 deletion, which lacks a para transaldolase of unknown function and was also reported in the Saccharomyces genome da (https://www.yeastgenome.org; searched on November 7, 2019) to be synthetically lethal, cou be confirmed, since segregants carrying the double deletions produced viable progeny (Figu HOD408 was then used as a recipient for plasmids carrying some of the heterologous gene transformants were sporulated and again subjected to tetrad analyses ( Figure 3C). Viabi segregants carrying both deletion markers (zwf1::kanMX rpe1::KlLEU2) in conjunction with UR an indicator of the presence of the respective plasmid confirmed that the wild-type ZWF1 (pJJH2111) from S. cerevisiae restored growth to the double deletion, as did the human gene (HsG on plasmid pJJH2223) and AtG6PD1 (pJJH2125). The latter produced smaller colonies in accor Logarithmically growing cultures of strain HOD269-1C (zwf1::kanMX) carrying the genes indicated on a CEN/ARS vector were diluted in fresh selective medium to an OD600 = 0.1 and subjected to ten-fold serial dilutions as indicated below.A quantity of 3 µL of each dilution was dropped onto the plates indicated and growth was documented after 3 days incubation at 30 • C.
In order to obtain a more stringent phenotype for the assessment of heterologously expressed genes, we further confirmed that the zwf1 deletion showed synthetic lethality with an rpe1 deletion in the genetic background of the strain used herein. Therefore, HOD269-1C (zwf1::kanMX) was crossed to an isogenic HD56-5A derivative, in which RPE1 was substituted by a heterologous marker from Kluyveromyces lactis (rpe1::KlLEU2), and subjected to tetrad analysis. While segregants of the resulting diploid HOD408 showing resistance to G418 and segregants prototrophic for leucine were readily obtained, double deletions displaying both markers did not produce any viable progeny ( Figure 3A). In contrast, a combination of zwf1 with a nqm1 deletion, which lacks a paralog of transaldolase of unknown function and was also reported in the Saccharomyces genome database (https://www.yeastgenome.org; searched on November 7, 2019) to be synthetically lethal, could not be confirmed, since segregants carrying the double deletions produced viable progeny ( Figure 3B). HOD408 was then used as a recipient for plasmids carrying some of the heterologous genes and transformants were sporulated and again subjected to tetrad analyses ( Figure 3C). Viability of segregants carrying both deletion markers (zwf1::kanMX rpe1::KlLEU2) in conjunction with URA3 as an indicator of the presence of the respective plasmid confirmed that the wild-type ZWF1 gene (pJJH2111) from S. cerevisiae restored growth to the double deletion, as did the human gene (HsG6PD1 on plasmid pJJH2223) and AtG6PD1 (pJJH2125). The latter produced smaller colonies in accordance with its limited capacity to complement the zwf1 deletion. Neither of the respective segregants grew on medium containing 5-FOA, an antagonist toxic to strains with a wild-type URA3 gene, confirming that growth depended on the presence of the plasmid. In contrast, no viable segregants of the double deletions were obtained from diploids transformed with pJJH2064 (empty vector) or pJJH2495 (LcZWF1).
Microorganisms 2020, 8,546 11 of 2 with its limited capacity to complement the zwf1 deletion. Neither of the respective segregants grew on medium containing 5-FOA, an antagonist toxic to strains with a wild-type URA3 gene, confirming that growth depended on the presence of the plasmid. In contrast, no viable segregants of the doubl deletions were obtained from diploids transformed with pJJH2064 (empty vector) or pJJH2495 (LcZWF1).  rpe1 deletion A) or a nqm1 deletion B) and subjected to tetrad analyses on rich medium. Growth was documented after 3 days of incubation at 30 °C. Colored circles and squares highlight single-and double-deletion mutants as indicated. C) The heterozygous diploid strain HOD408 (ZWF1/zwf1::kanMX RPE1/rpe1::KlLEU2) was transformed with CEN/ARS plasmids carrying the genes indicated, sporulated and subjected to tetrad analyses on rich medium (upper pannels). Growth was documented after 3 days of incubation at 30°C, before replica-plating onto the indicator media shown in the lower pannels. Growth there was documented after 1-2 days of incubation at 30 °C. Red circles indicate viable double deletions complemented by the plasmid-encoded gene.

Mutational Analysis of Plant G6PD1
Results on complementation of the slow-growth phenotype of the yeast zwf1 mutant and of th specific enzyme activities reported above indicated that G6PD1 from A. thaliana lacking its plastidi signal sequence (see legend of Table 2 for details) is barely functional in the heterologous expression system. Since this might be attributed to an inefficient formation of disulfide bridges in the cytoplasm of yeast as opposed to chloroplasts in plant cells, we intended to substitute the two cysteine residue involved by a salt bridge, i.e., the two cysteines were substituted by an aspartate and a lysine residue respectively (AtG6PD1 C100D/C108K ). Indeed, the respective plasmid was able to rescue the growth defec Figure 3. Analysis of synthetic lethality of a yeast zwf1 deletion and complementation by heterologously expressed genes. Strain HOD269-1C (zwf1::kanMX) was crossed to isogenic strains carrying either a rpe1 deletion (A) or a nqm1 deletion (B) and subjected to tetrad analyses on rich medium. Growth was documented after 3 days of incubation at 30 • C. Colored circles and squares highlight single-and double-deletion mutants as indicated. (C) The heterozygous diploid strain HOD408 (ZWF1/zwf1::kanMX RPE1/rpe1::KlLEU2) was transformed with CEN/ARS plasmids carrying the genes indicated, sporulated and subjected to tetrad analyses on rich medium (upper pannels). Growth was documented after 3 days of incubation at 30 • C, before replica-plating onto the indicator media shown in the lower pannels. Growth there was documented after 1-2 days of incubation at 30 • C. Red circles indicate viable double deletions complemented by the plasmid-encoded gene.

Mutational Analysis of Plant G6PD1
Results on complementation of the slow-growth phenotype of the yeast zwf1 mutant and of the specific enzyme activities reported above indicated that G6PD1 from A. thaliana lacking its plastidic signal sequence (see legend of Table 2 for details) is barely functional in the heterologous expression system. Since this might be attributed to an inefficient formation of disulfide bridges in the cytoplasm of yeast as opposed to chloroplasts in plant cells, we intended to substitute the two cysteine residues involved by a salt bridge, i.e., the two cysteines were substituted by an aspartate and a lysine residue, respectively (AtG6PD1 C100D/C108K ). Indeed, the respective plasmid was able to rescue the growth defect of the yeast zwf1 mutant in the presence of hydrogen peroxide much better than the one with the original AtG6PD1 clone, as demonstrated by drop dilution assays ( Figure 4A). However, to our surprise, the increased complementation capacity was also observed for clones in which only the first cysteine residue was substituted for an alanine (AtG6PD1 C100A ), or in a mutant with an alanine-lysine combination (AtG6PD1 C100A/C108K ), not expected to form a salt bridge. This result indicated that it is rather the formation of a disulfide bridge than its absence, which is deleterious for the function of the enzyme in yeast. This cannot be attributed to a stabilization of the mutant proteins in vivo, since they display slightly decreased steady-state levels as compared to the wild-type, as judged from their 3xHA-tagged variants in a Western blot analysis ( Figure 4B).    Figure 1C, setting the amount of ScZwf1 to 100%. C) Sensitivity of different G6PD enzymes produced in the yeast zwf1 deletion strain towards reductive (DTT) and oxidative (diamide) conditions. Specific enzyme activities were either directly determined after preparation of crude extracts (yellow columns) or after pre-treatment of the same crude extracts with 1 mM DTT  Figure 1C, setting the amount of ScZwf1 to 100%. (C) Sensitivity of different G6PD enzymes produced in the yeast zwf1 deletion strain towards reductive (DTT) and oxidative (diamide) conditions. Specific enzyme activities were either directly determined after preparation of crude extracts (yellow columns) or after pre-treatment of the same crude extracts with 1 mM DTT (green columns) or 0.5 mM diamide (blue columns). Activities were at least determined in triplicate, with the error bars giving the standard deviations of these technical replicates.
In order to assess the influence of the redox environment on the activity of the AtG6PD1 variants and the other enzymes from A. thaliana investigated herein, we determined the specific activities in crude extracts treated with either dithiothreitol (DTT) or diamide as reductive or oxidative conditions, respectively. As expected from previous data on the plant enzyme [17], AtG6PD1 activity depended on its redox state and its activity decreased significantly when incubated with DTT, while those of AtG6PD5 and AtG6PD6 did not ( Figure 4C). The yeast G6PD, which was employed as a control, was insensitive to DTT treatment, too. Treatment with diamide generally led to a moderate reduction in specific activities, indicating that the stability of all enzymes in the yeast crude extracts was slightly affected.

Chronological Life Span of S. cerevisiae Is Drastically Reduced by the zwf1 Deletion and Restored by Heterologous Expression of Functional ZWF1 Homologs
In the course of inoculating the zwf1 deletion strain carrying the vector pJJH2064 without an insertion for several controls, we noticed that they did not survive prolonged periods of storage at 4 • C, in contrast to those carrying the wild-type ScZWF1 gene. Therefore, the chronological life span of a strain lacking its native glucose-6-phosphate dehydrogenase was investigated and compared to a wild-type strain carrying the vector, and the deletion strain carrying eiher an episomal native ScZWF1 gene, or the ones encoding the human G6PD1 or the isoform 1 of A. thaliana. As evident from Figure 5, the deletion strain lacking G6PD activity rapidly lost viability and formed less than 1% of the colonies counted for the original stationary phase culture within four days of incubation. On the other hand, all other transformants showed a life span comparable to wild-type cells, displaying more than 80% viability after 4 days and taking at least 11 days to decrease to less than 1%. However, it should be noted that the lifespan of the clone carrying the AtG6PD1 gene appeared to be somewhat reduced as compared to the ones carrying either ScZWF1 or the human homolog. 5, the deletion strain lacking G6PD activity rapidly lost viability and formed less than 1% of the colonies counted for the original stationary phase culture within four days of incubation. On the other hand, all other transformants showed a life span comparable to wild-type cells, displaying more than 80% viability after 4 days and taking at least 11 days to decrease to less than 1%. However, it should be noted that the lifespan of the clone carrying the AtG6PD1 gene appeared to be somewhat reduced as compared to the ones carrying either ScZWF1 or the human homolog. Figure 5. Chronological lifespan of a yeast zwf1 deletion expressing heterologous G6PD genes. The indicated plasmids were introduced into strain HOD269-1C (zwf1::kanMX) and cultures were grown to stationary phase in synthetic selective medium. Samples were taken at the time points indicated, appropriately diluted to produce single colonies, plated on rich medium and incubated for three days prior to determination of colony-forming units (CFU). CFUs were set at 100% at day 0, with starting counts indicated at the upper-right corner, and relative viability was determined in three independent cultures, each. Error bars give the standard deviations of % survivors in these three biological replicates.

Discussion
In this work, we constructed a deletion mutant of the gene ZWF1, encoding the unique glucose-6-phosphate dehydrogenase enzyme (G6PD) in the yeast Saccharomyces cerevisiae to assess its suitability as a host for heterologous expression of G6PD homologs from various organisms throughout the biological kingdoms, in order to study their molecular structure and biochemical Figure 5. Chronological lifespan of a yeast zwf1 deletion expressing heterologous G6PD genes. The indicated plasmids were introduced into strain HOD269-1C (zwf1::kanMX) and cultures were grown to stationary phase in synthetic selective medium. Samples were taken at the time points indicated, appropriately diluted to produce single colonies, plated on rich medium and incubated for three days prior to determination of colony-forming units (CFU). CFUs were set at 100% at day 0, with starting counts indicated at the upper-right corner, and relative viability was determined in three independent cultures, each. Error bars give the standard deviations of % survivors in these three biological replicates.

Discussion
In this work, we constructed a deletion mutant of the gene ZWF1, encoding the unique glucose-6-phosphate dehydrogenase enzyme (G6PD) in the yeast Saccharomyces cerevisiae to assess its suitability as a host for heterologous expression of G6PD homologs from various organisms throughout the biological kingdoms, in order to study their molecular structure and biochemical properties. Unexpectedly, we noted a distinct growth retardation of haploid deletion strains on standard media. This stands in contrast to data on zwf1 deletions in most other yeast strains, which were reported to grow as wild-type under such conditions [30][31][32]56]. However, the zwf1 deletion was lethal in another common background of strain Σ1278b [57], underlining the substantial degree of nucleotide polymorphisms observed between commonly employed yeast strains [45]. The intermediate phenotype of the deletion strain used herein thus combines the advantage of being viable with that of reflecting the complementation capacity of heterologously expressed G6PD genes directly by the size of the colonies produced upon transformation. A more stringent test for complementation of growth phenotypes is provided by the hypersensitivity of zwf1 deletions towards oxidative stress, which is shared by our mutant with all viable strains tested in the works cited above. Moreover, we confirmed the synthetic lethality of the zwf1 deletion with that of the RPE1 gene [58] in our strain background as yet another means to test for complementation capacity. On the other hand, zwf1 nqm1 double deletions did not excerbate the growth retardation of single zwf1 deletions, i.e., they did not display a synthetic defect. This could be explained by the role of the ribulose-5-phosphate-3-epimerase encoded by RPE1 in the non-oxidative part of the pentose phosphate pathway, thus rendering both parts of the pathway non-functional in the zwf1 rpe1 double deletion. We, therefore, conclude that the PPP serves essential functions in S. cerevisiae, which can be met by either the oxidative or the non-oxidative part through their connections with glycolysis, but not if the two branches are blocked. Accordingly, NQM1 is supposed to encode a transaldolase of unknown function, which obviously is not required for the non-oxidative PPP, were the TAL1-encoded transaldolase works.
Another phenotype observed for the zwf1 deletion employed in this work is the drastic reduction in chronological lifespan (CLS), which is also alleviated by the expression of heterologous G6PD genes from humans and plants. This phenotype is most likely related to the fact that the production of reactive oxygen species (ROS) is a primary cause of ageing from yeast to humans [59], and the G6PD reaction provides the NADPH for its detoxification [3,4]. That CLS has not been assessed for zwf1 deletions in yeast until now may be attributed to the lack of growth phenotypes under standard conditions discussed above. Our findings thus also provide the basis to study the contribution of G6PD enzymes to ageing in the genetically easily tractable yeast model. Although all heterologously expressed G6PD genes herein complemented the yeast zwf1 deletion, they did so in various degrees, judging from the two parameters of specific enzyme activity and growth of the transformants in the presence of hydrogen peroxide. These two parameters did not always correlate. For instance, the three isoforms of AtG6PD barely conferred growth to the respective transformants in the presence of oxidative stress, whereas the bacterial enzyme from L. mesenteroides did not. Yet, the latter had specific activities that were at least double those of the plant enzymes measured in crude extracts from such transformants. On the other hand, specific activities of G6PD from E. coli were in the lower range of those of the plant isozymes, but transformants grew almost like wild-type controls in the presence of hydrogen peroxide. Differences in the growth complementation assays could also not be explained by varying stabilities of the enzymes produced in the yeast recipient strain, since Western blots showed higher steady-state levels for the HA-tagged plant isozymes, which did not complement well, than for the enzyme from the closely related milk yeast K. lactis, which did. We conclude that the cytosolic environment within the yeast cell substantially affects the actual in vivo activity of the heterologously produced G6PD chloroplast proteins, as observed in plants [18]. Clearly, the enzymes from human and yeasts work best, followed by the one from E. coli and the plant isozymes. The lack of complementation by the other bacterial G6PD, LmZwf1, despite its high specific activity in crude extracts can be attributed to its special cofactor dependence. While all the other enzymes tested herein exclusively use NADP as a cofactor, the enzyme from L. mesenteroides is less specific and can also employ NAD [60,61]. Since S. cerevisiae does not have a transhydrogenase which could interconvert the reduced forms of these cofactors [62], we assume that redox balance under oxidative stress is severely impaired by the production of this rather unusual heterologous enzyme. The fact, that transformants of the yeast zwf1 deletion with LmZWF1 produce even smaller colonies than those with the empty vector indicates that the heterologous enzyme exerts a negative effect already under standard growth conditions.
With regard to the AtG6PD1 isoform of A. thaliana, we reasoned that it may not properly form the disulfide bridge in the yeast cytosol, which appears to be crucial in plant chloroplasts for its activity in the dark phase [16]. Although the attempt to substitute the disulfide by an ionic bridge appeared to improve the complementation capacity of the enzyme at first glance, further mutant analyses indicated that the reason for this was rather the lack than the presence of the bridge. As the wild-type AtG6PD enzyme displayed higher steady-state levels in the Western blots of HA-tagged variants, the improved complementation capacity cannot be attributed to enhanced protein stability in the yeast host. Since specific activities in crude extracts were also not drastically changed by the mutations, it must be the interplay of enzyme structure and activity in the yeast cytosol in vivo that causes better performance. Nevertheless, the observation that the activity of the wild-type AtG6PD1 isoform can be influenced by applying reducing conditions to the crude extracts, whilst the mutant enzymes incapable of forming a disulfide bridge do not react, underlines the suitability of the yeast expression system to assess the biochemical properties of single heterologously produced plant isoforms.
Finally, it should be noted that the manipulation of the ZWF1 gene in S. cerevisiae receives increasing attention for several biotechnological applications. Thus, the deletion of the gene improved ethanol production from xylose in a recombinant yeast strain, which was attributed to an increase in NADH production at the expense of NADPH [63]. Vice versa, overexpression of ZWF1 increased the resistance of a production strain towards furfural, an inhibitor present in the fermentation of lignocellulosic substrates [64]. The reaction is also required for the use of S. cerevisiae in fermentations to cope with the ROS produced in the presence of inhibitory phenolic compounds [65]. In addtion, overproduction of ZWF1 in combination with other genetic manipulations was employed to improve the production of isoprenoids and carotenoids in yeast [66,67], and a zwf1 deletion carrying the E. coli pyruvate dehydrogenase complex was found to accumulate cytosolic acetyl-CoA and thus may serve as an important platform strain for many biosynthetic processes, including the production of biobutanol [68]. In the context of such biotechnological uses, the substitution of the endogenous yeast G6PD by heterologous enzymes with altered biochemical properties, as demonstrated in this work, could well be an approach in metabolic design strategies.

Conlcusions
We demonstrated that the zwf1 deletion obtained herein serves as a valuable expression system for heterologous G6PD genes in S. cerevisiae, especially when the presence of different isozymes impedes the biochemical analysis in the original host organism. Moreover, the relation of G6PD to oxidative stress response and ageing can be more conveniently studied in the extremely well-established model yeast. We, therefore, believe that this system will have broad applications not only in biotechnology, but also in medicine and agriculture.