Yap1-Regulated Glutathione Redox System Curtails Accumulation of Formaldehyde and Reactive Oxygen Species in Methanol Metabolism of Pichia pastoris

Strains and media. P. pastoris PPY12 (arg4 his4) (32) was used as the wild-type strain. Escherichia coli DH10B was routinely used for plasmid propagation.

The P. pastoris strains were grown on either YPD medium (2% glucose, 1% Bacto yeast extract, 2% Bacto peptone) or YNB medium (0.67% yeast nitrogen base without amino acids) supplemented, when required, with an appropriate amino acid(s) (200 μg/ml for arginine, 200 μg/ml for histidine) or Zeocin (50 μg/ml; Invitrogen, Carlsbad, CA). The following were used as carbon sources in YNB medium: 2% (wt/vol) glucose (SD), 2% (vol/vol) methanol (SM), and 0.5% (vol/vol) oleate (SO). Tween 80 was added to the medium containing oleate at a concentration of 0.05% (vol/vol). We used 2% methanol to clearly detect the phenotypic differences between the wild type and gene disruptants, although similar differences could also be observed at lower methanol concentrations (e.g., 0.7%). For organelle fractionation, YPM medium (2% methanol, 1% Bacto yeast extract, 2% Bacto peptone) was used. The initial pH of the medium was adjusted to 6.0. Cultivation was performed aerobically at 28°C with reciprocal shaking, and the growth of the yeast cells was monitored by measuring the optical density at 610 nm (OD₆₁₀).

E. coli was grown at 37°C in LB medium (1% Bacto tryptone, 0.5% Bacto yeast extract, 0.5% NaCl) supplemented with ampicillin (50 μg/ml).

Construction of disruption vectors and one-step disruption of P. pastoris genes.A 1.5-kb upstream region of the PpYAP1 gene was amplified by PCR with primers PpYAP1-up-NotI and PpYAP1-up-PstI (Table 1) with genomic DNA as the template. Analogously, a 1.4-kb downstream region of the PpYAP1 gene was amplified with primers PpYAP1-down-PstI and PpYAP1-down-XhoI. The PCR products were recombined as NotI-PstI and PstI-XhoI fragments, respectively, with a 4.9-kb NotI-XhoI fragment of plasmid SK⁺-Zeo^r which included the Zeocin resistance (Zeo^r) gene (26), yielding the PpYAP1 disruption vector pDyap1. After being linearized with PstI, pDyap1 was transformed into the wild-type strain by electroporation. Zeocin-resistant colonies were selected on SD medium supplemented with arginine, histidine, and Zeocin. Disruption of the PpYAP1 gene was confirmed by Southern blot analysis with EcoRI-digested genomic DNA of transformants and a 0.7-kb ClaI-NaeI fragment from the downstream region of the PpYAP1 gene as the probe.

The isogenic Ppcta1Δ, Ppfld1Δ, Ppfgh1Δ, Ppfdh1Δ, Ppglr1Δ, Ppgpx1Δ, and Pppmp20Δ mutant strains were generated by replacing the respective gene coding regions with the fragment of plasmid SK⁺-Zeo^r as described above.

Expression of PpGlr1 and PpYap1 in P. pastoris.To visualize the localization of PpGlr1, a strain expressing a yellow fluorescent protein (YFP)-tagged PpGlr1 (PpGlr1-YFP) fusion protein was constructed as follows. First, a fragment with the 1-kb 5′ untranslated region and the coding region of PpGLR1 was obtained by PCR with primers Pglr1-EcoRI-Fw and PpGLR1-SpeI-Rv (Table 1) by using genomic DNA as the template. Next, YFP was amplified by PCR with primers YFP-SpeI-Fw and YFP-SphI-Rv by using vector pEYFP-N1 (Clontech) as the template. These two fragments were cloned into promoter-free integration expression vector pIB1 (36). The resultant plasmid was linearized with NcoI and then introduced into the Ppglr1Δ mutant strain.

To visualize the localization of PpYap1, a strain expressing a DsRed-monomer fluorescent protein (mRed)-PpYap1 fusion protein was constructed as follows. First, with genomic DNA as the template, a 1.4-kb SpeI-SphI fragment containing the PpYAP1 coding region and a 0.8-kb NdeI-EcoRI fragment with the PpYAP1 promoter (P_YAP1) region were amplified by PCR with primer sets PpYAP1-SpeI-Fw/PpYAP1-SphI-Rv and Pyap1-NdeI-Fw/Pyap1-EcoRI-Rv, respectively. Then, a 0.7-kb EcoRI-SpeI fragment with the mRed coding region was amplified by PCR with primers mRed-EcoRI-Fw and mRed-SpeI-Rv by using vector pDsRed-Monomer-N1 (Clontech) as the template. These three fragments were cloned into pIB1. The resultant plasmid was linearized with SalI and transformed into P. pastoris strains.

Quantitative real-time PCR (qRT-PCR).Cells were grown aerobically to mid-exponential phase in SM or SD medium. Total RNAs were isolated from cells with an RNeasy mini kit (Qiagen) monitored by on-column DNase digestion. cDNAs were synthesized from 1 μg total RNA with Random Primers (Promega) and ReverTra Ace (Toyobo). After reverse transcription for 50 min at 42°C, samples were heated for 5 min at 99°C to terminate the reaction and 0.5 μl of RNase H was added. qRT-PCR was performed in 20-μl mixtures in glass capillary tubes in a LightCycler (Roche Diagnostic). The PCR cycling reaction was performed with 1× SYBR Premix Ex Taq (Takara) according to the following parameters: first cycle, 10 min of denaturation at 95°C; second cycle with 40 repetitions, 95°C for 20 s, 55°C for 20 s, and 72°C for 20 s (all temperature transitions, 20°C s⁻¹). The primers used for the reactions are listed in Table 1. As a negative control, PCR was done without ReverTra Ace. The relative abundance of mRNAs was standardized against the levels for PpACT1.

Measurement of intracellular glutathione and enzyme assays.The amounts of GSH and GSSG were determined as described by Tietze (41). The activities of glutathione reductase, catalase, and mitochondrial cytochrome c oxidase were measured as described previously (2, 38, 42). Proteins were quantified by a Bradford (3) assay kit (Bio-Rad) with bovine serum albumin as the standard. Cells were grown aerobically to mid-exponential phase in SM, SD, or SO medium.

Subcellular fractionation.For subcellular fractionation experiments, wild-type cells were grown aerobically to mid-exponential phase in YPM medium. Preparation of spheroplasts and cell homogenates, as well as the subsequent differential and Histodenz density gradient centrifugation, were performed as described by Monosov et al. (25).

Fluorescence microscopy and image acquisition. P. pastoris cells carrying YFP- and cyan fluorescent protein (CFP)-tagged proteins were incubated from a starting OD₆₁₀ of 0.5 in SM, SD, or SO medium and labeled with MitoTracker Red CMXRos to visualize mitochondria. Nuclear staining was performed as follows. Cells grown on SM medium were harvested, washed once, and fixed with 1 ml of 70% ethanol for 30 min at room temperature. Fixed cells were then washed, resuspended in 15 μl sterilized water, and stained with 150 μl of 0.125 μg/ml DAPI (4′,6′-diamidino-2-phenylindole; Dojin) solution. After 10 min of incubation, cells were observed under an inverted fluorescence microscope (IX70; Olympus) equipped with a UPlan Apo 100×/1.35 oil iris objective lens with a U-MWIG mirror/filter unit (Olympus) for DsRed-monomer and MitoTracker Red CMXRos (Molecular Probes), a U-MWU2 mirror/filter unit (Olympus) for DAPI, a U-MF2 (Olympus) filter set (XF114-2; Omega Optical, Inc.) for CFP, and a U-MF2 filter set (XF104-2; Omega Optical, Inc.) for YFP. Image data were captured by a charged-coupled device camera (SenSys; PhotoMetrics) with MetaMorph. 6.0 and saved as Photoshop files (Adobe). The TIFF image files were optimized for contrast in Photoshop CS3 and compiled on PowerPoint (Microsoft).

Nucleotide sequence accession numbers.The sequences of the PpYAP1, PpCTA1, PpPMP20, PpGLR1, PpGPX1, PpFGH1, and PpFDH1 genes have been submitted to DDBJ and assigned accession numbers AB472084, AB472085, AB472086, AB472087, AB472088, AB472089, and AB472090, respectively.

Increase in glutathione levels and change in thiol redox ratio during growth on methanol.Not only methanol but also oleate is known to induce peroxisomes in P. pastoris (34). To investigate the role of glutathione in relation to peroxisome metabolism, the total amounts of glutathione (GSH plus GSSG) were compared among cells grown on methanol, oleate, or glucose as a single carbon source. The total amount of glutathione was more than twofold higher in cells grown on methanol (SM) than in cells grown on glucose (SD) or oleate (SO) (Fig. 2A). In methanol-grown cells, the intracellular GSH/GSSG ratio, the best index of the cellular thiol redox balance (29), was 70 to 60% lower than that in glucose- or oleate-grown cells (Fig. 2B), indicating that the intracellular redox state of thiols in methanol-grown cells is more toward their oxidized form. Since oleate metabolism did not alter the glutathione redox state by more than 15%, the intracellular redox regulation observed during methanol metabolism does not seem to be a general feature of peroxisome metabolism but appears specific to methanol metabolism. Therefore, glutathione metabolism during growth on methanol was studied further.

• Open in new tab
• Download powerpoint

FIG. 2.

Effects of carbon sources on the glutathione system. Wild-type P. pastoris was grown on methanol (SM), glucose (SD), or oleate (SO) medium to the mid-exponential phase. (A) Intracellular levels of total glutathione (GSH plus GSSG). (B) Intracellular glutathione redox ratio (GSH/GSSG). (C) Relative transcriptional levels of genes involved in the glutathione redox system on methanol compared to those on glucose. The relative abundance of these mRNAs was standardized against the levels for PpACT1. Results are means ± standard deviations (n = 3).

Adaptation to methanol culture elevated the total amount of glutathione and decreased the intracellular GSH/GSSG ratio. From the P. pastoris genome sequence database, we identified the PpGLR1, PpGPX1, and PpGSH1 genes by homology searches. While three glutathione peroxidase genes (ScGPX1, ScGPX2, and ScGPX3) were found in S. cerevisiae (16), only one was found in P. pastoris (PpGPX1), as in Schizosaccharomyces pombe (SpGPX1) (49). Next, we examined the transcriptional levels of PpGLR1, PpGPX1, and PpGSH1 during growth on methanol by qRT-PCR analysis. Among them, PpGPX1 was found to be the most highly induced during growth on methanol (Fig. 2C).

Growth properties of gene disruptants on methanol.We hypothesized that mutants with the glutathione redox system deleted would show a growth defect on methanol. It has been reported that GSH-deficient mutants of the methylotrophic yeast Hansenula polymorpha failed to grow on methanol due to the accumulation of toxic amounts of HCHO (37).

To investigate the role of the glutathione redox system and its putative regulator PpYap1, the growth on methanol of PpGLR1, PpGPX1, PpPMP20, and PpYAP1 mutant strains was compared. The Pppmp20Δ mutant strain completely lost the ability to grow on methanol, as observed with the Cbpmp20Δ mutant strain (15; Fig. 3A). The lag phase of the Ppglr1Δ, Ppgpx1Δ, and Ppyap1Δ mutant strains was significantly prolonged. The Ppglr1Δ mutant strain showed the most severe growth defect. On the other hand, all of the strains tested grew well on glucose or oleate (data not shown).

• Open in new tab
• Download powerpoint

FIG. 3.

Phenotypes of strains carrying disruptions of genes involved in the glutathione redox system. (A) Growth properties of the wild type (WT) and the Ppglr1Δ, Ppgpx1Δ, Pppmp20Δ, and Ppyap1Δ disruptants on methanol (SM). The intracellular total glutathione levels (B) and the intracellular GSH/GSSG ratios (C) in the wild type and the Ppglr1Δ, Ppgpx1Δ, Pppmp20Δ, and Ppyap1Δ disruptants grown on methanol (SM), glucose (SD), or oleate (SO) medium are also shown. Results are means ± standard deviations (n = 3). (D) Growth of the wild type and the Ppglr1Δ disruptant on minimal methanol (SM, left panel) or glucose (SD, right panel) medium not supplemented or supplemented with 50 μg/ml GSH or GSSG.

Intracellular redox states of gene disruptants.Cells were cultivated to exponential phase, and the intracellular GSH/GSSG ratio, together with the total amount of glutathione, was analyzed. The total amount of glutathione was twofold higher and the GSH/GSSG ratio was decreased by 30 to 60% in cells of all of the disruptants tested when they were grown on methanol compared to those of cells cultured on glucose or oleate (Fig. 3B and C). During growth on methanol, all of the disruptants tested showed lower GSH/GSSG ratios than wild-type cells did. Especially in the Ppglr1Δ mutant strain, the GSH ratio was drastically decreased. No glutathione reductase activity was detected in the Ppglr1Δ mutant strain (see Fig. S1 in the supplemental material), indicating that there is only one gene that encodes Glr in P. pastoris and that GSSG could not be reduced to GSH by another pathway. Interestingly, the total amount of glutathione in the Ppglr1Δ mutant strain was significantly higher than in other strains, suggesting that de novo GSH synthesis in the Ppglr1 disruptant partially complemented the lack of regeneration of GSH from GSSG. The long lag phase of this strain when transferred to methanol (Fig. 3) indicated that this compensatory mechanism took a while to become fully active under the new growth conditions. Indeed, when Ppglr1Δ mutant cells in mid-exponential phase on methanol were diluted to fresh methanol medium, the lag phase was shorter for the second culture (data not shown), suggesting a higher level of GSH at the exponential phase of methanol-glucose-grown Ppglr1Δ mutant cells. We tested whether the growth defect of the Ppglr1Δ mutant cells on methanol could be complemented by adding GSH or GSSG to the culture medium (Fig. 3D, left panel). The retarded growth of the Ppglr1Δ mutant was partially complemented by the presence of GSH (50 μg/ml) but not by the presence of GSSG (50 μg/ml). In the presence of GSH (200 μg/ml), the growth of the Ppglr1Δ mutant strain was comparable to that of wild-type cells (data not shown). On the other hand, addition of GSH (50 μg/ml) did not affect its growth on glucose (Fig. 3D, right panel). These results indicate that GSH, not GSSG, plays a critical role during methanol metabolism in methylotrophic microorganisms.

Function of GSH during methanol metabolism.We next asked whether the growth defect of the gene disruptants mainly came from the failure to detoxify either HCHO or ROS. First, the HCHO levels in the culture medium of the gene disruptants were determined (Fig. 4A). In the wild-type strain, HCHO accumulation peaked after 20 to 30 h of incubation and then gradually decreased. The maximum HCHO levels in the Ppgpx1Δ, Ppyap1Δ, and Ppglr1Δ mutant strains were two- to fourfold higher than in wild-type cells, with the highest peak level in the Ppglr1Δ mutant. Disruption of PpPMP20, which encodes a glutathione peroxidase specifically involved in the detoxification of ROS but not in that of HCHO, resulted in an almost constant high level of HCHO. As shown in Fig. 4B, this high HCHO level is owing to the remarkably small Pppmp20Δ mutant cell numbers, since this strain was inviable on methanol due to ROS generated through methanol metabolism. In contrast, the Ppglr1Δ mutant strain, which also had a severe growth defect on methanol, was hypersensitive to HCHO, suggesting that GSH, which has been found to function mainly in the detoxification of HCHO during growth on methanol (51), is not sufficiently abundant to sustain viability under these growth conditions. This is in line with the high HCHO levels (Fig. 4A) and low GSH concentrations (Fig. 3C) detected in this strain.

• Open in new tab
• Download powerpoint

FIG. 4.

HCHO sensitivity. (A) HCHO level in culture medium during growth on methanol. Samples were obtained from the culture medium of Fig. 3A. (B) Cell viability on HCHO or methanol. Cells were incubated in YNB medium containing HCHO (2 mM) or methanol (2%) for the indicated times, spotted onto an SD plate, and grown for 2 days at 28°C. WT, wild type.

Intracellular localization of the glutathione regeneration system in P. pastoris.Our present data indicate that Glr1-catalyzed regeneration of GSH has a critical role in HCHO detoxification during methanol metabolism. On the other hand, the glutathione peroxidase activity of PpPmp20 requires GSH within peroxisomes for the detoxification of ROS (14). In other yeasts, such as S. cerevisiae and S. pombe, glutathione reductase showed a bimodal distribution between mitochondria and the cytosol (27, 39). In addition, this enzyme was reported to be localized in peroxisomes of pea (Pisum sativum) leaves (7, 18). Therefore, we examined the localization of PpGlr1 during the methylotrophic growth of P. pastoris. We expressed PpGlr1 C-terminally tagged with YFP (PpGlr1-YFP) under the control of its native promoter (P_GLR1) in the Ppglr1Δ mutant strain and compared the fluorescence to that obtained for a CFP-tagged peroxisomal marker, SKL (25). PpGlr1-YFP localized to the mitochondria and the cytosol during growth on methanol (Fig. 5A), glucose, or oleate (data not shown). The localization of PpGlr1 was also analyzed in the wild-type strain by subcellular fractionation, which separated the intracellular components into a cytosolic supernatant and an organelle pellet consisting mainly of peroxisomes and mitochondria. The glutathione reductase activity colocalized with the mitochondrial marker protein cytochrome c oxidase (4) on a Histodenz gradient (Fig. 5B) but not with peroxisomal catalase. Previously, we showed that methanol-induced peroxisomes purified from the methylotrophic yeast Candida boidinii also did not exhibit any detectable glutathione reductase activity (14). Therefore, we found no evidence of the presence of a glutathione regeneration system within peroxisomes of methylotrophic yeasts.

• Open in new tab
• Download powerpoint

FIG. 5.

Glutathione reductase localizes to mitochondria and cytosol in P. pastoris. (A) Strains carrying PpGlr1-YFP and CFP-labeled peroxisomes were analyzed during growth on methanol. The merged images comprise PpGlr1 (green) and peroxisome (Ps, blue) signals and mitochondrion (Mt, red) signals stained with MitoTracker Red CMXRos (Molecular Probes). The right panel is the differential interference contrast (DIC) image. Bar, 2 μm. (B) The organellar pellet obtained after centrifugation at 27,000 × g was further fractionated by Histodenz equilibrium density gradient centrifugation. Fractions were assayed for relative catalase (○), cytochrome c oxidase (•), and glutathione reductase (▴) activities.

PpYap1 transiently localized to the nucleus and positively regulated the expression of genes related to the glutathione redox system during adaptation to methanol.In S. cerevisiae, the nuclear enrichment of transcription factor ScYap1 in response to oxidative stress (21, 50) resulted in the transcriptional activation of genes that encode enzymes related to the glutathione redox system, such as ScGLR1, ScGPX2 (one of the three glutathione peroxidase genes), and ScGSH1, as well as numerous genes that are involved in preventing oxidative damage to the cells (17, 19, 35, 40, 43). To investigate the role and behavior of Yap1 in P. pastoris, we have cloned the PpYAP1 gene.

We expressed a DsRed-monomer-PpYap1 fusion protein (mRed-PpYap1) under the control of the native promoter (P_YAP1) in the Ppyap1Δ mutant strain and confirmed this by Western blot analysis (data not shown). The Ppyap1Δ mutant strain was hypersensitive to various oxidative stresses, and under oxidative stress conditions, mRed-PpYap1 accumulated in the nucleus (T. Yano and Y. Sakai, unpublished results). These experiments confirmed that activation of mRed-PpYap1 can be monitored in real time by its accumulation in the nucleus, as observed in S. cerevisiae.

The Ppyap1Δ mutant strain also exhibited a prolonged lag phase (Fig. 3A) and accumulated a considerable amount of HCHO (Fig. 4A), suggesting that PpYap1 is involved in the detoxification of HCHO during growth on methanol. To see the activation of PpYap1 under physiological conditions, the mRed-PpYap1-expressing Ppyap1Δ mutant cells were grown to exponential phase in glucose medium (SD) and then shifted to methanol or oleate medium. mRed-PpYap1 was predominantly localized to the cytosol before the shift. At 3 to 14 h after a shift to methanol medium, mRed-PpYap1 was transiently localized to the nucleus (Fig. 6A and B). In contrast, mRed-PpYap1 stayed in the cytosol during growth on glucose or oleate (data not shown).

• Open in new tab
• Download powerpoint

FIG. 6.

mRed-PpYap1 is transiently localized to the nucleus and positively regulates the transcription of genes related to the glutathione redox system during methanol metabolism. (A) Subcellular localization of mRed-PpYap1. The Ppyap1Δ mutant strain expressing mRed-PpYap1 was grown in glucose (SD) medium to exponential phase and shifted to methanol (SM) medium. (B) Nuclear accumulation of mRed-PpYap1 during growth on methanol. mRed-PpYap1-expressing cells incubated on methanol for 6 h were stained with DAPI. Bars, 2 μm. (C) Relative transcriptional levels of PpGLR1, PpGPX1, and PpGSH1 in the Ppyap1Δ mutant compared to those in the wild type. The relative abundance of these mRNAs was standardized against the levels for PpACT1. Results are means ± standard deviations (n = 3).

This transient nuclear localization of PpYap1 in methanol-grown cells appears to activate the expression of genes related to the glutathione redox system (PpGLR1, PpGPX1, and PpGSH1); as in the Ppyap1Δ mutant, the transcriptional activation of these genes was suppressed (Fig. 6C). The promoters of these genes contain various putative Yap1 response elements (13; data not shown). These results are consistent with the finding that the total glutathione amount and the GSH/GSSG ratio were lower in the Ppyap1Δ mutant strain during growth on methanol than in the wild-type strain (Fig. 3B and C). The retarded growth of the Ppyap1Δ mutant strain on methanol and HCHO accumulation may be due to the retarded expression of these genes, especially PpGLR1.

Activation of PpYap1 in various mutants.Since we identified a novel physiological role for PpYap1 in HCHO detoxification, we studied how PpYap1 is activated during adaptation to methanol in various gene disruptants. mRed-PpYap1-expressing strains were precultured in glucose medium (SD) and shifted to methanol medium (SM containing 0.05% yeast extract to support the growth of the mutants) to monitor the localization of mRed-PpYap1 periodically (time, 1 to 24 h).

First, we tested mutations affecting the GSH-dependent HCHO oxidation pathway (Ppfld1Δ, Ppfgh1Δ, and Ppfdh1Δ). In cells with these mutations, the intracellular HCHO level was expected to increase without enhancing the generation of ROS (Fig. 1), which made the Ppfld1Δ and Ppfgh1Δ mutants (but not the Ppfdh1Δ mutant) hypersensitive to HCHO in the medium (see Fig. S2B in the supplemental material). As shown in Fig. 7, in both the Ppfld1Δ and Ppfgh1Δ mutant strains, mRed-PpYap1 accumulated in the nucleus at earlier time points than in the wild-type strain when these strains were grown on methanol. This nuclear PpYap1 localization could not be released in Ppfld1Δ and Ppfgh1Δ mutant cells growing on glucose medium to which HCHO was added (see Fig. S2A in the supplemental material). Moreover, both the Ppfld1Δ and Ppfgh1Δ mutant strains showed HCHO sensitivity but the Ppfdh1Δ mutant strain did not (see Fig. S2B in the supplemental material). These results indicate that HCHO toxicity or the deficiency of GSH triggers the nuclear localization of PpYap1 in the Ppfld1Δ and Ppfgh1Δ mutant strains and suggested activation of PpYap1 by HCHO. We also confirmed that HCHO itself could induce the nuclear accumulation of mRed-PpYap1 (Yano and Sakai, unpublished).

• Open in new tab
• Download powerpoint

FIG. 7.

Subcellular localization of mRed-PpYap1 in various mutants on methanol. Mutants carrying mRed-PpYap1 were grown in glucose (SD) medium to exponential phase and then shifted to methanol (SM) medium containing yeast extract (0.05%). Bar, 2 μm. WT, wild type.

P. pastoris has two peroxisomal antioxidant enzymes, i.e., peroxisomal catalase Cta1 and Pmp20 glutathione peroxidase; a homologue of cytosolic catalase CTT1 could not be found in the P. pastoris genome. The growth defect of the Pppmp20Δ mutant was more severe on methanol medium than that of the Ppcta1Δ mutant (Fig. 4B; see Fig. S2C in the supplemental material), as was the case for their C. boidinii homologues (15). In the Pppmp20Δ mutant strain, mRed-PpYap1 localized to the nucleus 1 to 2 h after a shift to methanol medium, whereas this took another hour in Ppcta1Δ mutant and wild-type cells (Fig. 7). Thus, increased levels of peroxisomal ROS resulting from the lack of PpPmp20 activity cause a more rapid activation of PpYap1 than H₂O₂ accumulation, which could only be detected in the absence of PpCta1 (data not shown). HCHO by itself did not cause a growth defect in the Pppmp20Δ mutant, but methanol as a carbon source did (Fig. 4B). It is the conversion of methanol to HCHO that causes the generation of ROS as a by-product (Fig. 1), which is thus most damaging to Pppmp20Δ mutant cells. Taken together, our results show that the absence of PpPmp20 made the cells highly susceptible to oxidative stress.