Expression level of methanol-inducible peroxisomal proteins and peroxisome morphology are affected by oxygen conditions and mitochondrial respiratory pathway function in the methylotrophic yeast Candida boidinii


Correspondence: Tomoyuki Nakagawa, Faculty of Applied Biological Sciences, Gifu University, 1-1 Yanagido, Gifu 501-1193, Japan. Tel.: +81-58-293-2927;

fax: +81-58-293-2927;



In the methylotrophic yeast, Candida boidinii, methanol-inducible peroxisomal proteins, for example alcohol oxidase (AOD), dihydroxyacetone synthase (DAS), and peroxisomal glutathione peroxidase (Pmp20), were induced only under aerobic conditions, while expression of PMP47 encoding peroxisomal integral membrane protein Pmp47 was independent of oxygen conditions. Expression of the methanol-inducible peroxisomal enzymes was repressed by inhibition of the mitochondrial respiratory chain. In the respiratory-deficient (ρ0) mutant strain, their induction was at very low levels despite the presence of oxygen, whereas the expression of PMP47 was unaffected. Taken together, these facts indicate that C. boidinii can sense oxygen conditions, and that mitochondrial respiratory function may have a profound effect on induction of methanol-inducible gene expression of peroxisomal proteins. Peroxisome morphology was also affected by oxygen conditions and respiratory function. Under hypoxic conditions or respiration-inhibited conditions, cells induced by methanol contained small peroxisomes, indicating that peroxisome biogenesis and the protein import machinery were not affected by oxygen conditions but that peroxisome morphology was dependent on induction of peroxisomal matrix proteins.


Methylotrophic yeasts are capable of growing on methanol as their sole carbon and energy source. In these yeasts, the first step of methanol metabolism is the oxidation of methanol by alcohol oxidase (AOD), which is localized in peroxisomes. As AOD is strongly induced by methanol until AOD proteins comprise 20% to 30% of the total cellular protein, heterologous gene expression systems using the strong methanol-inducible gene promoters, such as the AOD gene promoter, were established in various methylotrophic yeast strains, including Candida boidinii, Pichia pastoris, Hansenula polymorpha, and Pichia methanolica (Raymond et al., 1998; Sakai et al., 1999; Cregg et al., 2000; Gellissen, 2000; Nakagawa et al., 2006). With these expression systems, large quantities of a number of useful proteins have been produced, for example, enzymes, antibodies, cytokines, plasma proteins, and hormones (Gellissen, 2000; Yurimoto & Sakai, 2009).

We have studied the molecular mechanism of methanol-dependent gene expression in detail to improve the gene expression system using C. boidinii (Yurimoto, 2009; Yurimoto & Sakai, 2009; Yurimoto et al., 2011). Recently, we identified two transcriptional factors, Trm1p and Trm2p, which are responsible for methanol-inducible gene expression in C. boidinii (Sasano et al., 2008, 2010). Trm1p is a transcription factor responsible for methanol-specific gene activation, while Trm2p is a regulator of derepression in C. boidinii. Both proteins are necessary for activation of the methanol-inducible promoters, that is, PAOD1, PDAS1, PPMP20, and PPMP47 (Sasano et al., 2010).

Oxygen is required by methylotrophic yeast for the AOD-catalyzed reactions within peroxisomes for methanol metabolism in addition to its role as the terminal electron acceptor for the respiratory chain in the mitochondria. Therefore, the yeast cells must have a mechanism for controlling the methanol metabolism, which responds to oxygen levels, to balance the oxygen needs between these two organelles. Indeed, the P. methanolica AOD isozymes, which are octameric enzymes consisting of two different subunits with different affinities for oxygen, Mod1p, and Mod2p (Nakagawa et al., 1996, 1999, 2001), are induced only under aerobic conditions, with the ratio of the two subunits in the AOD octamer depending on oxygen concentration and respiratory chain activity (Nakagawa et al., 2006; Fujimura et al., 2007). We believe that the regulation of methanol metabolism responding to oxygen conditions is an important system common to all methylotrophic yeast strains, and that maintenance of the oxygen consumption balance between peroxisomes and mitochondria is one of the key points for regulation of methanol metabolism. Therefore, analysis of the regulation of the methanol metabolism responding to oxygen conditions has potential for applications to improve the heterologous gene expression system in methylotrophic yeast.

In this study, we showed that expression of the methanol-inducible genes in C. boidinii was dependent on oxygen conditions and demonstrated that their expression was directly influenced by respiratory activity. Finally, we showed that peroxisomal morphology was affected by oxygen conditions, and the size of the peroxisome was determined by the quantity of peroxisomal matrix proteins.

Materials and methods

Yeast strains and cultivation conditions

The Cboidinii strains used in this study are listed in Table 1. Candida boidinii strain S2 (Tani et al., 1985) was used as the wild-type strain. Strains GFP-PTS1/wt and GFP-PTS1/aod1∆ were used for the study of peroxisomal morphology. Strains GFP-PTS1/wt and GFP-PTS1/aod1∆ harbored the GFP-PTS1 expression vector, pGFP-PTS1, which was constructed by introducing the GFP gene tagged with AKL sequence as a peroxisome targeting signal 1 at the carboxy terminus (GFP-PTS1) under the actin promoter of pACT1 (Sakai et al., 1998b), in the genome of Cboidinii TK62 and Cboidinii aod1ura3 strains (Nakagawa et al., 1999), respectively. The respiratory-deficient (ρ0) mutant of strain S2, strain S2ρ0, was constructed by treatment of the wild-type strain with ethidium bromide (Rickwood et al., 1988).

Table 1. Candida boidinii strains used in the study
S2Tani et al., (1985)
S2-ρ0ρ0This study
TK62 ura3 Sakai et al. (1991)
APPAOD1-ScPHO5 (URA3) ura3Yurimoto et al. (2000)
DPPDAS1-ScPHO5 (URA3) ura3Yurimoto et al. (2000)
20PPPMP20-ScPHO5 (URA3) ura3Yurimoto et al. (2000)
47PPPMP47-ScPHO5 (URA3) ura3Yurimoto et al. (2000)
GFP-PTS1/wtPACT1-GFP-PTS1 (URA3) ura3Sakai et al. (1998a, b)
GFP-PTS1/aod1PACT1-GFP-PTS1 (URA3) ura3 aod1This study

Strains AP, DP, 20P, and 47P, which contain the PHO5 gene under the control of methanol-inducible promoters, PAOD1, PDAS1, PPMP20, and PPMP47 respectively, were used in the promoter–reporter assay (Yurimoto et al., 2000; Sasano et al., 2008).

YP (0.5% yeast extract and 1% peptone) and synthetic salt media (Sakai et al., 1998a, b) were used for cultivation of C. boidinii. The following carbon sources were used: 1% glucose, 1% glycerol, and 1% methanol. Methanol induction of these strains was performed by transferring YP-glucose (YPD)-grown cells to YP-methanol (YPM) medium at an optical density at 610 nm of 1.0 and subsequently incubating them for 16 h. Aerobic cultivation was performed at 28 °C at 150 r.p.m. Hypoxic cultivation was performed in static cultures, using screw-capped tubes flushed with nitrogen.

Inhibition test for gene expression by respiratory inhibitors

Sodium azide (NaN3) (Wako Pure chemical industries, Ltd., Osaka, Japan) was used as respiratory inhibitor. The inhibition test was performed 16 h after addition of NaN3 into the incubation flasks. Final concentration of NaN3 in the induction media was 0.0005%.

Preparation of cell-free extracts and Western blot analysis

Candida boidinii cells grown on several conditions were suspended in a fivefold volume of 50 mM sodium phosphate buffer (pH 7.5) and transferred to a 2-mL microtube containing an equal volume of 0.5-mm-diameter zirconium beads (Biospec Products, Inc, Bartlesville, OK). The tubes were shaken vigorously for 30 s on a 3110BX mini-beadbeater (Biospec Products, Inc) and chilled on ice for 30 s. This procedure was repeated 3 times, and the cell debris was removed by centrifugation at 16 000 g for 5 min at 4 °C.

Western analysis was performed as described by Towbin et al. (1979) using the Amersham ECL detection kit (Arlington Heights, IL). Standard 10% SDS–polyacrylamide gels with separating gels of pH 9.2 were employed. The VA9 monoclonal anti-Pmp20 antibody, IVA7 monoclonal anti-Pmp47 antibody, and G358 polyclonal anti-AOD antibody were kindly provided by Dr Goodman J.M. (University of Texas, Southern Medical Center at Dallas, TX). Information of the anti-DAS antibody was previously described (Sakai et al., 1996).

The protein concentrations of cell-free extracts were determined by the method of Bradford with a protein assay kit (Bio-Rad Laboratories, Hercules, CA) using bovine serum albumin as the standard.

Northern analysis

Northern analysis was performed as previously described (Nakagawa et al., 2004). Total RNA was extracted from the cells using the ISOGEN reagent (Nippon Gene Co, Ltd, Toyama, Japan), and 10 μg of the RNA sample was electrophoresed on a 1.0% agarose gel containing 20 mM morpholine propane sulfonic acid (MOPS) buffer, 1 mM EDTA, and 2.2 M formaldehyde. After electrophoresis, the RNAs were transferred to a nylon membrane (Hybond-N+; Amersham Pharmacia Biotech UK Ltd, Buckinghamshire, UK) by capillary action in 20 × SSC. The DNA probes containing the entire coding regions of AOD1, DAS1, PMP20, and PMP47 were each labeled with the AlkPhos DIRECT kit (Amersham Pharmacia Biotech UK Ltd).

Acid phosphatase assay for promoter activities of methanol-inducible genes

The promoter activities of AOD1, DAS1, PMP20, and PMP47 genes were measured by acid phosphatase (AP) activity of strains, AP, FLP, 20P, and 47P (Yurimoto et al., 2000). AP activity was assayed as described by Toh-e et al. (1973).

Fluorescent microscopic analysis of mitochondrial membrane potential and GFP

Wild-type and S2ρ0 cells were grown on YPD medium. Cells were washed twice with 0.85% NaCl and stained with 200 nM MitoTracker Green FM (Molecular Probes Inc, Eugene, OR) to assess mitochondrial membrane potential (Tani et al., 2008).

Cells expressing the GFP fusion proteins and stained with MitoTracker Green FM were placed on a microscope slide and examined with an ECLIPSE 80i (Nikon Co, Tokyo, Japan) microscope equipped with a Plan APO 100x/1.40 (oil) objective and Nomarski attachments and set on the fluoresce-in-isothiocyanate channel. Images were acquired using a charge-coupled device camera, Penguin 600 CL (Pixera Co, Los Gatos, CA).


Induction of peroxisomal methanol-metabolizing proteins is affected by oxygen conditions, while PMP47 was not

We had previously demonstrated that expression of the genes for peroxisomal methanol-metabolizing proteins, AOD1, DAS1, PMP20, and PMP47, was induced by methanol in C. boidinii, while PMP47 was slightly expressed in the glucose medium (Yurimoto et al., 2000). However, we have never determined whether the expression of these genes was oxygen dependent. In this study, we examined the effect of oxygen conditions on the induction of these methanol-inducible genes at the translational and transcriptional levels. Even under hypoxic conditions, C. boidinii cells could grow slowly on YPM medium (data not shown). Under aerobic conditions, proteins of AOD, DAS, and Pmp20, which are peroxisomal matrix proteins, were strongly induced by methanol. However, they were not induced under hypoxic conditions, even when methanol was the sole carbon source (Fig. 1a and b). Induction patterns of AOD, DAS, and Pmp20 proteins coincided with expression patterns of gene promoters for AOD1, DAS1, and PMP20 (Fig. 1a and b). In contrast, the PMP47 expression was obviously induced by methanol even under hypoxic conditions (Fig. 1a–c). Therefore, the cells grown on YPM medium under hypoxic conditions were able to produce sufficient doses of ATP to synthesize mRNAs and proteins. These results indicate that methanol-inducible genes can be classified into two groups, one group for genes that are significantly affected by oxygen, including AOD1, DAS1, and PMP20, and another group that is little affected by oxygen, including PMP47.

Figure 1.

Induction and expression of methanol-inducible enzymes by methanol under aerobic and hypoxic conditions. (a) Western analysis. α-AOD, anti-AOD antibody; α-DAS, anti-DAS antibody; α-Pmp20, anti-Pmp20 antibody; and α-Pmp47, anti-Pmp47 antibodies. Lane 1, aerobic conditions. Lane 2, hypoxic conditions. (b) Relative activities of AP. Closed bars, cells induced by methanol under aerobic conditions (oxygen absorption rate, 16.0 mM min−1L−1); open bars, cells induced by methanol under hypoxic conditions; and gray bar, cells grown on glucose. The specific activities of AP under aerobic conditions are set to 100%. Data are the averages of results from three independent experiments. Specific activities of AP in strains AP, DP, 20P, and 47P grown on methanol under aerobic conditions are 21.4 ± 8.01, 48.6 ± 4.98, 6.59 ± 0.35, and 0.96 ± 0.329 U per OD660 nm, respectively. (c) Time course test of AP induction in strain 47P grown on methanol under hypoxic condition. The maximal level of AP activity (12 h) was set to 100%.

Suppression of mitochondrial function decreases oxygen-depended expression of methanol metabolic enzymes

In our previous study, we showed that induction of AOD isozymes in P. methanolica was suppressed completely by respiratory chain inhibitors antimycin A or sodium azide (NaN3), with alterations in the AOD isozyme zymogram patterns in response to NaN3 or antimycin A concentration (Fujimura et al., 2007). Further, expression of methanol-inducible genes AOD1, DAS1, and PMP20 was shown to be completely repressed by addition of NaN3, concomitant with their proteins not being detectable by Western blot analysis (Fig. 2). On the other hand, expression of PMP47 was unaffected by addition of NaN3 (Fig. 2). However, as NaN3 may inhibit cellular functions outside of the respiratory chain, we constructed the S2ρ0 strain, which lacks a mitochondrial genome (ρ0), to correlate any effect of the respiratory chain on the oxygen responsiveness of methanol-inducible genes. The S2ρ0 strain was unable to grow on methanol or glycerol and exhibited slow growth on glucose. Mitochondria of this strain had lost their function as they were not stained by MitoTracker Green FM (Fig. 3). When cells were incubated in methanol medium, the expression levels of AOD1, DAS1, and PMP20 were greatly decreased in the S2ρ0 strain compared with that in the wild-type strain (Fig. 4), and AOD, DAS, and Pmp20 proteins were detectable only at low levels by Western blot analysis (Fig. 4). On the other hand, expression of PMP47 was unaffected in the S2ρ0 strain (Fig. 4), and the strain can grow slowly in the YPM medium.

Figure 2.

Effect of respiratory inhibition on induction of methanol-inducible enzymes by methanol. (a) Western analysis. α-AOD, anti-AOD antibody; α-DAS, anti-DAS antibody; α-Pmp20, anti-Pmp20 antibody; and α-Pmp47, anti-Pmp47 antibodies. (b) Relative activities of AP. Closed bars, cells induced by methanol under aerobic conditions (oxygen absorption rate, 16.0 mM min−1 L−1); open bars, cells induced by methanol under hypoxic conditions; and gray bar, cells grown on glucose. The specific activities of AP under aerobic conditions are set to 100%. Data are the averages of results from three independent experiments.

Figure 3.

Phenotypes of the Candida boidinii S2ρ0 strain. (a) Growth on several carbon sources, 1% (w/v) methanol; 1% (v/v) glycerol; and 1% (w/v) glucose. (b) The mitochondrial membrane potential of C. boidinii strains. The wild-type and S2ρ0 strains grown on the YPD medium were stained with 200 nM MitoTracker Green FM.

Figure 4.

Effect of ρ0 phenotype on induction of methanol-inducible enzymes by methanol. (a) Western analysis. α-AOD, anti-AOD antibody; α-DAS, anti-DAS antibody; α-Pmp20, anti-Pmp20 antibody; and α-Pmp47, anti-Pmp47 antibodies. (b) Northern analysis. The DNA probes containing the entire coding regions of AOD1, DAS1, PMP20, and PMP47 were each labeled. Lane 1, wild-type strain and lane 2, S2ρ0 strain.

These results suggest that the respiratory chain might participate in the oxygen response for the methanol-inducible peroxisomal genes, AOD1, DAS1, and PMP20, but not for expression of PMP47.

Oxygen conditions and mitochondrial respiratory function affect peroxisome morphology

The peroxisome is the organelle that hosts several oxidase reactions which use oxygen directly as a substrate. Induction of three peroxisomal matrix proteins, AOD, DAS, and Pmp20, was clearly affected by oxygen conditions as described above. From these observations, we posit that peroxisome biogenesis and peroxisome morphology might be affected by oxygen conditions. To confirm this hypothesis, we examined the peroxisomal morphology of wild-type cells grown on methanol using GFP-PTS1, which expresses GFP tagged with an AKL sequence at its carboxyl terminus (Sakai et al., 1998b). When the cells were grown in methanol under aerobic conditions, large clustered peroxisomes (2.47 ± 0.564 peroxisomes per cell) were observed (Fig. 5). In contrast, the cells induced with methanol under hypoxic conditions contained only a few small peroxisomes (1.43 ± 0.277 peroxisomes per cell) (Fig. 5). These results suggest that proliferation of peroxisomes in C. boidinii is affected by oxygen conditions.

Figure 5.

Effect of oxygen and respiratory inhibition on peroxisome morphology in the GFP-PTS1/wt and GFP-PTS1/aod1∆ strains. Candida boidinii strains expressing GFP-PTS1 were grown on YPM medium.

Moreover, as expression of AOD1, DAS1, and PMP20 genes was affected by inhibition of the respiratory chain, we asked whether or not this inhibition extended to the proliferation of peroxisomes. When strain GFP-PTS1/wt was exposed to NaN3 under aerobic conditions, cells contained clustered small peroxisomes (2.99 ± 0.243 peroxisomes per cell) (Fig. 5), suggesting that the respiratory function is not important to proliferation of peroxisomes. However, this morphology of peroxisomes was very similar to that of strain GFP-PTS1/aod1∆ under aerobic conditions without NaN3 (2.95 ± 0.511 peroxisomes per cell) (Fig. 5). These results suggest that while proliferation of peroxisomes was not affected by oxygen conditions, the size of the peroxisomes was affected by induction of peroxisomal matrix proteins, that is, AOD, whose induction depended on oxygen conditions.


For methanol metabolism, methylotrophic yeasts require large amounts of oxygen for AOD-catalyzed methanol oxidation in the peroxisomes and for oxidative phosphorylation of the respiratory chain in the mitochondria. Therefore, the cell has to control the oxygen consumption balance between the peroxisomes and mitochondria. In this study, we showed that induction of gene expression for methanol metabolism and proliferation of peroxisomes in the methylotrophic yeast C. boidinii were affected by oxygen conditions. Induction of three methanol metabolic enzymes, AOD, DAS, and Pmp20p, was shown to be oxygen dependent. In a previous report, we showed that the induction patterns of AOD isozymes in P. methanolica were controlled by oxygen concentration (Fujimura et al., 2007). These findings indicate that methylotrophic yeast species have the ability to sense oxygen conditions and to control methanol metabolism accordingly. Moreover, induction of methanol metabolism in C. boidinii is affected by respiratory activity, and the respiratory activity may be related to oxygen sensing for the expression of methanol-inducible genes, as it is in P. methanolica. It seems that synchronous regulation between methanol metabolism and respiratory activity is a reasonable mechanism to control the oxygen consumption balance between peroxisomes and mitochondria in the cell, because the respiratory chain acts downstream of methanol metabolism and has a lower Km value for oxygen than AOD.

On the other hand, induction of Pmp47 by methanol was not affected by oxygen conditions and respiratory activity. Pmp47, which encodes the peroxisomal membrane protein, functions as an ATP transporter in the peroxisome and is also involved in peroxisome proliferation (Nakagawa et al., 2000). It is known that the cell needs to always have a core peroxisome present, even under noninducing conditions, to assure an immediate response to environmental changes. Therefore, Pmp47 must be induced even under hypoxic conditions, and its induction mechanism is different from that of the methanol metabolic enzymes. Methanol-inducible promoters from C. boidinii have several consensus sequences (Yurimoto et al., 2000; Sasano et al., 2008). However, these elements may not participate in the oxygen response because PMP47 promoter, which is expressed even under hypoxic conditions, also has these consensus sequences. Our next study is to identify the transcription factors and the consensus sequences in the methanol-inducible promoters for oxygen response.

The morphology of peroxisomes was also affected by oxygen conditions. The size and number of peroxisomes increased during growth on methanol in C. boidinii (Sakai et al., 1998b). Under hypoxic conditions, however, methanol-induced proliferation of peroxisomes in size or number was not observed. These results suggest that peroxisome morphology is related to oxygen conditions. On the other hand, wild-type cells exposed to NaN3 under aerobic conditions contained very small peroxisomes, and its morphology was very similar to that of the aod1∆ strain. It is very possible that peroxisome proliferation is not affected directly by oxygen conditions, as the machinery for peroxisome biogenesis and protein import has baseline levels of function even under hypoxic conditions. Therefore, we propose that peroxisome morphology is dependent upon induction of peroxisomal matrix proteins, that is, AOD and DAS.

In this study, we provided evidence that methanol metabolism and peroxisome morphology in C. boidinii are coordinated by induction of methanol-metabolizing enzymes whose induction level was affected by oxygen conditions, and that a key factor for its induction is the mitochondrial respiratory chain. On the other hand, there were some reports that peroxisomes can also function both as an intracellular signaling compartment and as an organizing platform for orchestrating developmental decisions from inside the cell (Titorenko & Rachubinski, 2004; Dixit et al., 2010). Ivashchenko et al. (2011) reported that the mitochondrial redox balance was perturbed in catalase-deficient cells and upon generation of excess ROS inside peroxisomes, so that peroxisomes and mitochondria regulate intracellular redox balance in a coordinated manner. Moreover, it was reported that RTG genes that function in communication between mitochondria and the nucleus were required for the expression of genes encoding peroxisomal proteins in Saccharomyces cerevisiae (Chelstowska & Butow, 1995; Epstein et al., 2001). Taking all these reports in account with our data strongly suggests that both these organelles engage in exchanging information with each other. As demonstrated in this study, we believe that respiratory activity functions as a sensor for oxygen, and that there are some signaling pathways to the nucleus for oxygen sensing, which are recognized by the respiratory chain. However, the signaling factors and mechanisms for an oxygen-sensing system that can distinguish subtle differences in oxygen concentration have yet to be elucidated. As such, our future work will concentrate on the oxygen-sensing system and identifying the factors functioning between mitochondria and peroxisomes.


This research was supported in part by a Grant-in-Aid for Scientific Research (C), No. 23580109 to T.N., from the Ministry of Education, Culture, Sports, Science and Technology, Japan.