Pichia methanolica is a methylotrophic yeast, able to use methanol as a sole carbon and energy source. This organism has been widely used in both basic and applied studies, i.e. the methanol-metabolic pathway (Nakagawa et al., 2002, 2004), peroxisome biogenesis and degradation (Kulachkovsky et al., 1997; Lusta et al., 2000; Nakagawa et al., 2005), yeast prions (Chernoff et al., 2000; Kushnirov et al., 2000), fatty acid production (Aoki et al., 2002) and biosensors (Lobanov et al., 2001; Reshetilov et al., 2001). One of the primary uses of P. methanolica has been as a host for heterologous gene expression (Raymond et al., 1998), similar to other methylotrophic yeasts, e.g. P. pastoris, Hansenula polymorpha and Candida boidinii (Sakai et al., 1999; Gellissen, 2000; Gerngross, 2004).
Heterologous gene expression systems in methylotrophic yeasts generally use promoters derived from alcohol oxidase (AOD) genes, because AOD promoters are tightly regulated by carbon source, being strongly induced by methanol and completely repressed by glucose. AOD-based gene expression systems in methylotrophic yeasts are among the major tools that have been used for overproduction of various proteins. For example, the P. methanolica expression system has been used to produce a number of recombinant proteins, including human glutamate decarboxylase, transferrin, lignin peroxidase, laccase and hepatitis B virus polymerases (Choi et al., 2002; Mayson et al., 2003; Wang et al., 2004; Guo et al., 2005).
P. methanolica possesses two genes encoding AOD, MOD1 (AUG1) and MOD2 (AUG2) (Raymond et al., 1998; Nakagawa et al., 1999). In P. pastoris, AOD is encoded by AOX1 and AOX2 (Koutz et al., 1989) and in Pichia sp. strain 159, by AOXA and AOXB (Szamecz et al., 2005). In P. pastoris, AOX2 was reported to be expressed at a significantly lower level than AOX1, although AOX1 expression was strongly induced by methanol (Cregg et al., 1989). On the other hand, in P. methanolica, expression of the second AOD gene, MOD2, was strongly induced by methanol, as was MOD1 (Nakagawa et al., 2002). In other words, P. methanolica produces large amounts of two AOD subunits during growth on methanol, which is different from other methylotrophic yeasts. Nine AOD isozymes are formed from these two gene products, Mod1p and Mod2p (Nakagawa et al., 1996, 1999, 2001). We speculated that MOD1 and MOD2 expression was controlled through distinct regulatory mechanisms as a function of carbon and nitrogen source (Nakagawa et al., 1996) and methanol concentration (Nakagawa et al., 1999), and reasoned that it should be possible to develop a unique and useful gene expression system to exploit the regulatory differences between the two promoters. To our knowledge, no studies have reported attempts to construct a gene expression system utilizing PMOD2, or have compared MOD1 and MOD2 gene expression at the promoter level.
In this report we describe development of a new P. methanolicaPMOD1- and PMOD2-based promoter assay system, utilizing the S. cerevisiae acid-phosphatase (AP) gene (ScPHO5) as a reporter, and compare regulation of these two methanol-inducible P. methanolica AOD promoters.
Materials and methods
Bacterial strains, media and cultivation
P. methanolica IAM 12 901 was used as the wild-type strain and was the source of chromosomal DNA. P. methanolica strain PMAD11 (ade2) (Invitrogen Co., Carlsbad, CA, USA) was used as a host for transformation. Escherichia coli DH5α was used for plasmid propagation.
Complex yeast peptone–dextrose medium (YPD) and synthetic MI medium (Sakai et al., 1998) were used for growing the P. methanolica strains. The following were used as carbon sources: 1% w/v glucose, 1% v/v glycerol, 1% v/v methanol, or 1% v/v glycerol plus 1% v/v methanol. The initial pH of the media was adjusted to 6.2. Yeasts were grown under aerobic conditions at 28 °C with reciprocal shaking, and growth was monitored by measuring optical density at 660 nm.
E. coli was grown at 37 °C in 2 × YT medium, supplemented with ampicillin (50 µg/ml) as needed.
Yeast DNA was isolated by the method of Cryer et al. (1975) or Davis et al. (1980). Transformation of P. methanolica PMAD11 (ade2) was performed as described (Raymond et al., 1998). DNA fragments were sequenced by the dideoxy cycle sequencing method with a PRISM DyeDeoxy Terminator Cycle Sequencing kit and DNA Sequencer Model 377 (Applied BioSystems, Foster, CA, USA).
Southern analysis was performed as follows. After digestion with several restriction enzymes, 5 µg DNA samples were electrophoresed in a 0.7% agarose gel containing TAE buffer and transferred to a nylon membrane (Hybond-N+; Amersham Biosciences Corp., Piscataway, NJ) in 20 × SSC (1 × SSC = 0.15 M NaCl plus 0.015 M sodium citrate). The entire coding region of ScPHO5 was labelled as a probe with an AlkPhos DIRECT kit (Amersham Biosciences Corp.).
Construction of the ScPHO5-expression cassette and its integration into P. methanolica chromosomal DNA
The ScPHO5 gene was amplified using the 5′-Xho I-PHO5 primer, 5′-CgAgCTCgAgAAAATgTTTAAATCTgTTgTTTATTCAA-3′, and the PHO5-Not I-3′ primer, 5′-ATAgTTTAgCggCCgCTTATTgTTTTAATAgggTATCATTg-3′, using pSH39 (a kind gift from Professor S. Harashima; Mukai et al., 1993) as template. The approximately 1.5 kb amplification product was subcloned into a pT7blue T-vector (Novagen Inc., Madison, WI, USA). The resultant vector was digested with XhoI and NotI, and the resulting 1.5 kb fragment was ligated into the XhoI–NotI sites of pMET B (Invitrogen Co., Carlsbad, CA, USA) to produce pPMOD1–ScPHO5.
PMOD2 was amplified by PCR using the PMOD2-5′ primer, 5′-CgAgCTCgATCCACTACAgTTTACCAATT-3′, and the PMOD2-3′ primer, 5′-CCgCTCgAgAATTTTAgTTTTAgATAgATA-3′, and P. methanolica IAM12901 genomic DNA as template. The approximately 1.6 kb amplified fragment was subcloned into pT7Blue to produce pMOD2p. pPMOD2 was digested with PstI and XhoI, and the PMOD2 fragment was ligated into the PstI–XhoI sites of PstI–XhoI-digested pPMOD1–ScPHO5. The nucleotide sequence of PMOD2 (1685 bp) has been assigned Accession No. AB223040 in the DDBJ/EMBL/GenBank nucleotide sequence database.
PstI-linearized pPMOD1–ScPHO5 and pPM-OD2–ScPHO5 were each introduced into P. methanolica strain PMAD11 (ade2) by non-homologous recombination (Hiep et al., 1993; Raymond et al., 1998), and Ade transformants were isolated. ScPHO5 copy number in the transformants was determined by band intensity on a Southern blot, using EcoRI-digested chromosomal DNA. Among the transformants, strains that produced the least intense ScPHO5 signal by genomic Southern analysis (data not shown) were named MOD1-P and MOD2-P, respectively. The two strains exhibited the same level of AP activity on 0.5% methanol medium. These results were in agreement with the zymogram and expression patterns of Mod1p and Mod2p, which showed the same levels of induction on 0.5% methanol medium (Figures 2, 3). Therefore, we conclude that the MOD1-P and MOD2-P strains had the same copy number of ScPHO5-expression cassettes.
Histochemical staining and AP assay
P. methanolica transformants were plated on YNB plates in a medium containing one of the following carbon sources: 1% v/v methanol, 1% v/v glycerol or 1% w/v glucose. After 16 h incubation at 28 °C, each plate was stained by the diazo-coupling method of Dorn (1965). AP activity was assayed as described (Torriani, 1960), and the specific AP activity of a cell suspension (units/unit OD660nm) was then determined (Toh-e et al., 1973).
Preparation of cell-free extracts, AOD analysis and electrophoresis
P. methanolica cells were harvested and washed in distilled water by centrifugation (1000 × g for 5 min at room temperature) and resuspended in 50 mM sodium phosphate buffer, pH 7.5, in a 2 ml microfuge tube containing an equal volume of 0.5 mm diameter zirconium beads (Biospec Products, Inc., Bartlesville, OK, USA). The tube was shaken vigorously for 30 s in a 3110BX mini-beadbeater (Biospec Products, Inc.) and then chilled on ice for 30 s. This procedure was repeated six times, after which the pellet was removed by centrifugation at 16 000 × g for 10 min at 4 °C. The cell-free supernatant was immediately subjected to enzyme assay and isozyme analysis by electrophoresis.
The protein concentration of cell-free extracts was determined by the Bradford (1976) method, using BSA as a standard (Bio-Rad Laboratories Protein Assay Kit, Hercules, CA).
AOD activity was assayed by the ABTS/POD method (Tani et al., 1985) at 28 °C in 2 ml 50 mM sodium phosphate buffer, pH 7.0.
For zymogram analysis, cell-free extracts corresponding to approximately 20 µg protein were subjected to native PAGE using 5% polyacrylamide (Laemmli, 1970). After electrophoresis, the gels were stained for AOD (Lee and Komagata, 1983). Guaiacol (Wako Pure Chemical Industries, Osaka, Japan) was used as the peroxidase substrate (Nakagawa et al., 1996).
Construction of a promoter assay system utilizing ScPHO5 in P. methanolica
In order to assess PMOD1 and PMOD2 expression levels in P. methanolica, a promoter assay system was constructed, using the S. cerevisiae PHO5 (ScPHO5) gene for acid phosphatase (AP) as a reporter because ScPHO5 encodes periplasmic AP in S. cerevisiae and has previously been used as a reporter for methanol-inducible promoters in the methylotrophic yeast C. boidinii (Yurimoto et al., 2000).
Histochemical staining of AP activity was performed on wild-type P. methanolica and on the MOD1-P strain containing the PMOD1-ScPHO5 cassette. For the wild-type, AP activity was not detected, regardless of whether methanol, glycerol or glucose was used as the carbon source (Figure 1). On the other hand, colonies of strain MOD1-P grown on methanol or glycerol medium stained dark red, indicating high AP activity, while colonies grown on glucose did not (Figure 1). These results are consistent with the known carbon source regulation of MOD1 (Nakagawa et al., 1996, 2001).
Strain MOD2-P containing the PMOD2-ScPHO5 cassette was also assayed. Colonies of strain MOD2-P grown on methanol stained dark red, like strain MOD1-P, but colonies grown on glucose or glycerol did not (Figure 1). These results are also consistent with the known carbon source regulation of MOD2 (Nakagawa et al., 1996, 2001). Expression levels of PMOD1 and PMOD2 were then assessed with the constructed reporter assay system (Figure 2). PMOD1 and PMOD2 were found to be strongly induced by methanol, exhibiting approximately the same level of AP activity, indicating that PMOD2, like PMOD1, has the ability to induce high levels of heterologous gene expression.
Regulation of PMOD1 and PMOD2 by carbon source
We have previously used zymogram analysis to examine carbon source regulation of MOD1 and MOD2 expression because the band profile reflects the ratio of subunit species (Nakagawa et al., 1999, 2002). To clarify transcriptional regulation, ScPHO5 expression levels under the control of the PMOD1 and PMOD2 promoters as a function of carbon source were assessed by AP activity (Figure 2).
PMOD1 and PMOD2 were found to be strongly induced by methanol but not by glucose. PMOD1 was also found to be induced by glycerol but PMOD2 was not. Because the regulatory patterns obtained with the established promoter assay system are in agreement with the zymogram patterns of the AOD isozymes (Figure 2), regulation of PMOD1 and PMOD2 appears to be controlled primarily at the transcriptional level.
The level of AP activity under PMOD1 with glycerol was approximately 70% of that with methanol, which was almost the same level reported for PMOX in H. polymorpha (Gödecke et al., 1994). It has been reported that expression of P. pastorisPAOX1 is completely repressed by glycerol (Inan and Meagher, 2001) and that in C. boidinii it is slightly induced (ca. 3% of that with methanol; Yurimoto et al., 2000). With respect to carbon source regulation, it appears that the regulation systems for PMOD1 and PMOD2 are similar to those for PMOX from H. polymorpha and PAOX1 from P. pastoris, respectively. However, glycerol did not repress methanol-induced expression of PMOD1 or PMOD2 (Figure 2), although glucose completely repressed expression of both (data not shown). Therefore, we speculated that PMOD2 is regulated by a novel system, different from the regulation reported for other AOD-gene promoters.
Next, AP expression induced by PMOD1 and PMOD2 was assayed over a time course, initially in cells grown on glycerol, but also following an addition of methanol to the medium 10 h post-inoculation. On glycerol as sole carbon source (0–10 h), PMOD1 was strongly expressed, but PMOD2 was not expressed (Figure 3A). In contrast, following the methanol addition, PMOD2-induced ScPHO5 expression was found to be very strong when assayed after 20 h. These expression patterns are in excellent agreement with the zymogram analysis of the AOD isozymes. These results indicate that PMOD2 expression can be regulated simply by the timing of methanol addition to glycerol-containing medium, showing that the production phases of two heterologous proteins can be differentially controlled, using PMOD1 and PMOD2 in the same host cell and in the same flask.
Effect of methanol and oxygen on PMOD1 and PMOD2 expression
We have previously shown that the composition of AOD-isozyme species responds to the methanol concentration in the medium, i.e. the ratio of Mod2p to Mod1p increased with an increase in methanol concentration (Nakagawa et al., 2002). To determine whether expression of PMOD1 and PMOD2 is indeed regulated by methanol concentration, PMOD1 and PMOD2 expression was assayed as a function of methanol level.
As shown in Figure 4, PMOD1 expression was observed even in the absence of carbon source (derepressed conditions). At 0.1% methanol, the expression level of PMOD1 was higher than when cells were derepressed. With higher methanol concentrations (up to 1%), PMOD1 expression was nearly constant (Figure 4A). On the other hand, PMOD2 expression was not observed under derepressed conditions (0% methanol). With an increase in the methanol concentration (up to 0.3%), PMOD2 expression drastically increased, with expression becoming approximately constant at higher methanol concentrations (up to 1%). Thus, PMOD1 expression predominated at low methanol concentrations, while PMOD2 expression predominated at high methanol concentrations. These findings are consistent with the band profile of an AOD zymogram and the composition ratio of subunit species (Nakagawa et al., 1999), suggesting that the expression of PMOD1 and PMOD2 by methanol concentration and carbon source are controlled differentially.
The expression levels of PMOD1 and PMOD2 were also found to be influenced by shaking speed during growth. At a low shaking speed (50 r.p.m.), the expression of PMOD1 and PMOD2 in 0.5% methanol was the same (Figure 4A). However, at a high shaking speed, the expression level of PMOD2 in 0.5% methanol was about two-fold higher than at a low shaking speed, while that of PMOD1 did not change (Figure 4B). On the other hand, PMOD1 expression in 0.1% methanol rose with an increase in shaking speed, while expression of PMOD2 did not (Figure 4B). These results indicate that the expression of PMOD1 and PMOD2 is responsive to oxygen, and that PMOD1 and PMOD2 are suitable for production of heterologous proteins under low methanol conditions and with high methanol and high oxygen concentrations, respectively.
Two AOD promoters, PMOD1 and PMOD2, were evaluated for heterologous gene expression in P. methanolica. Both promoters were strongly expressed after methanol induction. Because the level of AP activity driven by PMOD2 was as high as that driven by a commonly used promoter in P. methanolica, PMOD1, PMOD2 has the potential to be used as a promoter for recombinant protein production in P. methanolica.
Expression of PMOD1 and PMOD2 was found to be controlled primarily at the transcriptional level, and this pattern of regulation was found to be in good agreement with previous analyses (Nakagawa et al., 1996, 2002). The expression pattern of PMOD1 was found to be similar to that of PMOX from H. polymorpha, because PMOD1 is responsive not only to methanol, but also to glycerol. For the production of foreign proteins in yeasts using AOD promoters, the fermentation process is usually divided into two phases, a growth phase (during which cells are grown on glucose or glycerol to high cell density), and the induction phase (during which gene expression is induced by transferring the cells to methanol). However, in the case of PMOD1-controlled gene expression, the desired protein could be produced not only during the induction phase, but also during the growth on glycerol, with the possibility of higher yields than in the usual two phase production system. In contrast, PMOD2 is induced only by methanol, although it is not repressed upon glycerol addition to the methanol medium. This regulation of PMOD2 is different than that previously reported for other AOD promoters.
These differences in the regulation of PMOD1 and PMOD2 could be applied to the expression of two heterologous proteins in the same host cells. For example, when foreign proteins, such as collagen (Vuorela et al., 1997), require a post-translational modification by an enzyme that is not present in yeast or are hetero-oligomeric, these proteins must be co-synthesized in the same host cells. Moreover, when production of toxic and non-toxic heterologous proteins are desired in the same host cells, the stage of gene expression for the toxic protein must be limited to the non-growth phase. The two AOD promoters, PMOD1 and PMOD2, could be used in a situation in which their expression could be differentially controlled by adjustment of methanol and oxygen supply. One problem limiting the potential use of PMOD1 and PMOD2 in a gene expression system for P. methanolica is the lack of selectable markers and appropriate auxotrophic host strains, as only an ADE2-based selectable marker system has been constructed (Raymond et al., 1998). We are currently developing a new host/marker system in P. methanolica.
We are grateful to Professor Satoshi Harashima and Dr Yoshinobu Kaneko, Department of Biotechnology, Graduate School of Engineering, Osaka University, for providing the ScPHO5 gene. We are also grateful to Mr Yoshiyuki Nakayama and Mr Tasuku Mizumura, Department of Food Science and Technology, Tokyo University of Agriculture, for their skillful assistance. This research was supported in part by a Grant-in-Aid for Young Scientists (B), No. 17780065, from the Ministry of Education, Culture, Sports, Science and Technology, Japan, to T.N.