The methanol sensor Wsc1 and MAPK Mpk1 suppress degradation of methanol-induced peroxisomes in methylotrophic yeast

Autophagy refers to the degradation pathway of cellular components after transport into the vacuole and/or lysosome, and plays versatile physiological roles in unicellular and multicellular eukaryotes (Mizushima and Komatsu, 2011). Extensive studies have shown that selective autophagy of organelles employs ‘receptor’ proteins on the surface of the autophagy target organelles that recruit the Atg (autophagy-related) protein machinery for formation of autophagosomes (Xie and Klionsky, 2007). In the methylotrophic yeast Komagataella phaffii (Kp; also known as Pichia pastoris), which serves as a model organism for studying pexophagy (Pang et al., 2019; Sakai et al., 1998), KpAtg30 exerts its role by interacting with the peroxins KpPex3, KpPex14 and KpAtg37, which are necessary for peroxisome biogenesis, and the Atg proteins KpAtg8 and KpAtg11 (Burnett et al., 2015; Farré et al., 2013, 2008; Nazarko et al., 2014). The interaction between KpAtg30 and KpAtg11 requires phosphorylation of KpAtg30 by the kinase KpHrr25, which serves as a cue for pexophagy induction (Zientara-Rytter et al., 2018). In the yeast Saccharomyces cerevisiae (Sc), ScHrr25 and the pexophagy receptor protein ScAtg36 function in a similar way (Tanaka et al., 2014).

In contrast to the molecular mechanism and machinery of autophagy, the physiological roles of pexophagy in yeasts and fungi have been less studied, with the exception of plant pathogenicity (Asakura et al., 2009; Oku and Sakai, 2016; Oku et al., 2014). We have previously reported that by adapting to the methanol concentration in the environment, the methylotrophic yeasts Candida boidinii and K. phaffii proliferate slowly on plant leaves, undergoing 3–4 cell divisions in 7–10 days (Kawaguchi et al., 2011). The methanol concentration on plant leaves exhibits diurnal fluctuations, with high concentrations in the dark period and low concentrations in the light period. When the methanol concentration is high in the dark period, peroxisomes containing major methanol-utilizing enzymes are highly induced. In contrast, when the methanol concentration becomes low in the light period, peroxisomes are degraded by pexophagy (Kawaguchi et al., 2011). The induction of peroxisomes and methanol-utilizing enzymes, as well as Atg30-dependent pexophagy, are necessary for methylotrophic yeast cells to proliferate on plant leaves, indicating the essential role of peroxisome homeostasis and its regulation in cellular survival on plant leaves (Kawaguchi et al., 2011).

We have also previously reported that the cell-surface proteins KpWsc1 and KpWsc3, homologs of ScWsc1 in S. cerevisiae, are responsible for sensing the concentration of environmental methanol and regulating methanol-inducible gene expression (i.e. genes encoding proteins involved in peroxisome synthesis and methanol-metabolizing enzymes) in K. phaffii (Ohsawa et al., 2017). ScWsc1 is a glycosylphosphatidylinositol (GPI)-anchored plasma membrane protein that senses the perturbation of cell wall integrity (CWI) (Dupres et al., 2009; Lodder et al., 1999) and initiates a phosphorylation cascade (MAPK cascade) leading to the activation of a MAPK (ScMpk1, also known as Slt2) to activate expression of cell wall-related genes, such as those related to glucan synthesis (Levin, 2011; Rodicio and Heinisch, 2010). Our previous study with K. phaffii showed that knockout of KpWSC1 leads to unbalanced methanol metabolism, resulting in formaldehyde accumulation and retardation of cell growth (Ohsawa et al., 2017). KpWsc1 and its downstream factor KpRom2 are involved not only in methanol-induced gene expression, but also in the high-temperature stress response in K. phaffii (Ohsawa et al., 2017), which is similar to the CWI pathway in S. cerevisiae. Signaling of the two pathways could be dissected using mutations in KpWsc1, and the two pathways were thus shown to be independent of each other (Ohsawa et al., 2017).

In the present study, we examined how the environmental methanol concentration affects and regulates pexophagy, that is, degradation of methanol-induced peroxisomes. We found that the presence of methanol, which was sensed by KpWsc1, activated the downstream MAPK and suppressed pexophagy. Cellular sensing of nutrient concentrations has been shown to play a critical role in maintaining organellar homeostasis. The physiological significance of a regulatory mechanism of selective autophagy in cellular survival in nature is discussed.

In order to compare pexophagy between two different strains of K. phaffii in response to the methanol concentration, we used a Beppu flask in which the two compartments of the flask are separated by a membrane (pore size 0.22 µm) that allows medium components, including methanol, to pass through and be maintained at the same level, but does not allow yeast cells to mix (Fig. 1A).

Pexophagy was compared between the wild-type strain and the pexophagy-deficient strain Kpatg30Δ. The wild-type cells consumed methanol, proliferated and reached stationary phase after ∼20 h of cultivation, with complete consumption of methanol (Fig. S1A). Pexophagy was monitored by observing cleavage of a YFP-fused form of the peroxisomal membrane protein KpPex11 (Pex11–YFP), which yields a YFP fragment resistant to vacuolar proteases. The cleaved YFP band appeared after ∼16 h (0.15% methanol) to 18 h (0.04% methanol), and after depletion of methanol at 20 h, the band of cleaved YFP became more intense (Fig. S1B).

The pexophagy process of the Kpatg30Δ strain inoculated in the other compartment of the Beppu flask was compared with that of the wild-type strain. KpAtg30 is a pexophagy factor acting as a receptor on the peroxisome membrane (Farré et al., 2008). Comparison of the wild-type and Kpatg30Δ strains in the Beppu flask revealed that the methanol concentrations in the two compartments of the flask showed the same kinetics of methanol consumption and that both strains showed similar growth curves (Fig. S1A). As expected, cleavage of Pex11–YFP was inhibited in the Kpatg30Δ strain (Fig. S1B). After 20 h, the wild-type strain expressing a peroxisome-targeted CFP (CFP–SKL) exhibited a diffuse CFP signal pattern in the vacuolar lumen (Fig. S1C, Table S4). In contrast, the percentage of cells with a diffuse pattern of CFP fluorescence in the vacuole was remarkably reduced in the Kpatg30Δ strain. These results collectively indicate that Atg30-dependent pexophagy is induced at the late exponential to early stationary phase (16–18 h) in batch cultures of K. phaffii when the methanol concentration falls below 0.15%.

Next, we compared pexophagy between the wild-type and Kpwsc1Δ strains. In a previous study, we have shown that the Kpwsc1Δ strain is impaired in sensing the environmental methanol concentration (Ohsawa et al., 2017). Although the Kpwsc1Δ strain showed a decreased growth yield in a Beppu flask, the two strains showed similar growth curves (Fig. 1A). Accumulation of formaldehyde in the Kpwsc1Δ strain, which was observed when this strain was cultivated alone, was avoided due to consumption of formaldehyde by the wild-type strain.

The Kpwsc1Δ strain exhibited an earlier onset of pexophagy than the wild-type strain. In the Kpwsc1Δ strain, higher ratios of the cleaved YFP band to the full length Pex11–YFP were observed as early as 12 h of cultivation (Fig. 1B), when more than 0.25% methanol was still present in the medium. As judged from the difference in the timing of pexophagy induction between the wild-type and Kpwsc1Δ strains (Fig. 1B), KpWsc1 is considered to suppress the degradation of methanol-induced peroxisomes in the presence of more than 0.15% methanol.

Next, we compared macropexophagy between the wild-type and Kpwsc1Δ cells by shifting methanol-induced cells to ethanol-containing medium. Macropexophagy was monitored by observing cleavage of YFP-fused KpAtg8 (Atg8–YFP). As shown in Fig. S2, we did not observe enhancement of pexophagy in the Kpwsc1Δ mutant. These results suggest that KpWsc1 is not involved in regulation of ethanol-induced macropexophagy.

To elucidate the molecular mechanism of pexophagy suppression by methanol and its sensor KpWsc1, we next analyzed phosphorylation of KpAtg30, which induces pexophagosome formation at the peroxisomal surface (Farré et al., 2008). In the following experiments, we used Kpatg11Δ as the background strain to observe the phosphorylation level of HA-tagged KpAtg30 (KpAtg30–3×HA). In the Kpatg11Δ strain, degradation of KpAtg30–3×HA through pexophagy was blocked, and phosphorylated bands of KpAtg30–3×HA were readily visible. In the periodically sampled lysate of the Kpatg11Δ strain, the phosphorylation level of KpAtg30–3×HA was found to increase gradually with cultivation time, giving multiple bands of less mobility than the non-phosphorylated form (Fig. 2A). Next, we examined whether KpWsc1 depletion affects the phosphorylation level of KpAtg30–3×HA during growth on methanol. KpAtg30–3×HA in the Kpwsc1Δatg11Δ strain was found to be more highly phosphorylated in samples from a 12 h culture and thereafter than that in the Kpatg11Δ strain (Fig. 2A), consistent with the observed enhancement of pexophagy by depletion of KpWsc1 (Fig. 1B).

In our previous study, two variants of KpWsc1 mutated at the tyrosine 53 residue were found to have distinct phenotypes: the KpWsc1 Y53A mutation (but not the Y53F mutation) results in impaired methanol-inducible gene expression, and the KpWsc1 Y53F mutation (but not the Y53A mutation) renders the cells more susceptible to high-temperature stress (Ohsawa et al., 2017). The phosphorylation levels of KpAtg30–3×HA in strains carrying these KpWsc1 variants were analyzed, and the wild-type KpWsc1 and its Y53A variant reverted the hyper-phosphorylation phenotype of KpAtg30–3×HA in the Kpwsc1Δatg11Δ strain. In contrast, the Y53F variant could not (Fig. 2B). These results indicate that signaling mediated by KpWsc1, which is similar to the high-temperature stress response, was required for suppression of KpAtg30 phosphorylation. In line with this result, another KpWsc1 mutant (310–316Δ) incapable of interacting with KpRom2 (a signaling factor downstream of Wsc1) failed to revert the hyper-phosphorylation phenotype in the Kpwsc1Δ atg11Δ strain (Fig. 2C).

Among the identified K. phaffii Wsc proteins, KpWsc3 has also been found to be involved in methanol-inducible gene expression, but not in the high-temperature stress response (Ohsawa et al., 2017). Overexpression of KpWsc3–FLAG failed to revert the hyper-phosphorylation phenotype of Kpwsc1Δatg11Δ, even though the protein expression level was higher than that of KpWsc1–FLAG (Fig. S3A). This reinforces the notion that KpWsc1-mediated pexophagy regulation that is distinct from the methanol-inducible gene expression pathway is responsible for suppression of KpAtg30 phosphorylation in K. phaffii.

In S. cerevisiae, the CWI pathway leads to phosphorylation and activation of the MAPK ScMpk1 by MAPK kinases (Irie et al., 1993; Lee et al., 1993). The Kpmpk1Δ strain showed normal growth on methanol, together with peroxisome proliferation. To assess the involvement of KpMpk1 in K. phaffii, the pexophagy process and phosphorylation of KpAtg30–3×HA was compared between Kpmpk1Δ and wild-type strains. We found an earlier onset of pexophagy in the Kpmpk1Δ strain cultured with the wild-type strain in a Beppu flask, similar to that observed in the Kpwsc1Δ strain (Fig. S3B), and depletion of KpMpk1 from the Kpatg11Δ strain further enhanced phosphorylation of KpAtg30-3xHA (Fig. 2D, lanes Kpatg11Δ and EV in Kpmpk1Δatg11Δ). While the wild-type Mpk1 tagged with YFP was able to revert the hyper-phosphorylation phenotype of the Kpmpk1Δatg11Δ strain, a KpMpk1 variant mutated in the conserved phosphorylation residues (TAYF mutant: T188 and Y190 residues replaced with alanine and phenylalanine, respectively) did not (Fig. 2D). These results strongly suggest that phosphorylated Mpk1 suppresses phosphorylation of KpAtg30 in the presence of methanol. Taken together with the data for the Kpwsc1 mutants, we conclude that the MAPK pathway in K. phaffii under the control of KpWsc1, KpRom2 and KpMpk1, was responsible for suppression of KpAtg30 phosphorylation.

In S. cerevisiae, ScMpk1 relays the CWI signaling cascade by phosphorylation of the transcriptional factors ScRlm1, ScSwi4 and ScSwi6, which are required for gene expression involved in cell wall synthesis (Levin, 2011; Watanabe et al., 1995). We tested whether these downstream transcription factors in K. phaffii were involved in phosphoregulation of KpAtg30. In the K. phaffii genome, we identified two putative genes encoding homologs of the above transcription factors, and named them KpRLM1 and KpSWI4, based on sequence similarity. We disrupted these genes in the K. phaffii atg11Δ strain, yielding the Kprlm1Δatg11Δ and Kpswi4Δatg11Δ strains. KpAtg30–3×HA was highly phosphorylated in Kprlm1Δatg11Δ cells, whereas the phosphorylation level of KpAtg30–3×HA in the Kpswi4Δatg11Δ strain was comparable to that in the Kpatg11Δ strain (Fig. 3A). We also found that cleavage of Pex11–YFP was enhanced in the Kprlm1Δ strain at 18 and 22 h, when methanol still remained in the medium (Fig. S4). These results suggest that KpMpk1 suppresses phosphorylation of KpAtg30 by transcriptional control under KpRlm1, and not KpSwi4.

After screening for phosphatases regulated by KpRlm1, we found that the transcript levels of KpMSG5 (XM_002492799.1) and KpPTP2A (XP_002491091.1; a paralog of ScPTP2) were reduced in the Kprlm1Δ strains (Fig. 3B), whereas the expression level of KpPTP2B (the other paralog of ScPTP2) was independent of KpRlm1 (data not shown). Next, we investigated whether these phosphatases were involved in the suppression of Atg30 phosphorylation (Fig. 4A,B). Depletion of the KpPtp2A phosphatase in the Kpatg11Δ strain enhanced phosphorylation of KpAtg30–3×HA after 17 h growth on methanol, when methanol was still present in the medium, and the Kpmsg5Δ strain also showed slight enhancement of Atg30 phosphorylation. The Kpptp2AΔmsg5Δatg11Δ strain exhibited higher phosphorylation levels than observed in other tested strains (Fig. 4C), although methanol still remained at higher concentrations with this triple disruption strain than other strains. These results indicate that the KpPtp2A and KpMsg5 phosphatases act downstream of KpRlm1 and have a role in the suppression of KpAtg30 phosphorylation in K. phaffii. Based on sequence similarity, both KpPtp2A and KpMsg5 are assumed to be tyrosine phosphatases, suggesting that these phosphatases do not act at serine 71 and serine 112 residues of KpAtg30, which are known to be important for binding of Atg8 and Atg11, respectively (Farré et al., 2013).

We have previously reported that KpWsc1 senses the environmental methanol concentration and transmits the signal to ensure an appropriate level of methanol-induced gene expression (Ohsawa et al., 2017). KpWsc1 gives maximum transcription output for peroxisome synthesis at methanol concentrations in the range of 0.05–0.1% (Ohsawa et al., 2017). In the phyllosphere, methylotrophic yeasts have been observed to respond to diurnal fluctuations of methanol (∼0–0.3%), alternating peroxisome synthesis and degradation in response to these changes (Kawaguchi et al., 2011). In this study, the methanol-sensing molecule KpWsc1 was found to regulate not only the level of peroxisome synthesis, but also pexophagy according to the methanol concentration by two distinct signaling pathways.

KpWsc1 has previously been found to initiate dual signaling pathways: one leading to expression of methanol-inducible genes and peroxisome biogenesis, and the other dictating the high-temperature stress response, similar to the S. cerevisiae CWI pathway (Ohsawa et al., 2017). These two pathways have been clearly distinguished by mutations at the tyrosine 53 residue of KpWsc1. KpWsc1 Y53A inhibits methanol-inducible gene expression but does not inhibit the high-temperature stress response, and KpWsc1 Y53F has the opposite effect. Our present results demonstrate that suppression of pexophagy and KpAtg30 phosphorylation were abolished by the KpWsc1 Y53F mutation but not by the KpWsc1 Y53A mutation. Therefore, KpWsc1 determines synthesis or degradation of peroxisomes depending on the environmental methanol concentration (Fig. 5). Molecules downstream of KpWsc1, namely KpRom2, KpMpk1 and the transcription factor KpRlm1, were also shown to regulate the phosphorylation level of KpAtg30 (Fig. 2B–D, Figs 3,5). By analogy with the S. cerevisiae CWI pathway, we speculate that K. phaffii homologs of Rho1, Pkc1, and the MAP kinase cascade (including the MAPK kinase kinase, the MAPK kinase and the MAPK KpMpk1) are involved in suppression of pexophagy and that the methanol concentration input signal branches out to two pathways: methanol-inducible gene expression (peroxisome synthesis) through transcriptional factors (such as KpMxr1, KpTrm1 and KpMit1) (Lin-Cereghino et al., 2006; Ohsawa et al., 2018, 2017; Parua et al., 2012; Sahu et al., 2014; Wang et al., 2016) and suppression of pexophagy through phospho-regulation of KpAtg30 (Fig. 5). The phosphatase KpPtp2A was also involved in phosphoregulation of KpAtg30, although we could not confirm its direct phosphatase action in vitro.

While this study shows that the KpWsc1–MAPK pathway negatively regulates pexophagy in K. phaffii, in the past the MAPK pathway has been reported to support pexophagy and mitophagy in a positive manner (Manjithaya et al., 2010; Mao et al., 2011). Although KpAtg30 and ScAtg36 are both receptor proteins for pexophagy, they are not homologs (Farré et al., 2008; Motley et al., 2012). The discrepancy may be due in part to the differences in pexophagy regulation and the molecular machinery between K. phaffii and S. cerevisiae. Methanol-induced peroxisomes in K. phaffii are larger than oleate-induced peroxisomes of K. phaffii and S. cerevisiae. Indeed, the peroxisome size has been reported to determine the requirement for Atg proteins in K. phaffii (Nazarko et al., 2009).

Previous reports have suggested the involvement of the MAPKs ScMpk1 and ScHog1 and their mammalian homologs the p38 MAPKs in regulation of mitophagy and pexophagy (Meijer and Codogno, 2006) (Manjithaya et al., 2010). Overexpression of ScMpk1 inhibits pexophagy in wild-type cells, but its moderate expression accelerates pexophagy (Manjithaya et al., 2010). However, because the input signal and/or environmental stimuli upstream of the MAPK are not known, the physiological impact of MAPKs on the regulation of autophagy is not understood well. Yeast pexophagy has been followed after a shift from peroxisome-proliferating medium (methanol or oleate medium) to glucose or ethanol medium, in which peroxisomes are not required for metabolism, or to nitrogen-starvation conditions. Since such drastic changes of the medium could trigger multiple input signals simultaneously, their complexity has made the analyses of signaling pathways and interpretation of data difficult. In the present study, we were able to focus on one single environmental input, methanol concentration, and analyzed the intracellular signaling pathway. Higher methanol concentrations (0.15–2.0%) were found to suppress pexophagy. Surprisingly, the single sensor molecule KpWsc1 transmitted the methanol concentration input signal and regulated not only peroxisome synthesis, but also pexophagy. This result provides a perspective on the molecular mechanism of organellar homeostasis in response to an environmental cue.

During batch culture of K. phaffii, the methanol concentration decreased with cultivation time, during which growing cells sense methanol via KpWsc1 and regulate the expression of genes encoding methanol-metabolizing enzymes (Ohsawa et al., 2017). In our previous study of C. boidinii proliferating on plant leaves, we observed that the processing of GFP–Atg8, which is dependent Atg1 but not on Atg30, occurs throughout the daily dark–light cycle. Conversely, Atg30-dependent pexophagy was only observed in the light period. The methanol concentrations at late exponential phase (after 16 h) are considered to resemble the plant leaf environment in the light period, where the methanol concentration decreases to 0–0.15%. We think that the molecular mechanism of negative regulation of selective autophagy – pexophagy during bulk autophagic conditions – plays a critical role by optimizing the quantity of peroxisomes as a survival strategy in the phyllosphere. Further physiological understanding at the molecular level will deepen our insights into the survival strategies of microbes in nature.

Escherichia coli DH10B (Takara Bio, Otsu, Japan) was used as a host strain for plasmid DNA propagation. E. coli cells were grown in LB medium (1% tryptone, 0.5% yeast extract, 0.5% NaCl) at 37°C.

The yeast strains used in this study are listed in Table S1. K. phaffii cells were grown on YPD (1% yeast extract, 2% peptone, 2% glucose) or YNB medium (0.67% yeast nitrogen base without amino acids). In the YNB medium, 2% (w/v) glucose (glucose medium) or 0.5% (v/v) methanol (methanol medium) were used as the carbon sources. Appropriate amino acids (100 μg/ml) were added to the synthetic media. All of the components other than the carbon sources used in these media were purchased from Difco Becton Dickinson (Franklin Lakes, NJ, USA). The growth of the yeast was monitored by the optical density (OD) at 600 nm (OD_600 nm).

The oligonucleotide primers used in this study are listed in Table S3. The plasmids used in this study are listed in Table S2. A deletion cassette for the KpATG11 gene was constructed as follows: primer pairs EcoRI-PpATG11-1-F/PpATG11-1-R and PpATG11-2-F/BamHI-PpATG11-2-R were used to amplify 1.0 kb fragments using genomic DNA as template. The primer pair PpATG11-BSD-F/PpATG11-BSD-R was used to amplify the blasticidin resistance gene using plasmid pPIC6A (Thermo Fisher Scientific, Waltham, MA, USA) as template. Using three fragments as the template, the primer pair EcoRI-PpATG11-1-F/BamHI-PpATG11-2-R was used to couple and amplify a 3.1 kb fragment by overlap PCR. This fragment was cloned into the EcoRI and BamHI sites of the vector pUC19 (Takara Bio, Otsu, Japan), yielding the KpATG11 disruption vector pIS100. A deletion cassette for the KpATG30 gene was constructed as follows: primer pairs HindIII-PpATG30-1-F/PpATG30-1-R and PpATG30-2-F/PstI-PpATG30-2-R were used to amplify 1.0 kb fragments using genomic DNA as template. The primer pair PpATG30-BSD-F/PpATG30-BSD-R was used to amplify the blasticidin resistance gene or zeocin resistance gene using plasmid pPIC6A or SK+Zeo^r, respectively. Using the three fragments as the template, the primer pair HindIII-PpATG30-1-F/PstI-PpATG30-2-R was used to couple and amplify a 3.1 kb fragment by overlap PCR. These fragments were cloned into the HindIII and PstI sites of the vector pIB1 (Takara Bio, Otsu, Japan), yielding the ATG30 disruption vectors pOH104 and pOH105. For the KpMPK1 deletion vector, primer pairs NotI-PpMPK1-1-F/KpnI-PpMPK1-1-R and SacI-PpMPK1-2-F/NotI-PpMPK1-2-R were used to amplify 1.0 kb fragments using genomic DNA as template. Using two fragments as the template, the primer pair KpnI-PpMPK1-1-R/SacI-PpMPK1-2-F was used to couple and amplify a 2.0 kb fragment by overlap PCR. This fragment was cloned into the KpnI and SacI sites of the vector SK+Zeo^r, yielding the KpMPK1 disruption vector pIS101. Deletion cassettes for the KpRLM1 and KpSWI4 genes were constructed similarly. Primer pairs NotI-PpRLM1-1-F/KpnI-PpRLM1-1-R and SacI-PpRLM1-2-F/NotI-PpRLM1-2-R were used for the KpRLM1 disruption vector pOH103. Primer pairs NotI-PpSWI4-1-F/KpnI-PpSWI4-1-R and SacI-PpSWI4-2-F/NotI-PpSWI4-2-R were used for the KpSWI4 disruption vector pOH106. For the KpMSG5 deletion vector, primer pairs PpMSG5-1-F/PpMSG5-1-R and PpMSG5-2-F/PpMSG5-2-R were used to amplify 1.0 kb fragments using genomic DNA as template. Primer pairs PpMSG5-ZEO-F/PpMSG5-ZEO-R were used to amplify the zeocin resistance gene using plasmid SK+Zeo^r. Using three fragments as the template, the primer pair PpMSG5-1-F/PpMSG5-2-R was used to couple and amplify a 3.1 kb fragment by overlap PCR. This fragment was ligated to Topo vector pCR2.1 (Thermo Fisher Scientific, Waltham, MA, USA), yielding the KpMSG5 disruption vector pOH107. The deletion cassette for the KpPTP2A was constructed similarly. Primer pairs PpPTP2A-1-F/PpPTP2A-1-R, PpPTP2A-2-F/PpPTP2A-2-R and PpPTP2A-ZEO-F/PpPTP2A-ZEO-R were used to construct the KpPTP2A disruption vector pOH108. The deletion cassette using ScARG4 as a marker for the KpMSG5 was constructed similarly. Primer pairs PpMSG5-1-F/(ScArg4)-PpMSG5-1-R and (ScArg4)-PpMSG5-2-F/PpMSG5-2-R were used to amplify 1.0 kb fragments using genomic DNA as template, and (PpMSG5)-ScARG4-F/(PpMSG5)-ScARG4-R was used to amplify an ScARG4 fragment using plasmid pSY8200 as template. Using three fragments as the template, the primer pair PpMSG5-1-F/PpMSG5-2-R was used to couple and amplify a 3.1 kb fragment by overlap PCR. This fragment was ligated to Topo vector pCR2.1, yielding the KpMSG5 disruption vector using ARG4 as a marker, pOH109.

pIS100 was digested with EcoRI/BamHI to disrupt the KpATG11 gene, and pOH104 and pOH105 were digested with HindIII/PstI to disrupt KpATG30. In order to disrupt the KpWSC1, KpMSG5 and KpPTP2A genes, each disruption vector – pOH100, pOH107 and pOH108, respectively – was digested with EcoRI. pIS101, pOH103 and pOH106 were digested with NotI to disrupt KpMPK1, KpRLM1 and KpSWI4. These digested DNA fragments were used to transform K. phaffii by electroporation. Proper gene disruptions were confirmed using colony PCR.

The KpATG30 promoter and its open reading frame (ORF) region without the STOP codon was amplified using primers XhoI-PpATG30-subclo-F/SphI-PpATG30-subclo-R with genomic DNA as template. The PCR fragment was cloned into the XhoI and SphI sites of the vector pSY006, resulting in pRN001. The primer pair KpnI-P_ACTI-F/SpeI-PpWSC1-subclo-R was used to amplify the KpWSC1 gene fragment using plasmids pOH202, pOH208 and pOH209 as the template, and cloned into the KpnI and SpeI sites of pOH206, resulting in pOH213, pOH215 and pOH216, respectively. The primer pair KpnI-P_ACTI-F/SpeI- PpWSC1(310-316Δ)-R was used to amplify the KpWSC1 gene fragment using plasmid pOH205 as the template. The PCR fragment was cloned into the KpnI and SpeI sites of pOH206, resulting in pOH214. The primer pair BamHI-P_ACTI-F/SpeI-PpWSC3-R was used to amplify the KpWSC3 gene fragment using plasmid pOH203 as the template. The PCR fragment was cloned into the BamHI and SpeI sites of pNT206, resulting in pOH217. The KpMPK1 promoter and ORF region without the STOP codon was amplified using primers XmaI-MPK1-F/(YFP)-PpMPK1-R, with genomic DNA as template. The YFP region for coupling the KpMPK1 ORF was amplified using primers (PpMPK1)-YFP-F/BamHI-YFP-R, using plasmid pNT205 as the template. Using the two fragments as templates, the primer pair XmaI-MPK1-F/BamHI-YFP-R was used to couple and amplify by overlap PCR. This fragment was cloned into the XmaI and BamHI sites of pNT204, resulting in pIS001. pIS001 was subjected to site-directed mutagenesis using PpMPK1(T188AY190F)-F and PpMPK1(T188AY190F)-R primers. The resultant plasmid was designated pIS002.

Aliquots of the culture medium were subjected to gas chromatography (GC) analysis. After removing yeast cells by centrifugation, 2 µl of the culture medium was injected onto a Porapak Q (Agilent technologies, Santa Clara, CA, USA) column. The injector temperature was 180°C. The detection temperature was held at 180°C for 4 min.

The samples for KpPex11–YFP cleavage assays were prepared from cells harvested at an OD_600 nm of 2.0. The cells were resuspended in 1 ml of solution I [0.2 M NaOH and 0.5% (v/v) 2-mercaptoethanol] and incubated on ice for 10 min, and then 0.1 ml of 100% (w/v) trichloroacetic acid (TCA) solution was added to the suspension. The lysates were subjected to centrifugation at 20,000 g at 4°C for 5 min. The supernatant was removed, and the pellet was resuspended in 1 ml of acetone by brief sonication. The obtained sample was subjected to centrifugation at 20,00 g at 4°C for 5 min, and the pellet was dried, dissolved in 80 μl of sample buffer [0.1 M Tris-HCl (pH 7.5), 2% (w/v) SDS, 1% (v/v) glycerol, 0.5% (v/v) 2-mercaptoethanol and 0.01% (w/v) Bromophenol Blue] by brief sonication, incubated at 65°C for 10 min, then subjected to centrifugation at 20,000 g for 1 min. A 5 μl volume of the supernatant was electrophoresed on a 10% SDS–PAGE gel.

To prepare samples for detection of phosphorylated KpAtg30–3×HA, harvested cells were suspended in lysis buffer [50 mM Tris-HCl pH 7.5, 50 mM NaCl, 0.1 mM EDTA, 0.1% Triton-X100, 10% (v/v) glycerol, 1 mM phenylmethylsulfonyl fluoride and EDTA-free complete protease inhibitor cocktail (Roche Diagnostics, Basel, Switzerland)] and then lysed using Multi-Beads Shocker (Yasui Kikai, Osaka, Japan). Cell extracts were subjected to centrifugation at 10,000 g for 5 min at 4°C to remove cell debris, dissolved in sample buffer (125 mM Tris-HCl pH 6.8, 4% SDS, 20% glycerol, a dash of Bromophenol Blue and 10% 2-mercaptoethanol) and boiled for 5 min. Each sample was electrophoresed on a 12% SDS–PAGE gel.

The proteins were transferred to a PVDF membrane using semidry blotting (ATTO, Tokyo, Japan). The blots were incubated overnight with anti-GFP antibody (JL-8; Clontech, Mountain View, CA, USA), anti-FLAG antibody (M2; Merck Millipore, Darmstadt, Germany), anti-HA antibody (F7; Santa Cruz Biotech, Dallas, TX, USA), anti-phospho-MAPK (Thr202/Tyr204; 20G11; Cell Signaling Technology, MA, USA) or anti-β-actin (ab8224; Abcam, Cambridge, UK) at a 1:1000 dilution in TBS-T [20 mM Tris, 0.1% (v/v) Tween 20, 0.8% (w/v) NaCl and 0.02% (w/v) KCl] buffer. The membranes were washed three times with TBS-T buffer and incubated with HRP-conjugated anti-mouse IgG (12-349; Merck Millipore, Darmstadt, Germany) or HRP-conjugated anti-rabbit IgG secondary antibodies (12-348; Merck Millipore) at a 1:10,000 dilution for 1 h. Finally, bound secondary antibodies were detected using Western Lightning (Perkin-Elmer Life Science, Waltham, MA, USA) and the signals were analyzed using a Light Capture system (ATTO, Tokyo, Japan).

K. phaffii cells were grown in 5 ml of YPD medium to the stationary phase at 28°C. Subsequently, 30 μl of this culture was transferred into 5 ml of fresh YPD medium and the cells were grown at 28°C for 5 h. The culture was then harvested by centrifugation at 176 g for 5 min, then resuspended in methanol medium containing 0.93 μg/ml FM4-64 (Molecular Probes, Eugene, OR, USA) to label the vacuolar membrane and grown at 28°C for 24 h. The cells were harvested by centrifugation and stored on ice until observation. Observations were carried out using an IX81 fluorescence microscope (Olympus, Tokyo, Japan). Fluorescent images were captured with a charged coupled device (CCD) camera (SenSys; PhotoMetrics, Tucson, AZ, USA) using MetaMorph software (Universal Imaging, West Chester, PA, USA).

A single colony was inoculated into YPD medium and cultivated overnight. Yeast cells were transferred into glucose medium to an OD_600 nm of 0.1 and cultivated to early exponential phase. The cells were shifted to methanol medium at an OD_600 nm of 1.0 and cultured at 28°C. Cells were harvested at the indicated time point by centrifugation at 10,000 g for 1 min at 4°C. The methods used for RNA extraction and reverse transcription were described previously (Ohsawa et al., 2017).

The quantitative real-time PCR was performed with a Light Cycler Instrument (Roche Diagnostics, Basel, Switzerland). The PCR reaction was performed with SYBR Premix Ex Taq (Takara) and the primers for KpGAP1, KpMAG5 and KpPTP2A are listed in Table S3. The copy numbers of each sample were estimated using the Light Cycler software Version 4.1.

We are grateful to Professor Emeritus Teruhiko Beppu, University of Tokyo, and Professor Kenji Ueda, Nihon university, for their advice and technical information on using Beppu flasks. We also thank Tomoyuki Takeya and other Sakai laboratory members for helpful discussion and their technical assistance.

The peer review history is available online at https://journals.biologists.com/jcs/article-lookup/doi/10.1242/jcs.254714