Yapsins are a family of glycosylphosphatidylinositol (GPI)-anchored aspartyl proteases that preferentially cut at the C-terminus of basic amino acids (Lys or Arg). In Saccharomyces cerevisiae, Yapsin 1 (formerly named Yap3p, for third yeast aspartic peptidase) was first identified as a proprotein convertase that could suppress the phenotype induced by disruption of another endopeptidase, Kex2. The second yapsin gene coding for ScMkc7p (alternatively called Yps2p) was also isolated as a multicopy suppressor of the cold sensitivity of the kex2-null mutant strain (Komano et al., 1995). Homology searches revealed the presence of three other yapsin genes, YPS3, YPS6 and YPS7, in the S. cerevisiae genome together with one pseudogene, ScYPS5, which codes for a truncated form (Olsen et al., 1999). Although ScYPS6 and ScYPS7 have a GPI-anchoring signal motif, their degree of similarity to ScYPS1, ScYPS2 and ScYPS3 is not greater than to other aspartic protease genes such as ScBAR1 and ScPEP4. Further, it is not yet biochemically defined whether ScYps6p and ScYps7p can specifically cleave the substrates at basic amino acids, as reported for ScYps1p, ScYps2p and ScYps3p (Ladds et al., 2000).
Several interesting characteristics of yapsins that distinguish them from other classes of aspartic peptidases, such as their unique domain organization (in association with a complex mode of zymogen activation) and unusual localization at the cell surface, have been important subjects for fundamental studies elucidating the in vivo functions and substrates of yapsins (Gagnon-Arsenault et al., 2008; Komano et al., 1995; Kuroda et al., 2007; Li et al., 2002). Recent reports of studies in S. cerevisiae have described the roles of yapsins in the maintenance of cell wall integrity by processing cell wall proteins (Gagnon-Arsenault et al., 2008; Krysan et al., 2005) and in a novel protein quality control by releasing overexpressed GPI-linked cell wall proteins (Miller et al., 2010). The important roles of yapsin-like proteases for virulence have been also indicated in pathogenic yeasts, such as Candida albicans and C. glabrata (Albrecht et al., 2006; Kaur et al., 2007). However, intensive biochemical characterization of purified yapsins and systematic phenotype analysis of multiple yapsin deletion mutant strains have not yet been carried out in yeast species other than S. cerevisiae and C. albicans (Albrecht et al., 2006; Schild et al., 2011). Therefore, the defined physiological substrates and in vivo function of yapsin members in other yeast species remains unclear and requires further study.
The thermotolerant methylotrophic yeast, Hansenula polymorpha (syn. Pichia angusta), is one of the most important industrial yeasts. This yeast has been highlighted as a promising host for recombinant protein expression, due to its well-established genetic tool boxes, high yield of protein production and simple process for high-cell density cultivation (Kang et al., 2005). Moreover, the yeast-specific N-glycosylation pattern in H. polymorpha was studied and engineered to produce mammalian-type glycans without yeast-specific immunogenic epitopes (Kim et al., 2004, 2006; Oh et al., 2008). On the other hand, peculiar physiological characteristics of H. polymorpha, such as resistance to heavy metals, oxidative stress and heat, make this yeast attractive for several biotechnological purposes (Blazhenko et al., 2006; Park et al., 2007). Particularly, its tolerance to elevated temperatures up to 48 °C has drawn attention to H. polymorpha as an emerging candidate for the production of biofuels in simultaneous saccharification and fermentation processes, which require thermotolerant microbial host strains (Dmytruk et al., 2008). In the present study, we identified two H. polymorpha yapsin genes based on sequence homology to S. cerevisiae yapsins from the whole genome of H. polymorpha, and investigated their domain structures, cellular localization, possible roles in cell wall integrity and proteolytic activities.
Materials and methods
Yeast strains, plasmids, culture conditions and drugs
The S. cerevisiae and H. polymorpha strains used in this study are listed in Table 1. H. polymorpha strains DL1-L (leu2) and DL1-LdU (leu2 ura3Δ::lacZ) were used as parental strains for the construction of the H. polymorpha yapsin-deficient strains. The primers used for the construction of strains and plasmids are listed in Table S1 (see Supporting information). The construction of the Hpyps1 and Hpyps7 null mutant strains are described in Supplementary materials and methods (see Supporting information). The plasmids used and constructed in this study are listed in Table 1 and the detailed descriptions for their constructions are described in Supplementary materials and methods (see Supporting information). Calcofluor white (Sigma-Aldrich, St. Louis, MO, USA), Congo red (Sigma-Aldrich), caffeine (Sigma-Aldrich), and caspofungin acetate (Merck, Whitehouse Station, NJ, USA) were used to prepare drug-supplemented yeast extract peptone dextrose (YPD) agar plates. Drug sensitivity was assayed by growth at 30 °C or 37 °C for 2–3 days.
Table 1. Yeast strains and plasmids used in this study
pBluescript containing the HpYPS7 gene disruption cassette
Purification of the C-terminal truncated HpYPS1 and HpYPS7 proteins
To induce the expression of H. polymorpha yapsins, transformants harbouring pMOX–YPS1ct–His or pMOX–YPS7ct–His were precultured overnight in a synthetic complete medium lacking leucine, after which they were transferred to flasks containing YPM medium (1% yeast extract, 2% peptone, 2% methanol) for 24 h at 37 °C. The resulting culture supernatants were concentrated by ultrafiltration (YM30 membrane, Millipore, Bedford, MA, USA) and dialysed against 50 mm sodium phosphate, pH 6.0, and 300 mm NaCl. The resulting His-tagged yapsins were purified using a Ni–NTA column on the ÄKTAprime chromatography system (GE healthcare, Buckinghamshire, UK). Immunoblot analysis of His-tagged yapsins was carried out using polyclonal antibodies raised against a histidine tag (Qiagen, Valencia, CA, USA). For deglycosylation experiments, the secreted HpYps1p and HpYps7p were digested with PNGase F under the conditions described by the supplier (New England Biolabs, Beverly, MA, USA).
Confocal microscopy and cell wall protein fractionation analysis
To investigate cellular localization, pDLMOX–yEGFP, pDLMOX–yEGFP–HpYPS1(C40) or pDLMOX–yEGFP–HpYPS7(C40) were constructed (see Supporting information, Supplementary materials and methods) and transformed to yeast cells grown in a methanol-containing medium for 20 h. Images of freshly cultured cells were acquired using an LSM510 confocal microscope (Carl Zeiss GmbH, Jena, Germany). H. polymorpha cell wall proteins (CWPs) were isolated by previously described methods with slight modification (Frieman et al., 2003; Pitarch et al., 2008). Briefly, yeast cells grown in YPM medium for 48 h were harvested, washed and broken with glass beads (425–600 µm) in a multi-bead beater (Precellysis 24, Bertin technologies, France). The crude lysates and cell wall pellets were obtained by centrifugation at 1000 × g for 5 min. To remove non-covalently linked proteins and intracellular crude lysate, the isolated cell wall pellets were washed sequentially with washing solution A (distilled water with 1 mm PMSF), washing solution B (5% NaCl with 1 mm PMSF), washing solution C (2% NaCl with 1 mm PMSF), washing solution D (1% NaCl with 1 mm PMSF), and again with washing solution A at 1000 × g for 5 min. SDS-extractable CWPs were twice extracted for 5 min at 100 °C with 50 mm Tris–HCl, pH 7.8, 2% SDS, 100 mm Na–EDTA and 40 mmβ-mercaptoethanol, and were precipitated by trichloroacetic acid. SDS-treated walls were washed three times with water and freeze-dried. GPI-anchored CWPs were released by resuspending the cell walls in 800 µl undiluted HF–pyridine (Sigma-Aldrich) at 0 °C and incubating them for 3 h. The reaction was quenched by diluting the reaction mixture with an equal amount of ice-cold H2O. Then, the HF–pyridine extractable CWPs were freeze-dried. Crude lysate, SDS- and HF–pyridine-extractable proteins were separated by SDS–PAGE, stained with Coomassie brilliant blue R-250 or analysed with Western blotting, using the antibody raised against GFP (Abcam, Cambridge, UK).
Northern blot analysis and quantitative real-time PCR
Yeast cells were grown to early-exponential phase (OD600 = 0.3) in YPD medium and then incubated for another 2 h after adding Congo red, Calcofluor white, caspofungin, caffeine, sodium dodecyl sulphate or tunicamycin, or applying heat shock. Total RNA was prepared using the hot-phenol method, electrophoresed on 1.2% agarose-formaldehyde gels, blotted overnight onto a Nylon+ membrane, and hybridized with 32P-labelled DNA probes. DNA probes for HpYPS1 and HpYPS7 were synthesized by PCR, using the following oligonucleotides pairs: YPS1(F)/YPS1(R) and YPS7(F)/YPS7(R), respectively. Probes were labelled with the RediprimeTM II random priming labelling system kit (GE Healthcare) according to the manufacturer's recommended method. For quantitative real-time PCR analysis, H. polymorpha yapsin gene-specific primer sets and β-actin primer as a reference gene were used (see Supporting information, Table S1). Briefly, cDNA was synthesized from 5 µg total RNA, using 200 units of SuperscriptTM reverse transcriptase (Invitrogen, Carlsbad, CA, USA). Quantitative real-time PCR was performed in Rotor-Gene Q (Qiagen), using 10 pg of the synthesized cDNA and 1 µl forward and reverse primers (10 pmol/l) with a QuantiMix SYBR Kit (Philekorea Technology, Daejeon, Korea). The specificity of amplification was confirmed by melting curve analysis with one single peak. Each sample was analysed in triplicate and normalized to the endogenous control, β-actin. The relative expression levels of YPS mRNAs were calculated using 2–ΔΔCT methodology.
Only two yapsin genes, HpYPS1 and HpYPS7, are present in the H. polymorpha genome
To identify H. polymorpha yapsin genes, we searched the entire H. polymorpha genome database (Ramezani-Rad et al., 2003), using BLAST with five S. cerevisiae yapsin genes (ScYPS1, -2, -3, -6 and −7) as queries. Only two open reading frames (ORFs) of H. polymorpha were found with significant similarities to S. cerevisiae YPS1 and YPS7. They were assigned as H. polymorpha YPS1 (HpYPS1) and YPS7 (HpYPS7) genes, based on their highest homology to the corresponding S. cerevisiae yapsin genes, respectively (see Supporting information, Table S2). The amino acid sequence identity and similarity between HpYps1p and ScYps1p were 37% and 52%, respectively, and 23% and 43% between HpYps7p and ScYps7p, respectively. Moreover, HpYps1p also showed 29% and 22% identities to Sap9p and Sap10p, the GPI-anchored yapsins in pathogenic yeast C. albicans, respectively, while HpYps7p showed only low identity to both C. albicans yapsins (12%).
Sequence analysis revealed that the ORFs of HpYPS1 and HpYPS7 encoded 576 and 549 amino acids proteins, respectively. Both of the ORFs have N-terminal signal peptides as well as putative glycophosphatidylinositol (GPI) anchor attachment signals at their C-terminus (Figure 1A). Following analysis with a big-PI predictor programme (Eisenhaber et al., 1999), the GPI-anchoring amino acids (the ω sites) of HpYps1p and HpYps7p were predicted to be Asn552 and Ser524, respectively. HpYps1p and HpYps7p, were predicted to potentially contain three and nine N-glycosylation sites, respectively. Moreover, both proteins also had a putative Ser/Thr rich domain for O-glycosylation in the front of GPI-anchoring site. Therefore, HpYps1p and HpYps7p have all three components required for GPI proteins: (a) an N-terminal signal peptide; and (b) a serine/threonine-rich region; followed by (c) an ω site for GPI attachment in the C-terminal region. The nucleotide sequences of HpYPS1 and HpYPS7 were deposited in GenBank with Accession Nos AF493990 and AY570527, respectively.
H. polymorpha yapsins are largely localized to the cell wall
One of the unique characteristics of yapsins is their surface localization, due to the presence of a GPI attachment signal that directs localization to either the plasma membrane or cell wall. The presence of a dibasic site near the ω site for GPI attachment has been proposed to be the signal for the retention of yeast GPI proteins at the plasma membrane, whereas the absence of this retention signal is considered to be the signal for movement to the cell wall (Caro et al., 1997). With the exception of ScYps7p, all four S. cerevisiae yapsins have dibasic motifs near ω sites and ScYps1p and ScYps2p have been experimentally confirmed to localize at the plasma membrane (Gagnon-Arsenault et al., 2006; Hamada et al., 1999). Intriguingly, HpYps1p and HpYps7p lack these dibasic motifs, suggesting that they might be localized at the cell wall.
In order to define the cellular localization of HpYps1p and HpYps7p, we constructed the vector for expression of GFP reporter proteins fused with the C-terminal 40 amino acids of HpYps1p and HpYps7p containing the GPI anchoring motif, respectively, together with the signal peptide of HpYps1p (Figure 1B, top). While the original GFP was expressed exclusively in the cytosol of the cell, the GFP reporters fused with the signal peptide and the GPI anchoring motifs [GFP-HpYps1(C40) and GFP-HpYps7(C40)] were detected both on the cell surface and inside the cell (Figure 1B, bottom). The green fluorescence observed inside the cells might be presumably vacuoles where misfolded reporter proteins had accumulated for degradation, which has been previously reported in S. cerevisiae (Li et al., 2002). These results clearly revealed that the 40 C-terminal amino acids of HpYps1p or HpYps7p together with the signal peptide were sufficient to direct the surface expression of GFP reporter.
To clarify further whether H. polymorpha yapsins localize at the plasma membrane or at the cell wall, we analysed SDS and HF–pyridine extractable proteins from the membrane and cell wall fractions of the H. polymorpha transformants expressing the GFP reporters. After extraction of proteins using SDS and DTT (Figure 1C, lane 2), the GPI-anchored CWP fraction was obtained by treatment of HF–pyridine (Figure 1C, lane 3), which is known to specifically cleave the phosphodiester bonds linking GPI-anchored CWPs to β(1,6)-glucan chains (de Groot et al., 2004). Western blot analysis clearly showed that both GFP–HpYps1(C40) and GFP–HpYps7(C40) were significantly enriched in the GPI-anchored CWP fraction obtained by HF–pyridine treatment. The bands detected in the total cell lysate fraction (Figure 1C, lane 1) were speculated to be misfolded forms of GFPs entrapped in vacuoles, with a minor smaller band representing a degraded form of the GFP reporter. The finding that HpYps1p was highly enriched in the HF–pyridine-cleavable fraction strongly supports that most HpYps1ps are linked to the cell wall via a GPI anchorage. On the other hand, HpYps7p was found to be equally distributed to both SDS and HF–pyridine extracted fractions, implying that a significant portion of HpYps7ps are localized at the plasma membrane as well as the cell wall. Together with the results of fluorescence microscopic analysis, these results strongly support that HpYps1p and HpYps7p are localized on the cell surface via the GPI anchoring motif, preferentially at the cell wall.
HpYPS1 and HpYPS7 differentially complement S. cerevisiae yapsin deficiencies
To examine whether HpYps1p and HpYps7p can act as functional homologues of S. cerevisiae yapsins, we carried out complementation tests in the S. cerevisiae yps-deletion mutant strains, Scyps1Δ, Scyps7Δ and Scyps1Δyps2Δyps3Δ, which show varying degrees of growth defects when plated with cell wall-perturbing reagents or at high temperatures (Cho et al., 2010; Krysan et al., 2005). The S. cerevisiae mutant strains were transformed with the plasmids, YEp352GAPII–HpYPS1 or YEp352GAPII–HpYPS7, which express HpYPS1 or HpYPS7 under the control of the S. cerevisiae GAP promoter, respectively.
Notably, the expression of HpYPS1 was able to complement all growth defects of the triple yps-deletion mutant, Scyps1Δyps2Δyps3Δ, as well as the single disruptant, Scyps1Δ (Figure 2A). The triple Scyps1Δyps2Δyps3Δ deletion mutant showed severe sensitivities to all of the cell wall-perturbing reagents tested and also displayed a temperature-sensitivity phenotype (Figure 2). It was remarkable that the introduction of only a single gene, HpYPS1, was able to overcome all of the growth retardation effects seen in Scyps1Δ yps2Δ yps3Δ, but not those of Scyps7Δ in the presence of Congo red, indicating a unique function of ScYps7p in maintaining cell wall integrity. Interestingly, the HpYPS7 gene could not complement the growth defects of Scyps7Δ on the plate containing Congo red; however, it did restore the growth retardation of Scyps7Δ to normal levels on a plate containing Calcofluor white. The hypersensitivity of Scyps7Δ to Calcofluor white, a dye that binds and disrupts chitin polymers, was previously reported to be a unique characteristic of Scyps7Δ distinguished from other single yps-disruptants (Krysan et al., 2005). Moreover, it was surprising that HpYPS7 was able to partially rescue the growth defect of Scyps1Δyps2Δyps3Δ on the plate containing Calcofluor white, although the complementation effect was quite marginal. It seems that HpYps7p was able to repair the defect caused by Calcofluor white in S. cerevisiae yapsin mutants, but not the sensitivity to Congo red. Thus, its complementation ability depended on the type of cell wall stress and was also limited compared to HpYps1p. Taken together, these results suggest that HpYps1p can completely substitute for the function of ScYps1p, which plays a crucial role in cell wall synthesis as well as cell wall stress responses, while HpYps7p shares only limited functions with ScYps7p.
Loss of yapsin function in H. polymorpha causes marginal growth defects under cell wall stress condition
To characterize the growth phenotype associated with loss of yapsin function in H. polymorpha, yapsin deletion mutants with a single HpYPS1 deletion (Hpyps1Δ), a single HpYPS7 deletion (Hpyps7Δ), or a double HpYPS1/HpYPS7 deletion (Hpyps1Δyps7Δ) were constructed and analysed for their growth phenotypes. The H. polymorpha yps mutants did not exhibit any apparent growth retardation under normal conditions and even at elevated temperature, while the Hpyps1Δ and Hpyps1Δyps7Δ strains showed slightly increased sensitivity to hygromycin B (Figure 3A). Especially, different from the Scyps1Δ and Scyps7Δ mutants, all of the H. polymorpha yps mutants did not show any significant growth defects induced by cell wall-perturbing reagents, such as Congo red, Calcofluor white, caspofungin and caffeine, when compared with the H. polymorpha wild-type strain (Figure 3B). Only when increasing the concentration of Calcofluor white up to 3 mg/ml, the Hpyps1Δ and the Hpyps1Δyps7Δ mutants showed slightly decreased growth phenotype (see Supporting information, Figure S1 and Table S3). Interestingly, the growth of H. polymorpha wild-type and yapsin deletion mutants was almost not affected by Congo red, Calcofluor white or caffeine at the tested concentrations, whereas the S. cerevisiae wild-type as well as yps1- and yps7-disrutant strains did show growth defects, reflecting the considerable difference in cell wall structure and organization between the two yeast species. Indeed, only to caspofungin did the H. polymorpha strains show more severe sensitivity than S. cerevisiae; however, there was still no significant difference between the H. polymorpha wild-type and yapsin deletion strains with respect to the sensitivity.
H. polymorpha YPS expression is modulated by environmental stresses
The unexpected resistance of H. polymorpha yapsin mutants to the tested cell wall stressor agents urged us to investigate the expression pattern of H. polymorpha yapsin genes under conditions of cell wall stress. In the case of S. cerevisiae, the expression of several yapsin genes have been reported to be modulated by environmental stresses (Ash et al., 1995; Bourbonnais et al., 2001; Gagnon-Arsenault et al., 2006). Specifically, the expression of ScYPS1 was shown to be dramatically upregulated in response to cell wall stress, such as heat shock and treatment with perturbing reagents (Krysan et al., 2005). Northern blot analysis of H. polymorpha RNA samples obtained after treatment with Calcofluor white, caspofungin and heat shock (Figure 4A, B) revealed that expression of the HpYPS1 gene was not induced under the tested conditions. In contrast to HpYPS1, expression of the HpYPS7 gene appeared to decrease in response to stress conditions. Similar repression of the ScYPS7 gene has been reported in several genome-wide transcriptome analyses against cell wall stresses (Causton et al., 2001; Garcia et al., 2004). Even at high concentrations of drugs which began to show inhibitory effects on the growth of both wild-type and yps-mutant strains, we could not observe any apparent induced expression of the HpYPS genes by quantitative real-time PCR analysis (Figure 4D). In an effort to enhance the phenotype effects induced by cell wall-perturbing reagents, we used 50% YPD plates but found that the expression changes induced by Congo red, Calcofluor white and caffeine were still marginal (data not shown). In an attempt to search cell wall-perturbing reagents to affect the HpYPS expression, we observed that the treatment of SDS, a detergent and protein destabilizer, significantly induced the expression of both HpYPS genes (Figure 4D).
Our recent microarray analysis of the H. polymorpha transcriptome under unfolded protein response conditions indicated that expression of H. polymorpha yapsin genes are induced upon the treatment with tunicamycin (TM), a N-linked glycosylation inhibitor (Moon et al. unpublished data). Consistently, we confirmed a remarkably increased expression of HpYPS1 upon TM treatment along with a moderate increase of HpYPS7 expression by Northern blot analysis (Figure 4C). A very recent study performed in S. cerevisiae also reported the increased expression of a reporter protein fused to the ScYPS1 promoter by TM treatment (Miller et al., 2010). It was previously shown that the cell wall integrity pathway was activated during TM-induced ER stress, indicating that TM could cause not only ER stress but also cell wall stress (Chen et al., 2005). Our results indicate that, although the induced expression of HpYPS genes was not apparent by the treatment of reagents perturbing chitin and glucan chain assembly, such as Calcofluor white and Congo red, the HpYPS expression would be more tightly regulated by signals generated by reagents destabilizing cell wall proteins, such as SDS and TM.
During real-time PCR analysis, we found that the expressions of HpYPS1 were generally five times higher than HpYPS7 (Figure 4D). This was consistent with the result of Northern blot analysis, displaying the much weaker signals of the HpYPS7 transcript compared to those of the HpYPS1 transcript. These results clearly indicate that the intrinsic steady-state level of HpYPS7 mRNA was much lower than that of HpYPS1 mRNA.
HpYps1p comprises two subunits, whereas HpYps7p consists of only one single polypeptide
For biochemical characterization of H. polymorpha yapsins, HpYps1p and HpYps7p were expressed and purified as soluble forms lacking a GPI signal (deletion from ω +4 position to C-terminus) but tagged with histidine residues at their C-terminus. The highly glycosylated forms of HpYps1p and HpYps7p were detected by Western blot analysis, using an anti-His-tag antibody (Figure 5A). HpYps1p displayed a doublet band in the range 65–100 kDa, while HpYps7p was detected as a single broad band of a higher molecular weight (~150–200 kDa). After cleavage of signal peptides, the molecular weights of HpYps1p and HpYps7p were predicted to be approximately 55 and 57 kDa, respectively. The discrepancy between the observed electrophoretic mobility and the predicted molecular weight could be explained partly by differences in the N-glycosylation (nine putative sites for HpYps7p and three sites for HpYps1p) and the domain structures resulting from proteolytic processing. As expected, PNGase F treatment removing N-glycans resulted in more dramatic changes to the electrophoretic mobility of HpYps7p compared to that of HpYps1p. However, even after PNGase F deglycosylation, both of the HpYps proteins continued to be detected as heterogeneous forms at higher molecular weights than calculated, implying the presence of O-glycosylation.
ScYps1p is reported to consist of two subunits, α and β, which are generated by cleavage at the loop region of the precursor protein (Azaryan et al., 1993; Olsen et al., 2000). To determine whether H. polymorpha yapsins are also subjected to similar proteolytic maturation, the purified C-terminally truncated HpYps1p and HpYps7p were analysed by both reducing and non-reducing SDS–PAGE (Figure 5B, C). The purified HpYps1p displayed a broad doublet band at 65–100 kDa under non-reducing conditions, while it showed two clearly separated fragments (60–85 kDa and ~25 kDa) under reducing conditions. It seems that the ~25 kDa fragment detected on the reducing gel corresponds to the N-terminal α-subunit of HpYps1, since the corresponding smaller band was not detected in Western blot analysis using the anti-His-tag antibody (cf. Figures 5A and 5B). The position of the lower band of the doublet on the non-reducing gel matched the upper band on the reducing gel, suggesting that the higher band may correspond to the whole HpYps1 protein comprising the αβ-dimer, and the lower band might correspond to HpYps1p containing only the β-subunit. Under reducing conditions, the intensity of the larger fragment band appeared to be much higher than that of the smaller band (α-subunit), even after considering the molecular weight difference. Thus, we speculated that the αβ-dimer and the C-terminal β-subunit were enriched, while a substantial portion of the smaller N-terminal α-subunit was lost during affinity purification using Ni-NTA column. The N-terminal amino acid sequence analysis revealed that the small α- and large β-fragments in the reducing gel had A50DDGSV and G139FLGGW sequences, respectively. Propeptide removal and the two-subunit splitting of HpYps1p also occurred after dibasic-(K48R49) and monobasic-(K138) cleavages, respectively. The domain structure of HpYps1p proposed in Figure 1A, comprising two subunits connected together by disulphide linkages, appears to be similar to the domain structure of ScYps1p. In contrast to HpYps1p, the purified HpYps7p analysed by SDS–PAGE under either non-reducing or reducing conditions was detected as a single band of ~150–200 kDa (Figure 5C), indicating that HpYps7p consists of only a single polypeptide. Despite several repeated attempts, we were unable to obtain information about the N-terminal amino acid sequences of purified HpYps7p.
H. polymorpha yapsins show proteolytic cleavage activity at cell surface
In order to investigate the proteolytic activities of H. polymorpha yapsins, the purified soluble HpYps1p and HpYps7p were tested for proteolytic activity, using human parathyroid hormone (hPTH) as a substrate. When recombinant hPTH is expressed in S. cerevisiae, the main problem is a reduced yield of the intact form due to the aberrant proteolytic cleavage that occurs during secretion (Gabrielsen et al., 1990). In our previous study, it was demonstrated that this proteolytic cleavage of hPTH was mainly mediated by ScYps1p but also caused by other yapsins (Cho et al., 2010; Kang et al., 1998). Figure 6 shows the results of SDS–PAGE analysis of the hPTH digestion by purified HpYps1p and HpYps7p at 37 °C. The purified HpYps1p (70 ng) cleaved hPTH (3 µg) into three fragments, D1, D2 and D3, after just a 15 min reaction time (Figure 6A). In contrast, it took more than 2 h for HpYps7p (100 ng) to begin proteolytic cleavage of the same amount of hPTH (3 µg) into two fragments, D1 and D2, and a substantial amount of hPTH remained in the intact form even after a 4 h treatment (Figure 6B). The proteolytic products D1 and D2, generated by the cleavage at a dibasic site (R25K26), were commonly observed in both HpYps1p and HpYps7p digestion products. In contrast, D3 could be seen only in the digestion with HpYps1p. We assumed that D3 would be a proteolytic product of D1 generated by an additional cleavage at its C-terminus, which was confirmed by N-terminal amino acid sequencing of the fragment. The results of our proteolytic study show that HpYps1p displays much stronger proteolytic activity for hPTH than HpYps7p.
The decreased protein degradation activity of the H. polymorpha yapsin-disrupted mutants was also examined by proteolysis analysis of intact hPTH, using culture supernatants derived from the Hpyps1Δ and Hpyps7Δ strains as enzyme sources (Figure 6C). This assay was designed based on a previous report showing that significant amounts of ScYps1p can be detected in the extracellular medium after overnight culture (Ash et al., 1995), due to phosphatidylinositol-specific phospholipase C activity and mechanical shearing forces. Intact hPTH was added to the culture supernatant collected from the H. polymorpha wild-type, Hpyps1Δ, Hpyps7Δ and Hpyps1Δyps7Δ strains. While hPTH underwent rapid degradation within 3 h in the culture supernatant of the wild-type strain, the proteolysis was greatly retarded in the culture supernatants of the Hpyps1Δ and Hpyps1Δyps7Δ strains (Figure 6C). More than 50% of hPTH remained in its intact form even after 9 h incubation with the culture supernatants of the Hpyps1Δ and Hpyps1Δyps7Δ strains. However, the disruption of HpYPS7 did not contribute to the reduction of the hPTH degradation, as the culture supernatant of Hpyps7Δ strain rapidly degraded hPTH within 3 h (data not shown), with the same extent of degradation observed with the wild-type strain. Considering the much higher steady-state level of HpYPS1 mRNA, these results strongly indicate that HpYps1p would be a major yapsin protease responsible for the proteolytic cleavage of hPTH in the culture supernatant of H. polymorpha.
The potential of the Hpyps1Δ strain as a useful host for the production of recombinant proteins was validated by the secretory expression of recombinant hPTH in the H. polymorpha wild-type and Hpyps1Δ strain. The recombinant hPTH secreted from the wild-type strain was subjected to rapid degradation, and thus any detectable form of hPTH was not found in the culture supernatant of the wild-type strain. In contrast, the recombinant hPTH secreted from the Hpyps1Δ strain remained mostly in an intact form after 12 h culture, and was still detected as a major band together with degraded fragments even after 24 h culture. This result shows that Hpyps1Δ could be efficiently employed as an expression host for the economic production of recombinant hPTH (Figure 6D). Reduction of aberrant cleavage of another secretory recombinant protein, human serum albumin fused with tissue inhibitor metalloproteinase-2 (HSA-TIMP2), was also clearly observed in the Hpyps1Δ strain (see Supporting information, Figure S2). Specifically, a single intact form of the HSA-TIMP2 (88 kDa) fusion protein of intact size was secreted from the Hpyps1Δ strain without any detectable fragmented products. Together, these results clearly indicate that the Hpyps1Δ strain is useful for the production of the secretory proteins, especially those derived from humans and mammals.
Yapsins are a family of GPI-anchored aspartyl proteases located on the cell wall or plasma membrane of yeast and fungal cells. In the present study, we characterized the purified yapsins and deletion mutant strains of the thermotolerant methylotrophic yeast H. polymorpha to investigate their domain structures, cellular localization and possible roles in cell wall integrity and secretory protein processing. In contrast to the five yapsin members in S. cerevisiae (Olsen et al., 1999), only two yapsin homologues, HpYPS1 and HpYPS7, were identified from the entire H. polymorpha genome database. H. polymorpha yapsins were shown to be preferentially localized at cell walls (Figure 1), which is consistent with the prediction made from sequences analysis that H. polymorpha yapsins lack the dibasic motifs near ω sites that act as a membrane retention signal (Hamada et al., 1999). Biochemical analysis of the purified H. polymorpha yapsin proteins revealed that HpYps1p is proteolytically processed into two fragments connected together by disulphide linkages, similar to ScYps1p. Conversely, HpYps7p consisted of a single polypeptide chain (Figure 5), which was predicted based on the absence of a dibasic site in the loop region of HpYps7p (Gagnon-Arsenault et al., 2006). The in vitro proteolysis reaction using hPTH as a substrate revealed that HpYps1p has a much stronger yapsin protease activity compared to HpYps7p (Figure 6). However, it is noteworthy that HpYps7p displayed a yapsin-like protease activity, even though it was much weaker compared to HpYps1p. As far as we know, this is the first biochemical data showing that purified Yps7p has yapsin-like activity with substrate specificity for dibasic amino acid pairs.
Recently, Gagnon-Arsenault et al. reported that three putative yapsins (ScYps7p, HpYps7p and Schizosaccharomyces pombe Yps1p) have a motif around the catalytic Asp residues that differs completely from other yapsins, and questioned whether these putative yapsins really belong to the yapsin family; a matter that will require experimental investigation (Gagnon-Arsenault et al., 2006). Our results clearly showed that HpYps7p has essential features that are characteristic of the yapsin family. Indeed, HpYps7p was able to cleave the C-terminus of paired basic amino acids and was localized at the yeast cell surface as a GPI-anchored protein; however, it remains to be elucidated whether the weak proteolytic activity of HpYps7p originated from the different catalytic mechanisms predicted by sequence analysis.
The YPS null mutant strains of S. cerevisiae showed growth defects on media containing cell wall-directed drugs and at high temperature. Consistent with its essential role in cell wall integrity, the expression of S. cerevisiae YPS1 was highly induced by treatment of cell wall-perturbing reagents, such as Congo red and Calcofluor white, and upon shift to 37 °C (Krysan et al., 2005). Similarly, the C. albicans mutant strains with deletion of Sap9 and Sap10 were more sensitive to Calcofluor white and Congo red (Albrecht et al., 2006), and the expression of SAP9 gene was induced by the treatment of caspofungin (Copping et al., 2005). Therefore, it was unexpected that disruption of H. polymorpha yapsins, including double disruption of both HpYPS1 and HpYPS7, did not generate significant change with respect to resistance to several cell wall-perturbing reagents and to heat shock. Further, we observed that H. polymorpha was, in general, more resistant to the tested drugs, except caspofungin, compared to S. cerevisiae (Figure 3). It is notable that expression of the HpYPS genes was significantly induced only by a subset of cell wall stressor agents affecting the integrity of cell wall proteins, such as SDS and TM, in H. polymorpha (Figure 4). This result may be related to the different cell wall organizations among yeast species. We also speculate that the effect of yapsin gene disruption might be minimized in H. polymorpha, due to its thermotolerance, which is closely associated with cell wall integrity (Fuchs et al., 2009).
On the other hand, from a biotechnological aspect, this feature makes H. polymorpha yapsin-disrupted mutants more attractive as an expression host. The culture conditions for industrial fermentation are sometimes very harsh with high cell density and shearing forces that can induce cell wall stress. Moreover, the treatment of TM, a reagent that causes ER stress, was shown to remarkably induce HpYPS1 expression (Figure 4C), implying that the level of HpYps1p can be increased upon overexpression of recombinant secretory proteins that often cause ER stress. Indeed, this may be another advantage of the yapsin deletion mutant strain as a host strain for secretory protein production. Therefore, the H. polymorpha yapsin-disrupted mutants, which were resistant to cell wall stress, should have a significant advantage in industrial applications. To understand the different requirement for the function of yapsins in maintaining cell wall integrity between S. cerevisiae and H. polymorpha, comparative analysis on the physiological substrates of yapsins and on the cell wall organization between two yeast species are currently in progress.
This study was partly supported by grants from the Korean Ministry of Education, Science and Technology (Microbial Genomics and Applications R&D Programme and KRIBB Research Initiative Programme) and by grants from the Korean Ministry of Rural Development Administration (Next-Generation BioGreen 21 Programme Nos ADFC2011-PJ008330 and SSAC-PJ008001).