Disruption of PMR1, Encoding a Ca²⁺-ATPase Homolog in Yarrowia lipolytica, Affects Secretion and Processing of Homologous and Heterologous Proteins

Abstract

The Yarrowia lipolytica PMR1 gene (YlPMR1) is a Saccharomyces cerevisiae PMR1 homolog which encodes a putative secretory pathway Ca²⁺-ATPase. In this study, we investigated the effects of a YlPMR1 disruption on the processing and secretion of native and foreign proteins in Y. lipolytica and found variable responses by the YlPMR1-disrupted mutant depending on the protein. The secretion of 32-kDa mature alkaline extracellular protease (AEP) was dramatically decreased, and incompletely processed precursors were observed in the YlPMR1-disrupted mutant. A 36- and a 52-kDa premature AEP were secreted, and an intracellular 52-kDa premature AEP was also detected. The acid extracellular protease activity of the YlPMR1-disrupted mutant was increased by 60% compared to that of the wild-type strain. The inhibitory effect of mutations in secretory pathway Ca²⁺-ATPase genes on the secretion of rice α-amylase was also observed in the Y. lipolytica and S. cerevisiae PMR1-disrupted mutants. Unlike rice α-amylase, the secretion of Trichoderma reesei endoglucanase I (EGI) was not influenced by the YlPMR1 disruption. However, the secreted EGI from the YlPMR1-disrupted mutant had different characteristics than that of the control. While wild-type cells secreted the hyperglycosylated form of EGI, hyperglycosylation was completely absent in the YlPMR1-disrupted mutant. Our results indicate that the effects of the YlPMR1 disruption as manifested by the phenotypic response depend on the characteristics of the reporter protein in the recombinant yeast strain evaluated.

In living cells, the control of intracellular calcium (Ca²⁺) concentrations is critically important to the regulation of cellular processes such as muscle contraction, neurotransmitter release, and cell proliferation (1). In addition, Ca²⁺ is involved in the transport of secretory proteins from the endoplasmic reticulum (ER) (30). Intracellular compartments involved in the secretory pathway, particularly the ER and Golgi apparatus, contain higher concentrations of Ca²⁺ (∼1 mM) than the cytoplasm (≤0.1 μM) under normal conditions (1, 28). This Ca²⁺ concentration differential is normally maintained by Ca²⁺-ATPases, and a massive Ca²⁺ efflux from subcellular compartments to the cytoplasm can cause a specific cellular response depending on the external signal.

Two types of Ca²⁺-ATPases, plasma membrane Ca²⁺-ATPases (PMCA) and sarco/endoplasmic reticulum Ca²⁺-ATPases (SERCA), have been classified based on genetic and biochemical analyses (7). Recently, the presence of a new type of Ca²⁺-ATPase distinct from PMCA and SERCA has been proposed (11). Rudolph et al. identified a novel P-type ATPase encoded by PMR1 which functions as a Ca²⁺ pump and affects transit through the secretory pathway in the yeast Saccharomyces cerevisiae (29). The PMR1 gene is localized to the Golgi apparatus and was shown to be required for normal Golgi function (2). Sorin et al. provided biochemical evidence that Pmr1 is indeed a Golgi-specific Ca²⁺ pump and is distinct from both SERCA and PMCA (31). Interestingly, the PMR1-disrupted mutant displays pleiotropic changes, such as a Ca²⁺-dependent growth defect, secretion of an unprocessed α factor, suppression of various sec mutants blocked in ER and/or Golgi and post-Golgi transport, and incomplete outer-chain glycosylation (2, 13). These phenotypes can be reversed by the addition of a high concentration of Ca²⁺ (10 mM) to the medium, implicating a direct role for exogenous calcium in Golgi function (2). Furthermore, disruption of the PMR1 gene resulted in a 5- to 50-fold increase in the secretion of bovine prochymosin, a bovine growth hormone, and a nonglycosylated variant of human urinary plasminogen activator (14, 29). In contrast, secretion of the plant protein thaumatin could not be improved to any significant extent by disruption of the PMR1 gene (14). These results indicate that the PMR1 gene product plays an important role in the yeast secretory pathway.

Yarrowia lipolytica secretes high levels of several extracellular enzymes, including alkaline extracellular protease (AEP), RNase, lipase, and acid proteases (5, 15). In fact, AEP is secreted at a level of more than 1 g/liter under optimal conditions. The high-level secretion capacity of Y. lipolytica coupled with detailed studies on AEP secretion and processing make this yeast strain an excellent model system for studying protein secretion (22, 23).

Recently, we have cloned the Y. lipolytica PMR1 gene (YlPMR1), which is a Saccharomyces cerevisiae PMR1 homolog, and characterized the YlPMR1-disrupted mutant in the yeast Y. lipolytica (26). In the present paper, the effects of the YlPMR1 disruption on the processing and secretion of homologous and heterologous proteins are described. AEP, acid extracellular protease (AXP), rice α-amylase, and Trichoderma reesei endoglucanase I (EGI) were used as reporter proteins to illustrate the findings.

MATERIALS AND METHODS

Strains and media.

The yeast strains and plasmids used in this work are described in Table 1. The SMS397A (MATa ade1 ura3 xpr2) strain derived from Y. lipolytica CX161-1B (MATa ade1; ATCC 32338) was used to construct the YlPMR1-disrupted mutant designated CS3 (26). The Escherichia coli strain DH5α was used for plasmid DNA propagation and subcloning (12). The S. cerevisiae PMR1-disrupted mutant (AA274) and its isogenic wild-type strain (AA255) were kindly provided by G. R. Fink (Massachusetts Institute of Technology) (2). Y. lipolytica and S. cerevisiae cultures were maintained on YM medium (0.3% Bacto yeast extract, 0.3% Bacto malt extract, 0.5% Bacto Peptone, 1% dextrose, 2% agar) and cultivated in YPD medium (yeast extract, 10 g/liter; Bacto Peptone, 10 g/liter; glucose, 20 g/liter). The production medium used for recombinant α-amylase and EGI was GPP (10 g of glycerol/liter, 3.4 g of yeast nitrogen base/liter without amino acids and ammonium sulfate, 3.4 g of Proteose Peptone/liter, 50 mg of uracil/liter, 50 mg of adenine/liter in 50 mM sodium phosphate buffer [pH 6.8]).

TABLE 1.

Strains and plasmids used in this study

Strain or plasmid	Relevant genotype or description	Source and/or reference
Strains
E. coli DH5α	supE44 ΔlacU169 (φ80lacZΔM15) hsdR17 recA1 endA1 gyrA96	Hanahan (12)
Y. lipolytica
CX161-1B	ade1 ura3 xpr2	D. M. Ogrydziak
SMS397A	MATa ade1 ura3 xpr2	D. M. Ogrydziak
CS3	MATa Ylpmr1::ADE1 ura3 xpr2	26
SMS-RA	SMS397A harboring pXOS103-In	27
SMS-FC	SMS397A harboring pXCSIn	25
CS3-RA	CS3 harboring pXOS103-In	This work
CS3-FC	CS3 harboring pXCSIn	This work
S. cerevisiae
AA255	MATa ade2 hisΔ200 leu2-3,112 lys2Δ201 ura3-52	Fink (2)
AA274	MATa pmr1-1::LEU2 ade2 hisΔ200 leu2-3,112 lys2Δ201 ura3-52	Fink (2)
AA255-RA	AA255 harboring pSCRA	This work
AA255-FC	AA255 harboring pSCFC	This work
AA274-RA	AA274 harboring pSCRA	This work
AA274-FC	AA274 harboring pSCFC	This work
Plasmids
pOS103	1.5-kb rice α-amylase in XbaI site of pBluescript KS	R. L. Rodriguez (19, 25)
pEndoI	4.2-kb T. reesei EGI genomic DNA at HindIII site of pBR322	M. Ward
pIMR52	XPR2 in pUC19	D. M. Ogrydziak
pIMR53	XPR2 URA3 ars18 in pBR322	D. M. Ogrydziak
pIMR100	XPR2 URA3 in pBR322	D. M. Ogrydziak
pDB20	2μm adh1p adh1t URA3 Ampr	Becker (4)
pXOS103-In	XPR2p::rice α-amylase presequence::rice α-amylase in pIMR100	27
pXCSIn	XPR2p::EGI presequence::EGI in pIMR100	25
pSCRA	adh1p::rice α-amylase presequence::rice α-amylase in pDB20	This work
pSCFC	adh1p::EGI presequence::EGI in DB20	This work

DNA manipulation and transformation.

General recombinant DNA techniques were performed as described in Sambrook et al. (30). E. coli transformation was performed by the SEM method (16). S. cerevisiae and Y. lipolytica transformations were carried out by the lithium acetate method as described by Ito et al. (17) and Gaillardin et al. (9), respectively, with HaeIII-digested E. coli DNA as the carrier DNA.

Construction of expression vectors.

Construction of the vectors pXOS103-In and pXCSIn(myc) for expression in Y. lipolytica of rice α-amylase and fungal EGI, respectively, was done as described by Park et al. (25, 27). The rice α-amylase expression vector for S. cerevisiae, pSCRA, was constructed by inserting the rice α-amylase coding sequence (HindIII fragment of pENO103 [19]) into pDB20, a 2μm vector containing the URA3 gene as a selection marker (4). The endoglucanase expression vector, pSCFC, was constructed by transferring the EGI coding sequence (HindIII fragment) from pXCS to pDB20.

SDS-PAGE and Western blot analysis.

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was conducted as described by Laemmli (20). After separation by SDS-PAGE, proteins were electroblotted onto a nitrocellulose membrane (Schleicher & Schuell, Inc., Keene, N.H.) in ice-cold transferring buffer (15.6 mM Tris, 120 mM glycine, 20% methanol [pH 8.3]) at 100 V for 1 h. The primary antibodies for Western blot analysis were anti-AEP, provided by D. M. Ogrydziak, anti-AXP provided by T. W. Young (University of Birmingham, Birmingham, United Kingdom), anti c-myc antibody (Invitrogen, San Diego, Calif.), and anti-barley α-amylase antibody provided by S. Katoh and M. Terashima (Kyoto University, Japan). The secondary antibodies were anti-mouse immunoglobulin G (IgG) and anti-rabbit IgG antibodies conjugated with peroxidase as supplied in the Phototope-HRP Western Blot Detection Kit (New England Biolabs). The detection procedure was performed according to the manufacturer’s recommendation. The Western blot analyses were performed for AEP from SMS397A and CS3 cells grown in GPP medium at pH 6.8 and 5.0, respectively, AEP from SMS397A and CS3 cells grown in YPD medium supplemented with 10 mM CaCl₂, AXP from SMS397A and CS3 cells grown in GPP medium at pH 6.8 and 5.0, respectively, and rice α-amylase from SMS397A-RA and CS3-RA grown in YPD medium supplemented with 10 mM CaCl₂. The protein samples were prepared by concentrating the culture supernatant and cell extracts 10-fold with trichloroacetic acid precipitation before loading. Ten microliters of the concentrated culture supernatants and cell extracts containing approximately 0.15 μg of total protein was loaded.

Preparation of cell extracts.

After the cells were grown in GPP or YPD medium for 36 h, 5 ml of cell suspension was collected and centrifuged. Cell pellets were dissolved with 0.5 ml of ice-cold protease inhibitor cocktail solution (Boehringer Mannheim Biochemicals, Indianapolis, Ind.) and disrupted by vortexing with 1 g of acid-washed glass beads (diameter, 212 to 300 μm; Sigma Chemical Co., St. Louis, Mo.). Ten microliters of crude extracts were mixed with 2.5 μl of 5× sample loading buffer, boiled for 5 min, and loaded onto an SDS-PAGE gel.

Enzyme activity assays.

The acid protease activity was estimated by measuring the hydrolysis of the standard hemoglobin by the method described by Larson and Whitaker (21). The substrate contained acid-denatured bovine hemoglobin (Sigma Chemical Co.) at a final concentration of 5 g/liter in 0.05 M acetate–0.05 M phosphate–5 × 10⁻⁶ M EDTA buffer adjusted to pH 3.5. Enzyme solution (0.4 ml) and substrate (3.0 ml) were combined and incubated for 1 h at 25°C. After the reaction was stopped with 3.0 ml of 10% trichloroacetic acid, the precipitate was filtered and 1.0 ml of the filtrate was assayed by using the Bio-Rad Protein Assay Kit (Bio-Rad Laboratories). One enzyme unit corresponds to the amount of protease causing an increase in absorbance of 0.1 at 595 nm after 1 h.

The starch-degrading activity of recombinant rice α-amylase was determined by monitoring reducing sugars by the modified dinitrosalicylic acid (DNS) method (27). Enzyme solution (0.5 ml) was added to 0.5 ml of substrate solution (100 mM sodium acetate buffer with 5 mM CaCl₂ and 1% soluble starch [pH 5]). After 10 min of incubation at 30°C, the reaction was terminated by adding 0.5 ml of DNS to 0.5 ml of reaction solution and boiling for 5 min. The solution was then diluted with 4 ml of distilled water, and absorbance was measured at 540 nm.

The carboxymethyl cellulose activity of recombinant EGI was determined by monitoring reducing sugars by the DNS method (3, 24). The reaction was initiated by adding 0.2 ml of enzyme solution to 1.8 ml of substrate (1% carboxymethyl cellulose in 50 mM sodium citrate [pH 4.8]). After incubation for 20 min at 50°C, 3 ml of DNS solution was added to terminate the reaction. Following 5 min of boiling, the absorbance at 540 nm was determined.

For both of these assays, glucose solution was used as a standard. One enzyme unit corresponds to the amount of enzyme required to produce 1 μmol of reducing sugar from the substrate per min.

Protein assays.

The Bio-Rad Protein Assay Kit was used for protein assays with bovine serum albumin as the standard. Absorbance at 595 nm was used to monitor the protein content of culture supernatants and cell extracts.

RESULTS

AEP processing and secretion in a Ylpmr1-disrupted mutant.

To examine the functional role of the YlPMR1 gene product with secretory proteins in Y. lipolytica, the effects of the YlPMR1 disruption on the secretion of endogenous proteins were investigated. SDS-PAGE results showed that the productivity of 32-kDa mature AEP, a major secreted protein in Y. lipolytica, was reduced dramatically in the CS3 strain compared to that in the wild-type SMS397A strain (data not shown).

Western blot analysis with anti-AEP antibody showed that the processing and secretion of AEP in a CS3 strain grown in GPP medium was significantly affected by the YlPMR1 disruption (Fig. 1, cf. the 32-kDa mature AEP bands in lanes 1 and 2 with those in lanes 3 and 4 at pH 6.8; also compare lanes 5 and 6 with lanes 7 and 8 at pH 5.0). The CS3 strain secreted 52- and 36-kDa AEP precursors in addition to the 32-kDa mature AEP (Fig. 1, lane 3), while the SMS397A strain secreted a large amount of only the 32-kDa mature AEP at pH 6.8 (Fig. 1, lane 1). Matoba et al. suggested that the 52- and 36-kDa polypeptides were possible precursors for the 32-kDa mature AEP (22). The 32-kDa mature AEP proteins in the samples taken from both the supernatant and cell extract of the CS3 strain almost completely disappeared at pH 5.0 (Fig. 1, lanes 7 and 8), while significant amounts of secreted and intracellular mature AEP were detected in the SMS397A strain at pH 5.0 (Fig. 1, lanes 5 and 6). The intracellular 52-kDa AEP precursor was found in the sample taken from the cell extract of the CS3 strain cultivated at pH 6.8 (Fig. 1, lane 4), while it was not detected in the cell extract sample from the wild-type SMS397A strain (Fig. 1, lane 2).

FIG. 1.

Western blot analysis of AEP from SMS397A (YlPMR1) and CS3 (Ylpmr1) strains grown in GPP medium at pH 6.8 and 5.0, respectively. Lanes 1 and 5, supernatants of SMS397A culture grown at pH 6.8 and 5.0, respectively; lanes 2 and 6, cell extracts of SMS397A culture grown at pH 6.8 and 5.0, respectively; lanes 3 and 7, supernatants of CS3 culture grown at pH 6.8 and 5.0, respectively; lanes 4 and 8, cell extracts of CS3 culture grown at pH 6.8 and 5.0, respectively. mAEP, mature AEP; pAEP, AEP precursors. The test for secreted proteins is shown in lanes 1, 3, 5, and 7, and the test for the intracellular proteins is shown in lanes 2, 4, 6, and 8.

A very significant effect of pH on the secretion of the 32-kDa AEP and on the intracellular processing of the 36-kDa precursor AEP in the CS3 strain was also found. The amount of the 32-kDa mature AEP secreted by SMS397A decreased significantly when the pH was lowered from 6.8 to 5.0 (Fig. 1, cf. lanes 5 and 1), while the amount of the intracellular 32-kDa mature AEP increased (Fig. 1, cf. lanes 6 and 2). The secreted 32-kDa mature AEP and the 36- and 52-kDa precursor AEPs from the CS3 strain disappeared when the pH was lowered from 6.8 to 5.0 (Fig. 1, cf. lanes 7 and 3). Almost all of the intracellular 32-kDa mature and 36-kDa precursor AEP disappeared when the pH was lowered from 6.8 to 5.0 (Fig. 1, cf. lanes 8 and 4).

The secretion of the 32-kDa mature AEP in the CS3 strain was restored when the YPD medium was supplemented with 10 mM CaCl₂ (Fig. 2). Both the extracellular and intracellular 32-kDa mature AEPs were not detected in the CS3 strain at a low-calcium concentration compared to the detection of these proteases in the SMS397A strain (Fig. 2, cf. lanes 3 and 4 with lanes 1 and 2). When 10 mM CaCl₂ was added to the YPD medium, secretion of the 32-kDa mature AEP and the 36-kDa premature AEP was restored in the CS3 strain (Fig. 2, cf. lanes 7 and 3), while the intracellular 32-kDa mature AEP and the 36-kDa premature AEP were not detectable (Fig. 2, cf. lanes 8 and 4) regardless of the concentration of calcium added to the medium. The amount of the 32-kDa intracellular mature AEP significantly decreased in the SMS397A strain at the higher calcium concentration (Fig. 2, cf. lanes 6 and 2).

FIG. 2.

Western blot analysis of AEP from SMS397A (YlPMR1) and CS3 (Ylpmr1) strains grown in YPD medium supplemented with 10 mM CaCl₂ at pH 6.8. The Western blot experiment was performed with anti-AEP antibody. Lanes 1 and 5, supernatants of SMS397A culture grown in low-calcium (∼180 μM) and high-calcium (10 mM) concentration media, respectively; lanes 2 and 6, cell extracts of SMS397A culture grown in low- and high-calcium concentration media, respectively; lanes 3 and 7, supernatants of CS3 culture grown in low- and high-calcium concentration media, respectively; lanes 4 and 8, cell extracts of CS3 culture grown in low- and high-calcium concentration media, respectively.

These results suggest that the YlPMR1 disruption in Y. lipolytica significantly alters the pH and Ca²⁺ effects on the processing and secretion patterns of mature and premature AEP proteins.

Activation of secreted AXP.

The effects of the YlPMR1 disruption in the CS3 strain on the secretion and processing of AXP were also studied. The AXP activity (92.0 U/ml) of the CS3 strain was about 60% higher compared to that of the SMS397A strain (57.3 U/ml) grown in GPP medium adjusted to an initial pH of 5.0 (Table 2). During the cultivation of the CS3 strain, the pH level remained constant at 5.0 while that of the SMS397A culture increased from 5.0 to 5.7. Glover et al. suggested that the expression of the AXP gene is regulated by the pH, and the highest level of AXP mRNA was detected at a pH range of 5.0 to 5.5 (10). It is likely that the lower pH in the CS3 strain slightly enhanced the expression of the AXP gene and this resulted in the higher enzyme activity in the CS3 strain (92.0 U/ml) compared to that of the SMS397A strain (57.3 U/ml). Western blot analysis revealed that both the SMS397A and CS3 strains produced and secreted the same 39-kDa AXP band (Fig. 3). As Glover et al. pointed out that secreted AXP proenzyme undergoes activation extracellularly by autocatalytic cleavage at acidic pH, the processing of AXP is independent of Ca²⁺ regulation in the Golgi apparatus (10). These results suggest that the YlPMR1 disruption affects the activity of the secreted AXP at the pH where its expression is induced, whereas the secretion process of premature AXP protein is not affected by the YlPMR1 disruption.

TABLE 2.

Effect of YlPMR1 disruption on the activity of AXP in Y. lipolytica SMS397A (YlPMR1) and CS3 (Ylpmr1) strains

Parameter	Values at initial pH of:
	6.8		5.0
	YlPMR1	Ylpmr1	YlPMR1	Ylpmr1
AXP activity (U/ml)	1.6	3.9	57.3	92.0
Cell density (OD600)a	15.0	15.1	14.9	15.6
Protein amt (mg/ml)	3.1	2.8	1.2	1.4
Final pH	6.4	6.2	5.7	5.0

^aCultures were grown in GPP medium for 36 h without pH control and worked up as described in Materials and Methods. 

FIG. 3.

Western blot analysis of AXP from SMS397A (YlPMR1) and CS3 (Ylpmr1) strains grown in GPP medium at an initial pH of 5.0. Lane 1, supernatant of SMS397A; lane 2, cell extract of SMS397A; lane 3, supernatant of CS3; lane 4, cell extract of CS3.

Inhibition of secretion of heterologous rice α-amylase.

The expression and efficient secretion of rice α-amylase in Y. lipolytica was previously reported (27). The expression vector (pXOS103-In) was transformed into SMS397A and CS3 strains to study the effects of the YlPMR1 disruption on the secretion of heterologous proteins. Integration of the expression vector into chromosomal DNA was confirmed by Southern blot analysis (27). The transformants of both strains were grown on YM-starch plates, and secretion of rice α-amylase was detected by using iodine vapor. In liquid culture, relatively high rice α-amylase activity (2.75 U/ml) was detected in the transformed SMS397A strain (SMS-RA), while very low enzyme activity (0.46 U/ml) was observed in the transformed CS3 strain (CS3-RA) during all stages of growth (Fig. 4b). Although the SMS397A-RA transformant showed significant α-amylase activity, the CS3-RA transformant showed no clear zone around the colony, indicating the absence of α-amylase activity on the YM-starch plate (Fig. 4a). The effect of the YlPMR1 disruption on rice α-amylase secretion was further confirmed by Western blot analysis. Western blot analysis revealed that neither extracellular nor intracellular rice α-amylase was detected in the CS3-RA strain cultured in YPD medium (Fig. 4c, cf. lanes 1 and 2 with lanes 3 and 4). When 10 mM CaCl₂ was added to the YPD medium, secretion of rice α-amylase was partially restored in the CS3-RA strain (Fig. 4c, cf. lanes 7 and 3), while intracellular rice α-amylase was not detected (Fig. 4c, cf. lanes 8 and 4) regardless of the concentration of calcium added to the medium. These results suggest that secretion of rice α-amylase is almost completely blocked by the YlPMR1 disruption.

FIG. 4.

Effects of YlPMR1 disruption on the secretion of rice α-amylase in Y. lipolytica. (a) Test of α-amylase activity on a YM-starch plate. 1, SMS-RA; 2, CS3-RA. (b) Secretion of recombinant rice α-amylase in a flask culture (GPP medium). ■, SMS-RA (growth); •, CS3-RA (growth); □, SMS-RA (activity); ○, CS3-RA (activity). (c) Western blot analysis of rice α-amylase from SMS397A-RA (YlPMR1) and CS3-RA (Ylpmr1) strains grown in YPD medium supplemented with 10 mM CaCl₂. The Western blot experiment was performed with anti-barley α-amylase antibody. Lanes 1 and 5, supernatants of SMS397A-RA culture grown in low-calcium (∼180 μM) and high-calcium (10 mM) concentration media, respectively; lanes 2 and 6, cell extracts of SMS397A-RA culture grown in low- and high-calcium concentration media, respectively; lanes 3 and 7, supernatants of CS3-RA culture grown in low- and high-calcium concentration media, respectively; lanes 4 and 8, cell extracts of CS3-RA culture grown in low- and high-calcium concentration media, respectively.

Since the secretion response of the Ca²⁺-ATPase-deleted S. cerevisiae pmr1 mutant (29) was found to be different from that of the Y. lipolytica Ylpmr1 mutant, the rice α-amylase gene was expressed in YlPMR1-disrupted S. cerevisiae (AA274) and its isogenic wild type (AA255) to examine whether the effect of secretory pathway Ca²⁺-ATPase disruption on heterologous protein secretion is species specific or protein specific. The rice α-amylase signal sequence and coding sequence were inserted into the yeast expression vector, pDB20, between the alcohol dehydrogenase promoter and terminator sequences. The resulting vector (pSCRA) was transformed into S. cerevisiae AA255 and AA274 separately, and secretion of rice α-amylase was determined in liquid and solid plate cultures. As shown in Fig. 5a, the AA255 transformant (AA255-RA) exhibited rice α-amylase activity, as indicated by the halo around the colony, while the AA274 transformant (AA274-RA) showed no enzyme activity. These results demonstrate that disruption of the YlPMR1 gene also created an inhibitory effect on secretion of heterologous rice α-amylase in S. cerevisiae. The secreted T. reesei EGI in S. cerevisiae AA255 and AA274 gave positive responses on an ostrazin brilliant red H-3B-conjugated hydroxyethyl cellulose (OBR-HEC)-containing YM plate (Fig. 5b). Thus, we can conclude that improvements in heterologous protein secretion by the disruption of the secretory pathway Ca²⁺-ATPase are not applicable to all proteins in Y. lipolytica and S. cerevisiae strains.

FIG. 5.

Effects of PMR1 disruption in S. cerevisiae on the secretion of rice α-amylase and T. reesei EGI. (a) α-Amylase activity on a YM-starch plate. 1, AA255-RA; 2, AA274-RA. (b) EGI activity on an OBR-HEC-containing YM plate (25). 1, AA255-FC; 2, AA274-FC.

Change in outer-chain glycosylation of heterologous fungal EGI.

The EGI expression vector pXCSIn(myc) was transformed in both the SMS397A and CS3 strains, and the secretion of recombinant EGI was detected on solid plate and liquid cultures (Fig. 6). Although the plate assay showed similar halos around the colonies of the SMS397A and CS3 transformants (SMS-FC and CS3-FC, respectively) (Fig. 6a), the liquid cultures showed that SMS-FC secreted more EGI than CS3-FC (Fig. 6b). However, if we consider the cell concentrations of both strains and the specific activities of the secreted proteins (196 U/mg of protein for SMS-FC and 176 U/mg of protein for CS3-FC), the amounts of EGI produced by both strains were almost the same. This result is also consistent with that shown in Fig. 5b.

FIG. 6.

Effects of YlPMR1 disruption on the secretion of T. reesei EGI in Y. lipolytica. (a) EGI activity on an OBR-HEC-containing YM plate. 1, SMS-FC; 2, CS3-FC. (b) Secretion of recombinant EGI in a flask culture (GPP medium). ■, SMS-FC (growth); •, CS3-FC (growth); □, SMS-FC (activity); ○, CS3-FC (activity). (c) Western blot analysis for recombinant EGI. Lane 1, culture supernatant of SMS397A (host cell); lane 2, culture supernatant of SMS-FC; lane 3, culture supernatant of CS3-FC; and lanes 4 and 5, culture supernatants of SMS-FC and CS3-FC, respectively, after endo H treatment.

The secreted EGI was examined by Western blot analysis with an anti-myc antibody (Fig. 6c). While the recombinant EGI produced by the SMS-FC strain was hyperglycosylated, the YlPMR1-disrupted strain displayed a homogeneous single band without any smear corresponding to hyperglycosylation. To determine the amounts of N-linked glycosylation in the secreted proteins, endo-β-N-acetylglucosaminidase H (endo H) treatment was performed. The endo H-treated EGI secreted by CS3-FC displayed a homogeneous single band around 55 kDa, which is smaller than the untreated protein (58 kDa) (Fig. 6c, cf. lanes 4 and 5 with lanes 2 and 3). Endo H cleaves high mannose and some hybrid oligosaccharides from the N-linked glycosyl group of glycoproteins. Although the recombinant EGI secreted from the YlPMR1-disrupted strain is not hyperglycosylated compared to the control strain (SMS-FC), it contains N-linked glycosyl residues (possibly core glycosylation) which can be removed by endo H.

DISCUSSION

This study clearly shows that the secretion patterns of the YlPMR1-disrupted mutant (CS3) and the isogenic wild-type strain (SMS397A) are different in Y. lipolytica. Secretion of homologous AEP is dramatically decreased and premature precursors that have not been processed completely (52- and 36-kDa proteins) were secreted by the YlPMR1-disrupted mutant (Fig. 1). Enderlin and Ogrydziak reported that the Xpr6p (Kex2p-like endoprotease), which is the primary protease required for Lys-Arg cleavage of AEP, is Ca²⁺ dependent and that mutation of XPR6 causes formation of the 52-kDa precursor (8). Based on these data, we can assume that disruption of the YlPMR1 gene causes the Ca²⁺ depletion in the Golgi apparatus and thereby affects the function of the Ca²⁺-dependent Xpr6p and results in the secretion of premature precursors of AEP. When 10 mM calcium was added to the YPD medium, the secretion of AEP was partially restored (Fig. 2). However, Western blot analysis with the AEP antibody in the YlPMR1-disrupted mutant showed two bands (36 and 32 kDa) while only a 32-kDa band was observed for the control strain. It has been suggested that the 36-kDa polypeptide is a possible precursor for the mature AEP (22).

It has been reported that the SMS397A strain displays a dimorphic characteristic (a mixture of mycelia and oval-shaped cells) while the CS3 strain shows clustered strings of bead-like cells in a GPP medium (26). The morphology of the CS3 strain is similar to that of the XPR6-defective mutant. This is additional evidence for our suggestion that the YlPMR1 gene disruption affects the function of Xpr6p in the CS3 mutant strain.

The findings that the AXP activity of the CS3 strain was significantly increased (60%) and the AEP secretion by SMS397A was inhibited at pH 5.0 are closely related to the perturbation of calcium concentration and culture pH. During cultivation, the CS3 strain remained constant at pH 5.0, but the SMS397A culture pH increased from 5.0 to 5.7 (Table 2). Disruption of the YlPMR1 gene caused the difference in the culture pH patterns of these two strains.

While rice α-amylase is efficiently secreted by the SMS-RA strain, very low levels of rice α-amylase secretion were observed for the CS3-RA strain (Fig. 4b). An inhibitory effect of a Ca²⁺-ATPase deletion in the S. cerevisiae PMR1-disrupted strain on the secretion of rice α-amylase was also observed (Fig. 5a). Our results show that recombinant rice α-amylase neither secretes into the culture medium nor accumulates in the YlPMR1-disrupted mutant cells and that an addition of exogeneous Ca²⁺ can partially recover the secretion of rice α-amylase (Fig. 4c). Jones et al. determined the effect of calcium on the secretion of α-amylase and other hydrolases from aleurone layers of barley (18). They observed that withdrawal of Ca²⁺ from the incubation medium of aleurone layers resulted in a 70 to 80% reduction of α-amylase secretion. Also, Bush et al. showed that barley α-amylase is irreversibly inactivated by the removal of Ca²⁺ from the protein and that Ca²⁺ stabilizes the tertiary structure of the enzyme (6). They have also shown that millimolar levels of calcium are necessary to stabilize barley α-amylase in the ER of the aleurone layer. Since rice α-amylase is highly homologous to cereal α-amylases including barley α-amylase, it is speculated that production of recombinant rice α-amylase in CS3-RA cells is almost completely inhibited by the intracellular Ca²⁺ deficiency which results from the YlPMR1 disruptions in Y. lipolytica.

Unlike rice α-amylase, the secretion of T. reesei EGI was not influenced by the YlPMR1 disruption. However, the secreted EGI from the YlPMR1-disrupted mutant did have different characteristics than that of the control (Fig. 6). While wild-type cells secreted the hyperglycosylated form of EGI, hyperglycosylation was completely absent from the mutant strain. This result confirms that disruption of the secretory pathway Ca²⁺-ATPase causes disrupted outer-chain glycosylation of secreted proteins in yeast (29). However, endo H treatment of the secreted EGI shows that there are some glycosyl groups at the N-glycosylation site, indicating only a partially defective outer-chain glycosylation in the YlPMR1-disrupted mutant.

In this study, the effects of disrupted secretory pathway Ca²⁺-ATPase in the yeast Y. lipolytica on the secretion and posttranslational processing of homologous and heterologous proteins were examined by using AEP, AXP, rice α-amylase, and fungal EGI as model proteins. In the YlPMR1-disrupted mutant, secretion of mature AEP decreased while the premature 36- and 52-kDa AEPs were secreted and intracellular 52-kDa premature AEP was also detected. Production of rice α-amylase was completely blocked in the YlPMR1-disrupted mutant. Production and secretion of recombinant fungal EGI was not affected, while hyperglycosylation of EGI was removed. Production of AXP was increased by the YlPMR1 disruption. Our results indicate that the effects of the YlPMR1 disruption depend on the characteristics of the target protein in the recombinant yeast strain evaluated.