Degradation of Rhizopus niveus aspartic proteinase-I with mutated prosequences occurs in the endoplasmic reticulum of Saccharomyces cerevisiae.

Rhizopus niveus aspartic proteinase-I (RNAP-I) is secreted by Saccharomyces cerevisiae extracellularly (Horiuchi, H., Ashikari, T., Amachi, T., Yoshizumi, H., Takagi, M., and Yano, K. (1990) Agric. Biol. Chem. 54, 1771-1779). The prosequence of RNAP-I has the function to promote correct folding of its mature part. Deletion (Deltapro) and amino acid substitutions (M1) in the prosequence block secretion of RNAP-I (Fukuda, R., Horiuchi, H., Ohta, A., and Takagi, M. (1994) J. Biol. Chem. 269, 9556-9561). In this study, little accumulation of Deltapro was observed in Western blot analysis of the cell extracts of the transformants producing Deltapro using anti-RNAP-I antisera. In contrast, M1 was accumulated in the yeast cells. Pulse-chase analysis revealed that they were synthesized at almost the same rates and that Deltapro was degraded in the cells more rapidly than M1. In subcellular fractionation analysis, Deltapro was found in the fraction that contained most of the activity of an endoplasmic reticulum (ER) marker enzyme, NADPH-cytochrome c reductase. In indirect immunofluorescence microscopy, Deltapro was observed in the ER. Similar result was also observed in a mutant which is deficient of the two vacuolar proteases, proteinase A and proteinase B. So, the vacuolar proteases are not involved in degradation of Deltapro. From these results, we concluded that RNAP-Is with the mutated prosequences, which probably could not be folded correctly, were retained and degraded in the ER.

In eukaryotic cells, most of nascent secretory proteins are targeted to the endoplasmic reticulum (ER) 1 after their signal sequences are recognized. Then these proteins are translocated across the ER membrane co-or post-translationally. Many secretory proteins have their final conformations in the ER lumen, then are transported to the Golgi compartment and secreted from cells. Some proteins that cannot be folded correctly or assembled into proper multimers are retained in the ER. Molecular chaperones such as immunoglobulin heavy chain binding protein, BiP, bind to those immature or abnor-mal proteins and keep them retained in the ER to stimulate their proper folding or multimerization (for review, see Refs. [3][4][5][6]. Recently, it has been revealed that protein degradation machinery is present in the ER and some secretory or membrane proteins that could not be folded or multimerized correctly are degraded in the ER (for review, see Ref. 7). But little is known about the mechanisms of recognition and degradation of malfolded proteins by these protein degradation machinery in the ER. Three kinds of mutated proteins are shown to be degraded in the ER of Saccharomyces cerevisiae; these are a secretiondefective mutant of human ␣-1-protease inhibitor (8), mutants of carboxypeptidase Y, carrying an Arg instead of a Gly in a highly conserved region, and proteinase A, in which 37 amino acids spanning the processing site of the prosequence are deleted (9). In addition, it is considered that the unassembled 100-kDa integral membrane subunit of the yeast vacuolar H ϩ -ATPase complex is degraded in the ER of the cell lacking Vma21p, which is suggested to be required for assembly of the integral membrane sector of V-ATPase in the ER (10).
Rhizopus niveus, a filamentous fungus, secretes large amounts of aspartic proteinases extracellularly. We have cloned and sequenced genes encoding aspartic proteinase-I (11) and aspartic proteinases II-V (RNAP-I to RNAP-V) (12). RNAP-I is synthesized as a precursor form with a pre-sequence (21 amino acid residues) and a prosequence (45 amino acid residues) at the N terminus of the mature part (323 amino acid residues). RNAP-I is also secreted extracellularly by S. cerevisiae efficiently when the encoding gene is introduced and expressed (1). The prosequence of RNAP-I is essential for both renaturation of the denatured mature part in vitro and secretion of the mature part in vivo in S. cerevisiae. Some RNAP-Is with the mutated prosequences are not secreted extracellularly at all (2). Therefore, the prosequence of RNAP-I is suggested to be essential for correct folding of its mature part in vivo.
In this paper, we examined the transport, accumulation, and degradation of those RNAP-Is with the mutated prosequences in S. cerevisiae. We found that these precursor RNAP-Is were retained and degraded in the ER. Therefore, it is concluded that the precursors that cannot be folded correctly in the ER lumen are specifically recognized and degraded by the degradation machinery in the ER lumen.
S. cerevisiae strain R27-7C-1C (MATa his3 leu2 ura3 trp1) was used as a host for production of RNAP-I with the wild-type or the mutated prosequences. Strain #334 (MAT␣ leu2 ura3 reg1 gal1 pep4 prb1) were used as hosts for production of ⌬pro. Yeast cells were cultivated aero-* This study was supported by a grant-in-aid for scientific research from the Ministry of Education, Science and Culture of Japan. This work was performed using the facilities of the Biotechnology Research Center, University of Tokyo. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Enzymes and Reagents-Nucleic acid modification enzymes were purchased from Takara Shuzo Co. (Kyoto, Japan) and used under conditions recommended by the manufacturer. An oligonucleotide for site-directed mutagenesis was synthesized with a model 391 DNA synthesizer (Applied Biosystems).
Plasmid Construction-Recombinant DNA manipulations were done by the standard methods (13). Yeast transformation was carried out using the lithium acetate procedure described by Ito et al. (14).
Plasmid pYPR28412, which encodes an RNAP-I derivative ⌬pro* in which Asp 100 of ⌬pro was replaced with Ala to inactivate the protease active site, was constructed as follows. A 0.4-kb EcoRI-SalI fragment from pYPR2841 (2) was cloned between EcoRI and SalI sites of pUC119 (resulting in pUC2841), and Asp 100 was replaced with Ala by sitedirected mutagenesis with a 26-mer synthetic oligonucleotide, 5Ј-AAACTTGATTTTGCAACTGGTTCTTC-3Ј. A 0.4-kb EcoRI-SalI fragment from the resultant double-stranded DNA was ligated with a 9.4-kb EcoRI-SalI fragment of pYPR2841 to form the plasmid pYPR28412.
Plasmid pYPR2841U, which was used to express ⌬pro in pep4 prb1 cells, was constructed as follows. A 2.2-kb HindIII fragment of pYPR2841, which contained glyceraldehyde-3-phosphate dehydrogenase gene promoter, ⌬pro, and glyceraldehyde-3-phosphate dehydrogenase gene terminator, was transferred into PvuII site of YEp24.
Accurate construction of all plasmids was confirmed by nucleotide sequencing.
Nucleotide sequencing was performed on the Applied Biosystems model 373A DNA sequencing system.
Preparation of Total Cell Extracts-Yeast cells were collected in a tube, washed with 10 mM sodium azide, and suspended in Laemmli sample buffer (15). One volume of glass beads was added to the tube, and then it was vortexed for 20 s and put on ice for 1 min. This step was repeated five times, and the preparation was boiled for 5 min. After centrifugation at 15,000 rpm for 5 min at 4°C, the supernatant was used as a total cell extract.
Western Blot Analysis-SDS-PAGE was done by the method of Laemmli (15). Detection of filter-bound antibodies was carried out according to the enhanced chemiluminescence method with ECL detection reagents (Amersham Corp.).
Pulse-chase Analysis-Transformants of S. cerevisiae were cultured in YNBDCU medium (YNBD medium containing 2% casamino acids and 50 g/ml uracil) for 12 h. Cells in a 1.1-ml culture were collected and inoculated in 1.1 ml of YNBDHLU medium (YNBD medium containing 20 g/ml histidine, 30 g/ml leucine, and 50 g/ml uracil) supplemented with the amino acid mixture (20 g/ml arginine, 30 g/ml tyrosine, 30 g/ml isoleucine, 30 g/ml lysine, 50 g/ml phenylalanine, 100 g/ml glutamic acid, 100 g/ml aspartic acid, 150 g/ml valine, 200 g/ml threonine, 400 g/ml serine). After 20 min at 30°C, cells were labeled with 0.1 mCi of [ 35 S]methionine for 5 min and an aliquot (200 l) was transferred to a chilled tube containing 2 l of 100 mM azide. Then the remaining cells were collected, washed, and incubated in 900 l of YNBDHLU medium supplemented with the amino acid mixture and 2 mg/ml methionine. An aliquot was removed into ice-cold azide at intervals.
Immunoprecipitation was performed by the method described by Franzusoff et al. (16) with some modifications. Total cell extract (40 l) was prepared as described by Yaffe and Schatz (17). Four volumes (160 l) of IP dilution buffer (1.25% Triton X-100, 190 mM NaCl, 6 mM EDTA, 60 mM Tris-HCl (pH 7.4)) were added to the cell extract in a tube. After the addition of 4 l of 1:10 diluted preimmune sera, the extract was incubated on ice for 1 h. Then 40 l of 10% IgG Sorb (The Enzyme Center, Inc., Malden, MA) were added, and the extract was incubated for 30 min on ice. After centrifugation at 10,000 ϫ g for 5 min, the supernatant was taken into a tube and 1:10 diluted anti-RNAP-I antisera was added, after which it was incubated for 1 h on ice. Then 40 l of 10% IgG Sorb were added, and the extract was incubated for 30 min on ice. After centrifugation, the pellet in the tube was washed four times with IP buffer (1% Triton X-100, 0.2% SDS, 150 mM NaCl, 50 mM Tris-HCl (pH 7.4)). After 40 l of Laemmli sample buffer was added, it was boiled for 5 min and the mixture was subjected to SDS-PAGE.
Subcellular Fractionation Analysis-Subcellular fractionation analysis was performed by the method of Nakano et al. (18) with some modifications.
S. cerevisiae spheroplasts were prepared by the method of Franzusoff et al. (16). Spheroplasts were disrupted with Potter-Elvehjem homogenizer (15 strokes). The extracts were centrifuged at 1,000 ϫ g for 10 min at 4°C to prepare the pellet (low speed pellet, LSP) and the supernatant (low speed supernatant). The low speed supernatant fraction was further centrifuged at 100,000 ϫ g for 1 h at 4°C to obtain the high speed pellet ( Indirect Immunofluorescence Microscopy-Indirect immunofluorescence microscopy was performed by the method described by Preuss et al. (22) with some modifications.
Transformants were grown in YNBDCU medium for 12 h and fixed by adding formaldehyde to the final concentration of 4% and incubating for 1 h at room temperature.
Staining of Vacuole-Vacuolar components were visualized with lucifer yellow CH by the methods of Wada et al. (26) with some modifications. Samples of 1 ml of exponentially growing cells in YNBD medium containing 2% Casamino acids were harvested, resuspended in YEPD containing 10 mg/ml of lucifer yellow CH, and incubated for 6 h with agitation. Labeled cells were diluted with 1 ml of ice-cold PBS and washed with the same buffer five times. The vacuoles were observed with a fluorescence microscope.

RNAP-Is with the Mutated Prosequences Were Degraded in
the Cells-The prosequence of RNAP-I is essential for extracellular secretion of the mature part in S. cerevisiae and RNAP-Is with the mutated prosequences, ⌬pro and M1, are not secreted at all (2). In ⌬pro, the whole prosequence (from Ser 22 to Ala 66 ) are deleted, and in M1, Ala 49 -Leu 50 is replaced by Asp-Pro (numbering is from the initiation codon Met).
To examine whether these RNAP-Is with the mutated prosequences were synthesized in the cells, the transformants harboring the plasmids encoding these RNAP-Is with the mutated prosequences were cultured for 14 h at 30°C, and the total cell extracts were prepared and subjected to Western blot analysis using anti-RNAP-I antisera. Both of the mutated as well as the wild-type RNAP-I precursors were detected in the cell extracts, but not in equal amounts. Little accumulation of ⌬pro was observed in the extracts of the cells (Fig. 1, lane 3), and the amount of M1 accumulated in the cells was more than that of ⌬pro (Fig. 1, lanes 3 and 4). To test the possibility of autocatalytic degradation of ⌬pro in the cells, the intracellular accumulation of ⌬pro*, in which one of the active sites, Asp 100 , of ⌬pro was changed to Ala to inactivate this enzyme was examined. The amount of ⌬pro* in the cells was less than that of the wild-type precursors and was almost equal to that of  (Fukuda et al., 1994), or pYPR28412 (⌬pro*, lane 5) was subjected to Western blot analysis using anti-RNAP-I antisera. The arrowheads indicates the bands corresponding to RNAP-I with the wild-type prosequence, ⌬pro, M1, or ⌬pro*. ⌬pro (Fig. 1, lanes 2, 3, and 5). Therefore, it was suggested that lower level of accumulation of ⌬pro was not due to autocatalytic degradation in the cells.
To investigate why ⌬pro was not accumulated so much as M1 in the cells, the pulse-chase analysis was done. Transformants producing ⌬pro or M1 were pulse-labeled for 5 min with [ 35 S]methionine and chased for 0, 30, or 60 min. ⌬pro and M1 were synthesized at almost the same levels (Fig. 2, lanes 1 and  4). However, after 30 min of chase, most of ⌬pro disappeared from the cell extract (Fig. 2, lane 2). The amount of M1 after 30 min of chase also decreased, but not so much as that of ⌬pro did (Fig. 2, lane 5). ⌬pro and M1 were not secreted extracellularly at all (2). These results indicated that ⌬pro and M1 were degraded in the cells and that the lower level of accumulation of ⌬pro in the cells was due to the faster rate of the intracellular degradation.
⌬pro and M1 Were Degraded in the ER-To determine the intracellular site of accumulation of ⌬pro and M1, subcellular fractionation analysis was performed. Spheroplasts of the transformants producing ⌬pro or M1 were homogenized, and the cell extracts were subjected to differential centrifugation at 1,000 ϫ g and 100,000 ϫ g. Three fractions, LSP, HSP, and HSS, were analyzed by Western blotting with anti-RNAP-I antisera. Both ⌬pro and M1 were recovered exclusively in the LSP fractions (Table I). Most of the activity of an ER marker enzyme, NADPH-cytochrome c reductase, was also collected in the LSP fraction. The activity of a Golgi marker enzyme, heatstable dipeptidylaminopeptidase, was recovered in both the LSP and the HSP fractions. The activity of a vacuolar enzyme, carboxypeptidase Y, was recovered in both the LSP and HSS fractions. Most of the activity of cytosolic enzyme, ␣-glucosidase, was collected in the HSS fraction. Therefore both ⌬pro and M1 were obtained in the fractions that included most of the ER marker enzyme activity.
To further define the localization of ⌬pro, indirect immunofluorescence microscopy with anti-RNAP-I antisera was carried out. The nuclei in these cells were simultaneously stained with 4Ј,6-diamidino-2-phenylindole. The transformants producing ⌬pro showed prominent staining of the region surrounding the nuclear DNA (Fig. 3A), while no staining was observed when non-related antibodies (anti-␤-galactosidase antibodies) were applied to those cells (Fig. 3C). The transformants producing M1 showed the same staining pattern as those producing ⌬pro (data not shown). As a marker for the ER, the staining of protein disulfide isomerase (PDI) was also performed with anti-PDI antisera. The staining pattern of PDI was similar to that of ⌬pro (Fig. 3B).
In eukaryotic cells, vacuole (lysosome in mammalian cells) contains various proteases and many proteins are degraded in it (23,24). If ⌬pro is degraded in vacuole, ⌬pro would be accumulated in vacuole of the vacuolar protease-deficient cells.
To address this possibility, the localization of ⌬pro was investigated in a mutant that is deficient of the major vacuolar proteases, proteinase A and proteinase B, by indirect immunofluorescence microscopical observation. In pep4 prb1 cells, prominent staining of the perinuclear region was observed as in the wild-type cells (Fig. 4A). The staining pattern of PDI was similar to that of ⌬pro (Fig. 4B). The staining pattern of ⌬pro was different from the pattern of vacuole visualized with lucifer yellow CH (Fig. 4D). Therefore, it was suggested that ⌬pro was degraded not in vacuole, but in the ER.  1 and 4), 30 (lanes 2 and 4), and 60 min (lanes 3 and 6). Cell extracts were prepared and immunoprecipitated using anti-RNAP-I antisera. The immune pellets were resolved by SDS-PAGE. The arrowheads indicate the bands corresponding to ⌬pro or M1.  In this paper, we analyzed the intracellular transport, accumulation, and degradation of RNAP-Is with the mutated prosequences, ⌬pro and M1, that were not secreted from S. cerevisiae. ⌬pro and M1 were synthesized at almost the same rate, but ⌬pro was degraded more rapidly than M1 in the cells. Therefore, the amount of M1 accumulated in the cells was more than that of ⌬pro. The subcellular fractionation analysis showed that both ⌬pro and M1 were recovered in the fractions that contained most of the activity of the ER marker enzyme, NADPH-cytochrome c reductase. In addition, indirect immunofluorescence microscopy with anti-RNAP-I antisera demonstrated that ⌬pro localized in the ER. The staining pattern of ⌬pro did not change even in pep4 prb1 cells. Thus, it was presumed that vacuolar proteases did not take part in the degradation of ⌬pro. Consequently, we concluded that these RNAP-Is with the mutated prosequences were retained and degraded in the ER.
It was revealed that denatured RNAP-I could renature in the presence of the wild-type prosequence in vitro, but not in the absence of the prosequence or in the presence of that of M1 (2). Therefore, ⌬pro and M1 seem to be unable to have the correct tertiary structure in the ER lumen of S. cerevisiae. In this case, they would be recognized and degraded specifically by the degradation machinery in the ER. Our results presented in this paper support this hypothesis. However, it is not clear why ⌬pro is degraded more rapidly than M1. ⌬pro is deficient of the whole prosequence, while M1 has the only prosequence with 2 amino acids substituted. So, it is speculated that ⌬pro cannot have correctly folded structure in the ER lumen, whereas M1 may have partially folded structure. Therefore, ⌬pro may be more easily recognized or degraded by the degradation machinery in the ER than M1.
Based on the results described in this paper, we propose a model for the transport, accumulation, and degradation of RNAP-I precursors in the ER lumen of S. cerevisiae. The wildtype precursor has the correct folding structure by the function of the prosequence and is secreted extracellularly. In contrast, ⌬pro and M1 cannot be folded correctly because of the mutations in their prosequences and are degraded in the ER by an unknown degradation mechanism.
We are interested in the mechanism via which unfolded or malfolded proteins such as ⌬pro or M1 are recognized by the degradation machinery and in proteases that participate in the degradation of those proteins. We have isolated several mutants of S. cerevisiae that have deficiencies in the degradation of ⌬pro and accumulate ⌬pro in the cells, probably in the ER. 2 It is expected that mutants defective in the recognition and/or degradation machinery in the ER would be present among them. Characterization of these mutants will be helpful to elucidate the mechanisms of recognition and degradation of proteins in the ER.