Vacuolar Targeting and Posttranslational Processing of the Precursor to the Sweet Potato Tuberous Root Storage Protein in Heterologous Plant Cells*

the tuberous root storage protein of the sweet potato, which is localized in vacuoles, is synthesized as a prepro-precursor with an N-terminal se- quence of amino acids that includes a signal peptide and an additional pro-segment of 16 amino acids. A full-length cDNA for sporamin was placed downstream of the 35 S promoter of cauliflower mosaic virus and introduced into tobacco and sunflower genomes by Ti plasmid-mediated transformation. A polypeptide of nearly the same size as mature sporamin from the sweet potato was detected in transformed calli of to- bacco and sunflower, as well as in the leaves, stems, and roots of regenerated, transgenic tobacco plants. Amino acid sequence analysis of the nearly mature- sized form of sporamin from the transformed tobacco cells revealed that it is actually longer by three amino acids at its N terminus than authentic sporamin puri- fied from the sweet potato. By pulse labeling of suspen-sion-cultured cells with

acids at its N terminus than the nearly mature-sized form of sporamin. These results suggest that at least two steps of posttranslational processing of the pro-form occurs sequentially in tobacco cells. The posttranslational processing of the pro-form of the precursor to sporamin was inhibited by monensin, suggesting that this step takes place in the acidic compartment, probably in the vacuole. All of the sporamin polypeptides synthesized in transformed tobacco cells were retained inside the cell and sporamin was localized in the vacuole, as judged from results of subcellular fractionation.
These results indicate that sporamin is appropriately targeted to the vacuole in tobacco cells.
The correct sorting and targeting of newly synthesized proteins to separate organelles is essential for the generation * This work was supported in part by a Grant-in-Aid for Special Project Research and a Grant-in-Aid for Cooperative Research (to K. N.) from the Ministry of Education, Science and Culture, Japan. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "aduertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
?I To whom correspondence should be addressed. and maintenance of the organization and function of eukaryotic cells. Recently, signals required for targeting proteins to specific organelles have been identified for many proteins, such as those found in mitochondria (l), chloroplasts (2), endoplasmic reticulum (3), lysosomes (4), and nuclei (5). In many cases, these targeting signals are encoded in the Nterminal amino acid sequence of the precursor, which is removed during maturation of the protein by proteolytic processing. In some cases, at least two signals are required for the correct targeting of a protein to specific organelles and to a specific intraorganellar space. Some nuclear-encoded chloroplast proteins require two signals in the transit peptide segment of the precursor, one for transport into the chloroplast and the other for translocation to the appropriate location inside of the organelle (2). Vacuolar hydrolases of the yeast Saccharomyces cereuisiae also require two signals for their correct targeting (6). In addition to the N-terminal signal peptide required for entry into the lumen of the endoplasmic reticulum, a second signal is required for directing this protein into the vacuole. The latter signal is located in the propeptide region of the precursor and is cleaved off after entry into the vacuole (6). However, the mechanism of signal recognition for the translocation of proteins and the proteases responsible for the maturation of precursor polypeptides have not yet been well characterized in most cases. It is obvious, however, that the presence of such mechanisms is a prerequisite for the functional expression of genes that encode organellar proteins.
In many plant cells, the vacuole is the largest organelle and it has diverse functions, such as the intracellular digestion of materials, the accumulation and storage of organic and inorganic nutrients and metabolites, and the generation of turgor, with specific function depending on the type of cell and on the stage of development of the cell (7). Vaculoes in certain cells that have differentiated to become storage cells may, themselves, differentiate into protein storage organelles or protein bodies by accumulating large amounts of storage proteins. Plant vacuolar matrix proteins are synthesized by membrane-bound polysomes as precursors with N-terminal signal peptides, and many of them are also subject to posttranslational processing prior to maturation (8). However, the exact nature of their vacuole-targeting signals and the cellular mechanisms required for their targeting and maturation have not yet been determined.
Systems for transgenic gene expression have proved useful for the analysis of the mechanisms of protein targeting in various systems. Several reports have previously described the expression of precursors to plant vacuolar proteins in heterologous systems. Bean phytohemagglutinin, a vacuolar storage protein in seeds, has been expressed in yeast cells and targeted to the vacuole (9), suggesting that some common mechanisms may be operating in the targeting of proteins to the vacuole in these two organisms. The precursor to phytohemagglutinin was also expressed in transgenic tobacco (lo), and the correct glycosylation, Golgi processing, and targeting to the vacuole were observed. Furthermore, it was shown that glycan side chains are not required for the correct targeting of phytohemagglutinin to the vacuole in transgenic tobacco seeds (11). Expression of other plant vacuolar proteins, pconglycinin from soybean seeds (12), lectin from barley seeds (13), 2 S albumin from Arubidopsis seeds (14), and patatin from potato tubers (15), in heterologous plant tissues has also been examined.
Sporamin is the most abundant protein in the tuberous root of the sweet potato, accounting for about 80% of the total soluble protein (16), and it is localized in the vacuole (17). It is encoded by a nuclear multigene family of more than 10 different genes and consists of a mixture of closely related polypeptides with molecular weights of approximately 20 kDa (16,18). Unlike many other vacuolar storage proteins in plants, sporamin is not glycosylated (16). It is synthesized by membrane-bound polysomes as a larger precursor (19) with an extra N-terminal sequence of 35 or 37 amino acids, the number depending on the specific gene (18,20). The structure of the N-terminal region of the precursor to sporamin can be divided into two segments: the N-terminal, hydiophobic signal peptide and an additional segment of 16 amino acids which contains an unusually high proportion of charged amino acids (18,20). We showed previously that only the Nterminal signal peptide segment of the precursor is removed by cotranslational processing in uitro with microsomal membranes (21), a result that strongly suggests the posttranslational removal of the additional segment of 16 amino acids. Here we report the processing and targeting to vacuoles of a precursor to sporamin expressed in transformed tobacco and sunflower cells. both of these forms are obtained under these conditions (21). Edman Degradation Analysis-Suspension-cultured transformed tobacco cells (0.5 ml) were labeled with 200 &i ofL-[2,5-3H]histidine for 1 h with a subsequent chase with 1 mM L-histidine for 1 h. Sporamin-related polypeptides were immunoprecipitated from the cell lysate with sporamin-specific antiserum and protein A-Sepharose, and then they were eluted by boiling in a mixture of 2% SDS, 1% pmercaptoethanol, and 20 mM ammonium carbonate (pH 9.0). Polypeptides in the supernatant were concentrated by ultrafiltration, mixed with bovine serum albumin, and subjected to sequential Edman degradation as described previously (21).
Preparation of the Vacuole Fraction-4-day-old, suspension-cultured, sporamin-expressing cells were harvested by centrifugation and washed once with 0.6 M mannitol. The packed cells were resuspended in 4 volumes of a mixture of 2% Cellulase Y-C, 0.5% Macerozyme R-10, 0.1% Pectolyase Y-23, 0.6 M mannitol (pH 5.2 with HCl), and incubated at 32 "C for 2 h, with occasional gentle pipeting, to convert most of the cells to protoplasts. Protoplasts were harvested by centrifugation and washed twice with 0.6 M mannitol.
15 ml of the suspension of protoplasts (lo6 protoplasts/ml) were centrifuged at 1500 X g for 10 min through a gradient that was composed, from bottom to top, of 5 ml of 10% Ficoll, 0.1% dextran sulfate, 7.5 ml of 5% Ficoll, 0.1% dextran sulfate, and 20 ml of 2.5% Ficoll, 0.6% DEAE-dextran. All of these Ficoll-dextran solutions were prepared in 10 mM MES, Tris (pH 6.9) and 0.6 M mannitol. Lysed cells at the interface between layers of 5 and 10% Ficoll were collected and resuspended in 5 volumes of 5% Ficoll, 0.1% dextran sulfate. 10 ml of 5% Ficoll, 0.1% dextran sulfate, and 5 ml of 0.1% dextran sulfate were overlaid on the suspension, and the entire preparation was centrifuged at 1500 X g for 10 min. Vacuoles and some surviving protoplasts that floated to the interface between 5 and 0% Ficoll were resuspended in 10% Ficoll to a volume of 12 ml, overlaid with 2 ml of 2.5% Ficoll, 0.1% dextran sulfate, and 1 ml of 0.1% dextran sulfate and centrifuged at 1500 X g for 10 min. The material at the interface between layers of 2.5 and 0% Ficoll, enriched with vacuoles, was collected and used as the vacuole fraction.
Enzymatic and Chemical Assays-Activities of phosphodiesterase and glucan synthase I were assayed by the methods of Boller and Kende (31) and Green (32), respectively. Other enzymatic activities and concentrations of protein were assayed as described previously (17).

Expression
of the Precursor to Sporamin in Transformed Tobacco and Sunflower Callus Tissues-A full length sporamin cDNA in pIM023 (19) encodes a precursor to sporamin of 230 amino acids. Processing of this precursor in vitro with dog pancreatic microsomal membranes (21) removes only the N-terminal signal peptide segment to yield a pro-form of the precursor which still retains an extra 16 amino acids at the N terminus of the mature form (Fig. lA). A polypeptide with the same electrophoretic mobility as the pro-form of the precursor is also observed when the precursor encoded by pIM023 cDNA is expressed in E. coli (21), or in S. cereuisiae pep4 mutant which lack vacuolar proteinase A (30)' (Fig. ZA, lane I). In order to examine the expression of the precursor to sporamin in heterologous plant cells, the cDNA insert of pIM023 was fused to the downstream end of the 35 S promoter from CaMV. We used a derivative of pIM023 in which the 5' poly(dG) tail and part of the 5'-noncoding region has been removed from the cDNA insert, for the construction of 35 S promoter-sporamin fusion gene, since the presence of the 5' poly(dG) tail inhibited the expression of the precursor to sporamin in yeast cells.' The 35 S promoter-sporamin fusion gene was integrated into the T-DNA region of the T: plasmid pTiB6S3trac, via an intermediate plasmid, to yield pTiSA16F (Fig. lC), as described under "Experimental Procedures." A. tumefaciens harboring pTiSA16F was used to generate crown gall-type transformed calli of tobacco and sunflower.
The tobacco and sunflower calli transformed with Agrobacterium that harbored pTiSA16F contained polypeptides that cross-reacted with sporamin-specific antiserum ( Fig. 2A, .&es  3 and 4). No immunoreactive polypeptides were detected in calli transformed with Agrobacterium that harbored pTi-BGS3tra" alone ( Fig. 2A, lunes 5 and 6). Although the level of expression of sporamin-related polypeptides varied considerably among individual transformants (data not shown), a polypeptide band of almost identical size was detected. The size of the material in the immunoreactive band was smaller than the size of the pro-form of the precursor to sporamin (Fig. 2A, lane 1), but it was slightly larger than the average size of a mixture of sporamin isoproteins isolated from the sweet potato ( Fig. 2A, lane 2 A, the primary structure of the N-terminal part of the precursor to sporamin encoded by pIM023 (19). Hydrophobic and charged amino acid residues are indicated by circles and hexagons, respectively. Thick underlining and dotted underlining indicate the mature part and the pro-segment, respectively. B, the structure of the T-DNA region of Ref. 24) and the critical part of pTiSA16F (lower). TL and TR, two T-DNA regions of pTiB6-806; filled arrowheads; the positions of 25-base pair border sequences. 35Sp, CaMV 35 S promoter; SPO, sporamin cDNA, derived from pIM023; NTP-II, neomycin phosphotransferase gene. C, the structure of the binary plasmid pVSAD. The EcoRI-Hind111 fragment carrying the 35 S promoter-sporamin fusion gene from pCSAD was inserted into the EcoRI-Hind111 sites of pGA469 (27). no@ and no.?, nos (nopaline synthase)-promoter and nos-terminator, respectively.

Sporamin
migrates anomalously on SDS-polyacrylamide gels during electrophoresis, with its mobility depending on the concentration of acrylamide (19). Sporamin isolated from the sweet potato migrates as a broad band, reflecting the microheterogeneity of the preparation, which has a peak size at an apparent molecular mass of 23.8 kDa under the conditions used in the present study, which is larger than the value of 19.95 kDa determined by analysis of a Ferguson plot or deduced from the nucleotide sequences (19,20). The apparent molecular masses of the pro-form and the protein in the immunoreactive band from tobacco cells were 26.8 and 24.0 kDa, respectively.
These results strongly suggest that the precursor to sporamin that is expressed in tobacco cells has been subjected to some kind of posttranslational processing in addition to the cotranslational removal of the signal peptide. Since the apparent molecular mass of 24.0 kDa of the protein in the immunoreactive band from tobacco cells was slightly larger than that of the purified sporamin, 23.8 kDa, the sporamin-related immunoreactive band detected in extracts of tobacco cells is henceforth referred to as the nearly mature-sized sporamin. Expression of the Precursor to Sporamin in Various Organs of Transgenic Tobacco Plants-The 35 S promoter-sporamin fusion gene was introduced into the binary vector pGA469 (27) to yield pVSAD (Fig. lc), and A. tumefaciens harboring both pTiBGS3tra" and pVSAD was used to obtain regenerated, transformed tobacco plants. A polypeptide with the same electrophoretic mobility as that of polypeptides present in transformed calli was detected as the major sporamin-related polypeptide in extracts from the leaves, stems, and roots of axenically grown, transgenic tobacco plants (Fig. 2B) indicating that the precursor to sporamin is processed in these organs in a similar manner to that in callus tissue.
N-terminal Amino Acid Sequence of the Nearly Maturesized Form of Sporamin in Transformed Tobacco Cells-In order to examine the localization and the processing in vivo of the precursor to sporamin that is expressed in tobacco cells, we first developed a suspension culture of tobacco cells that expressed the precursor using the transformed calli. Sporamin-related polypeptides could be detected only in the cell fraction, and they were not detected in the medium fraction of the cultures by immunblotting of SDS gels (data not shown).
The cells in the suspension culture were collected and the nearly mature-sized form of sporamin was purified from the soluble fraction by immunoaffinity column chromatography and SDS-polyacrylamide gel electrophoresis. Immunoaffinity column was prepared by covalent linking of IgG fraction of sporamin-specific antiserum to Affi-Gel. In addition to the nearly mature-sized form of the sporamin, several other proteins were eluted from the immunoaffinity column. Proteins in the eluted fraction were separated by SDS-polyacrylamide gel electrophoresis and transferred to Immobilon membrane. After staining, the band corresponding to the nearly maturesized form of sporamin was excised and its N-terminal amino acid sequence was determined.
The sequence, X-P-A-S-X-E-T-P-V-L-D-I-N-G-D, shown with the one-letter amino acid codes with X representing those which could not be unambiguously identified, corresponds to the amino acid sequence of the precursor to sporamin from positions GluZ5 to Asp"' (19; see Fig. IA). These results indicate that the nearly mature-sized form of sporamin in tobacco cells is actually longer by three amino acids than authentic sporamin purified from the sweet potato.
Labeling in Vivo of the Precursor to Sporamin in Transformed Tobacco Cells-When transformed tobacco cells in suspension culture were labeled with ["Slmethionine for 6 h and "'S-labeled cellular polypeptides were subjected to immunoprecipitation with sporamin-specific antiserum, a polypeptide with the same electrophoretic mobility as the nearly mature-sized sporamin was precipitated (data not shown). By contrast, when the cells were labeled with [""Slmethionine for 1 h, two labeled polypeptides were precipitated (Fig. 3A, lane 2). The mobility of the more slowly migrating band was identical to the mobility of the pro-form of the precursor to sporamin which was obtained by processing in vitro of the precursor to sporamin with dog pancreatic microsomal membranes (Fig. 3A, lane 1, lower band), and the mobility of the more rapidly migrating band was similar to that of the band detected by immunoblotting.
After a 2-h chase, the upper band disappeared and only the more rapidly migrating band was observed (Fig. 3A, lane 3). The time-course of the conversion of the pro-form to the more rapidly migrating band during the chase (Fig. 3B) indicated that an increase in the amount of material in the more rapidly migrating bands occurs concomitant with a decrease in the amount of the proform. The half time of this processing, as estimated from densitometric tracings of autoradiographic bands, was 21 min. In contrast to the appearance of the pro-form after pulse labeling, no precursor to sporamin with the signal peptide still attached was detected under these conditions in vivo. These results indicate that the signal peptide is very rapidly removed from the precursor and that the precursor is further processed posttranslationaly at a relatively slow rate, compared to the rate of cotranslational processing.

Posttranslational
Processing of the Pro-form of the Precursor to Sporamin-The more rapidly migrating band which appeared after the chase of the 35S-labeled pro-form was some-  In order to examine further the nature of the posttranslational processing of the precursor to sporamin in tobacco cells, we labeled cells with ["Hlhistidine for 1 h and chased with unlabeled histidine for 1 h. Under these conditions, the "H-labeled pro-form of the precursor was almost completely processed to the intermediate form, and the nearly maturesized form represented only a minor proportion of the labeled protein (data not shown; see Fig. 3B, lane 4). The 3H-labeled, sporamin-related polypeptides were immunoprecipitated and analyzed by the manual, sequential Edman degradation. A peak of radioactivity appeared in the second cycle (Fig. 4), indicating that the N-terminal amino acid of the 24.3-kDa intermediate form is Thr"" (see Fig. L4). That conversion of the 24.3-kDa band to the 24.0-kDa nearly mature-sized form occurs by removal of the Thr33-His34 dipeptide was supported by Edman degradation analysis of a mixture of the 24.3-kDa band and the 24.0-kDa band that had been obtained by pulse labeling with ["Hlproline for 1 h and subsequent chase with unlabeled proline for 2 h. In this case, peaks of radioactivity appeared at the second and fourth cycles (data not shown).

Effect of Monensin on Posttranslational
Processing of the Precursor to Sporamin-We examined the effects of monensin on the posttranslational processing of the precursor to sporamin in tobacco cells. Monensin has been reported to inhibit posttranslational processing of precursors to several vacuole-localized seed proteins (33, 34), and it is suggested that monensin causes alkalinization of the vacuolar matrix, which in turn inhibits the vacuole-localized processing enzyme(s) (33, 35). As shown in Fig. 5, the posttranslational processing of the pro-form of the precursor to sporamin to the nearly mature-sized form was almost completely inhibited by 5 pM monensin.

Vacuolar Localization of Sporamin Expressed in Trans-
formed Tobacco Cells-Sporamin-related polypeptides could not be detected in the culture medium of the transformed tobacco cells not only by immunoblotting of an SDS gel, but also by immunoprecipitation after long term labeling with [35S]methionine (data not shown). Moreover, no sporaminrelated polypeptides were detected in the culture medium after 30 min of pulse labeling and after a subsequent 2-h chase (Fig. 6). These results suggest that all of the sporamin-related polypeptides synthesized in tobacco cells are retained inside of the cell and none of them are secreted to the medium. In order to examine the intracellular site of localization of sporamin in tobacco cells, we prepared the vacuole fraction from protoplasts of tobacco cells grown in suspension culture, as described under "Experimental Procedures." The isolated vacuole fraction was found to be contaminated by less than 0.1% protoplasts when examined under the light microscope, and it was enriched with round vesicles which could be stained with Neutral Red. The average size of vacuolar vesicles was similar to that of protoplasts (a diameter of about 20 pm) suggesting that most of the vacuoles in this fraction were of the central vacuole type. Specific activities of vacuolar marker enzymes, namely, phosphodiesterase and a-mannosidase, were about ll-and 17-fold greater, respectively, in the vacuole fraction than in the protoplasts (Table I). By contrast, the specific activities of marker enzymes for other organelles (NADPH-cytochrome c reductase for the endoplasmic reticulum, glucan synthase I for Golgi, cytochrome c oxidase for mitochondria, catalase for microbodies, glucose-6-phosphate dehydrogenase for plastids and cytoplasm) in the vacuole fraction were lower than those in protoplasts ( Table I). The relative amount of the nearly mature-sized sporamin in the vacuole fraction was analyzed by immunoblotting of an SDS-polyacrylamide gel (Fig. 7). The amount of sporamin detected in a total of 5 pg of protein from the vacuole fraction was similar to that in 75 pg of protein from the protoplasts (Fig. 7, lanes 1 and 4). This difference indicates that sporamin was concentrated to about 15-fold during the purification of the vacuole fraction. This ratio of the concentration of sporamin in the vacuolar fraction to that in the protoplast was similar to those for the vacuolar marker enzymes described above. These results strongly suggest that the intracellular site of localization of sporamin in transformed tobacco cells is the interior of the vacuole, as is the case in the tuberous root of the sweet potato (17).

Structural
analyses (19,20) and processing in vitro by microsomal membranes (21) of the precursor to sporamin suggest that, in addition to cotranslational removal of a signal peptide, posttranslational removal of a pro-segment of 16 amino acids is required for the maturation of sporamin in the sweet potato (Fig. 1A). The subcellular site and the nature of the posttranslational processing of the precursor in the sweet potato are not known. In this paper, we described our analysis of the expression of a single precursor polypeptide for spora- A suspension culture of tobacco cells that expressed the precursor to sporamin was labeled with ["'Slmethionine for 30 min (chase 0 min) and chased for 120 min (chase 120 min). Polypeptides in the culture medium (M) and the cell extract (C) from 0.4 ml of the culture were immunoprecipitated with sporamin-specific antiserum and fractionated by SDS-polyacrylamide gel electrophoresis. pp, precursor; p, pro-form; m, mature sporamin. min in heterologous plant cells under control of the 35 S promoter from CaMV. This approach is particularly appropriate for studies of sporamin, especially for the analysis of posttranslational processing at the molecular level, since sporamin consists of a mixture of various isoproteins encoded by a multigene family (16,18). Comparison of amino acid sequences of six precursors for sporamin, as deduced from the nucleotide sequences, show sequence homologies of 77-98% between them, and the total number of genes for sporamin in the sweet potato genome is estimated to be about 60 (18). In order to avoid complexities due to the microheterogeneity at the sequence level, it is desirable to analyze the behavior of a single polypeptide precursor to sporamin. However, to date, transformation of the sweet potato has not been achieved. In crown gall-type calli of tobacco and sunflower and in the leaves, stems, and roots of axenically cultured, transgenic tobacco plants the precursor to sporamin was processed to a nearly mature-sized polypeptide (Fig. 2). Amino acid sequence analysis of this nearly mature-sized form of sporamin in tobacco cells indicated that it is actually longer by three amino acids at its N terminus than the authentic sporamin purified from the sweet potato. The occurrence of the posttranslational processing of the precursor to sporamin in tobacco cell is clearly indicated by the pulse-chase labeling experiments using the suspension-cultured cells. The signal peptide is very rapidly removed from the precursor and the resulting proform of the precursor is further processed in two sequential steps (Fig. 3). It is not known whether two sequential posttranslational processings also occurs in the sweet potato or not. However, the difference in the N-terminal structures suggests that the protease that is involved in the posttranslational processing of the precursor to sporamin in tobacco cells is different from that functioning in the tuberous root of the sweet potato. It will be interesting to determine whether this difference is due to the difference in plant species or to the presence of a particular protease in the tuberous root in which the expression of sporamin genes is specifically activated.
Only limited information is available at present concerning the proteases involved in the posttranslational processing of plant vacuolar proteins. In the case of yeast vacuolar proteins, vacuolar proteases catalyze the cleavage of the pro-segment from pro-proteins after their arrival in the vacuole. Thus, in S. cerevisiae pep4, a mutant deficient in vacuolar proteinase A (30), precursor forms of several vacuolar proteins accumulate in the vacuole (36). Many storage proteins and lectins in seeds, which accumulate in vacuoles or protein bodies, also undergo posttranslational proteolytic cleavage (8). Endoproteolytic activities for the processing of precursors to ricin and agglutinin in castor bean endosperm (37), lectin in rice embryo (33), and 11 S globulin in pumpkin cotyledon (35) have acidic pH optima and, in these cases, posttranslational processing most likely occurs at the sites of accumulation of the mature proteins. Hara-Nishimura and Nishimura (35) characterized a vacuolar thiol protease from pumpkin cotyledons which catalyzes the posttranslational processing of pro-globulin and pro-trypsin inhibitor. The enzyme seems to catalyze the processing of a wide variety of precursors to seed vacuolar protein, and a similar enzymatic activity is present in seeds of other plant species. Since the amino acid sequence of the pIM023encoded precursor to sporamin around the potential processing site does not contain the Asn-X sequence which is the site of posttranslational cleavage in 11 S globulin and other precursors to seed vacuolar protein (35, 3%40), the enzyme that processes the precursor to sporamin seems to be different from these seed enzymes. It seems likely that the posttranslational processing of the precursor to sporamin takes place in the vacuole, or at a late stage of its transport to the vacuole. When the precursor to sporamin is expressed in yeast, the pep4 mutation causes the appearance of the pro-form of the precursor (Fig. 24, lane l).* Furthermore, posttranslational processing of the pro-form of the precursor in tobacco cells is inhibited by monensin (Fig.  5). Monensin has been reported to inhibit posttranslational processing of the precursor to lectin in rice embryos (33) and to 11 S globulin in pumpkin cotyledons (34). In the case of pumpkin cotyledons, monensin does not inhibit the transport of proglobulin to the vacuole (34), and it has been suggested that monensin causes alkalinization of the vacuolar matrix, which in turn inhibits the activity of the processing enzyme (33, 35). In addition, posttranslational processing is a much slower process, with a half time of 21 min, than the very rapid cotranslational processing of the precursor to sporamin in transformed tobacco cells (Fig. 3). This time difference may be required for the transport of the pro-form by endomembrane transport systems from the lumen of the endoplasmic reticulum to the site of processing.
We did not detect any sporamin-related polypeptides in the culture medium of the suspension-cultured tobacco cells (Fig.  6) and sporamin polypeptides were localized in the vacuole fraction (Fig. 7). These results suggest that the precursor to sporamin can be appropriately transported to the vacuole in the heterologous tobacco cells. We have not examined the subcellular localization of sporamin-related polypeptides in various organs of transgenic tobacco plants. However, if the posttranslational processing of the precursor takes place in the vacuole, the fact that the sizes of sporamins detected in the leaves, stems, and roots of transgenic tobacco plants were almost identical to that of the sporamin in suspension-cultured, transformed tobacco cells (Fig. 2) suggests that precur-sors to sporamin in these organs are also appropriately targeted to the vacuole.
The targeting of proteins to lysosomes in animal cells and to vacuoles in yeast cells requires, in addition to a signal peptide, a second positive signal, namely, mannose 6-phosphate groups in the former case (4) and the polypeptide of the pro-segment in the latter case (6, [41][42][43]. Sporamin is a simple protein which does not contain glycans (16). If the targeting of sporamin to vacuoles in plant cells also requires a second positive signal, it is most likely to be contained in the structure of the pro-form of the precursor. Glycan side chains on the precursors to several plant vacuolar proteins are not required for the correct targeting to the vacuole (11,13,44). Heterologous expression of the precursor to sporamin in transformed tobacco cells, as described in this paper, should provide a useful experimental system with which to study mechanisms for targeting of proteins to plant vacuoles, and should help in identifying the vacuolar targeting signal in the protein. Our recent studies indicate that the pro-segment of the precursor to sporamin is required for correct targeting of sporamin to the vacuole."