The Aspartic Proteinase of Barley 1s a Vacuolar Enzyme That … · Plant Physiol. (1 994) 105: 321 -329 The Aspartic Proteinase of Barley 1s a Vacuolar Enzyme That Processes Probarley

Plant Physiol. (1 994) 105: 321 -329

The Aspartic Proteinase of Barley 1s a Vacuolar Enzyme That Processes Probarley Lectin in Vitro'

Pia Runeberg-Roos*, Jukka Kervinen, Valentina Kovaleva, Natasha V. Raikhel, and Susannah Cal

lnstitute of Biotechnology, P.O. Box 45, Karvaamokuja 3 A, FIN-00014, University of Helsinki, Finland (P.R.-R., J.K.); and Michigan State University-Department of Energy Plant Research Laboratory,

Michigan State University, East Lansing, Michigan 48824-1 31 2 (V.K., N.V.R., S.G.)

Previous work suggested that the aspartic proteinase from Hor- deum vulgare (HvAP) would be a vacuolar protein in plant cells. Based on N-terminal sequencing we show that the in vitro-translated protein was translocated into the lumen of microsomal membranes, causing a concomitant removal of 25 amino acid residues from the protein. Vacuoles were purified from barley leaf protoplasts and were shown to contain all of the aspartic proteinase activity found in the protoplasts. This vacuolar localization of HvAP was confirmed with immunocytochemical electron microscopy using antibodies to HvAP in both barley leaf and root cells. In an attempt to discern a function for this protease, we investigated the ability of HvAP to process the C-terminal proregion of barley lectin (BL) in vitro. Prolectin (proBL), expressed in bacteria, was processed rapidly when HvAP was added. Using severa1 means, we were able to determine that 13 amino acid residues at the C terminus of proBL were cleaved off, whereas the N terminus stayed intact during this incubation. lmmunohistochemical electron microscopy showed that HvAP and BL are co-localized in the root cells of developing embryos and germinating seedlings. Thus, we propose that the vacuolar HvAP participates in processing the C terminus of BL.

Proteolysis plays an important role in many biological processes, including protein degradation and tumover, protein processing, and pathogen attack. Proteinases have been classified as to the amino acid residues or cofactors involved in the proteinase activity, and the classification is usually based on the sensitivity of the activity to various specific inhibitors. The main classes include Ser proteinases, Cys proteinases, metalloproteinases, and Asp proteinases. There are examples of each of these types of proteinases in plants.

Asp proteinases (EC 3.4.23) are a class of endopeptidases that are active at acidic pH, inhibited by pepstatin A, and show a conserved three-dimensional structure with two Asp residues in the active cleft (Tang and Wong, 1987; Davies, 1990). This group of proteinases is generally in the secretory pathway and is activated either outside the cell or inside the lysosome/vacuole (Tang and Wong, 1987; Davies, 1990). In

The research was supported by grants from the Finnish Academy of Science to P.R-R., the Finnish Ministry of Agriculture and Forestry to P.R-R. and J.K., and the National Science Foundation, Washington, DC (grant No. MCB-9002652) and the United States Department of Energy, Washington, DC (grant No. DE-FG02-91ER20021) to N.V.R.

* Corresponding author; fax 358-0-434-6046. 321

animals, fungi, and viruses, Asp proteinases form an inter- esting group, often with specific physiological functions. Proteinases belonging to this group include the mammalian digestive enzymes pepsin and chymosin (Foltmann, 198 l), renin, which is involved in the regulation of blood pressure (Campbell, 1987), and cathepsin D and yeast proteinase A, which are involved in the processing of protein precursors (Ammerer et al., 1986; Woolford et al., 1986; Gottschalk et al., 1989; Nishimura et al., 1989). Retroviral proteinases involved in polyprotein processing also belong to this class (Davies, 1990).

Plant Asp proteinases have been purified from monocoty- ledonous (Doi et al., 1980; Belozersky et al., 1989; Sarkkinen et al., 1992) and dicotyledonous (Polanowski et al., 1985; Heimgartner et al., 1990; Rodrigo et al., 1991) species, as well as from gymnosperms (Bourgeois and Malek, 1991). The hydrolytic specificities of some plant Asp proteinases have been determined using synthetic substrates (Faro et al., 1992; Kervinen et al., 1993), but little is known about the biological function of these enzymes. In the insectivorous plant Ne- penthes, Asp proteinase activity has been found in the digestive fluid of the pitchers (Tokés et al., 1974). In tomato and tobacco, an Asp proteinase has been suggested to take part in the hydrolysis of extracellular pathogenesis-related proteins (Rodrigo et al., 1991). Asp proteinases have also been reported to be associated with the intracellular storage protein bodies of hemp (St. Angelo and Ory, 1970) and buckwheat seeds (Elpidina et al., 1990) and to take part in the hydrolysis of storage proteins in wheat (Belozersky et al., 1989; Du- naevsky et al., 1989) and cocoa (Biehl et al., 1993). In addition, Asp proteinases from seeds have been shown to process the Arabidopsis 2s albumin proteins in vitro (D'Hondt et al., 1993).

We have been characterizing the Asp proteinase HvAP from barley (Hordeum uulgare) grains. During purification of the aspartic proteolytic activity, two isoforms of the proteinase were found, one containing peptides of 32 and 16 kD, and the other containing peptides of 29 and 11 kD (Sarkkinen et al., 1992). Subsequent cloning (Runeberg-Roos et al., 1991) and comparison with the sequences of peptides derived from the various proteins purified (Sarkkinen et al., 1992) suggest that the two isoforms are derived from a single gene and

Abbreviations: BL, barley lectin; CTPP, C-terminal processed peptide from BL; HvAP, Hordeum uulgare aspartic proteinase; proBL, pro-form of BL.

https://plantphysiol.orgDownloaded on April 22, 2021. - Published by Copyright (c) 2020 American Society of Plant Biologists. All rights reserved.

https://plantphysiol.org

322 Runeberg-Roos et al. Plant Physiol. Vol. 105, 1994

represent differential processing of the protein precursor in barley grains. The protein sequence (Runeberg-Roos et al., 1991) and amino acid bond specificity (Kervinen et al., 1993) of the proteinase suggest that the most closely related enzymes are lysosomal cathepsin D and the yeast vacuolar proteinase A. Therefore, we wanted to determine if the HvAP would have celIular properties similar to these enzymes, namely localization in the secretory system of the cells and a role in protein processing. Here we show, by biochemical fractionation and immunocytochemistry, that HvAP is located in the vacuoles of barley leaf and root cells, and, in an attempt to determine the function of this vacuolar endopro- teinase, we show that it can process a secretory protein, proBL, in vitro.

MATERIALS AND METHODS

In Vitro Translation and Translocation of HvAP

mRNA encoding the entire HvAP message (Runeberg-Roos et al., 1991) was transcribed from the Gem3Zf vector under the T7 promoter and translated with rabbit reticulocyte lysate with or without canine pancreatic microsomes (Promega) in the presence of [35S]Met. For proteinase protection assays of the translation products, 10-pL samples were transferred to an ice-water bath, and 7 pg of proteinase K was added with or without 1% Triton X-100. The reaction was incubated on ice for 30 min, stopped by the addition of PMSF to 10 mM, and then immediately boiled in Laemmli sample buffer. Translation products were analyzed on SDS-PAGE. The cleavage site between the ER signal sequence and the proenzyme was determined by N-terminal radiosequencing of a [3H]Leu-labeled sample translated in the presence of microsomes. For visualization on SDS-PAGE, about 10,000 cpm of a [35S]Met-labeled sample was mixed with the [3H]Leu- labeled sample. The proenzyme was separated from the preproenzyme by SDS-PAGE and visualized by autoradiog- raphy. The proenzyme was electroeluted for radiosequencing of the N terminus. About 30,000 cpm of the [3H]Leu-Iabeled sample was analyzed by Edman degradation in a gas-pulsed liquid sequencer (Kalkkinen and Tilgmann, 1988). The radioactivity released in each cycle was measured by liquid scin- tillation counting. The N-terminal radosequencing was performed by Dr. N. Kalkkinen (Institute of Biotechnology, University of Helsinki, Finland).

Preparation of Protoplasts and Vacuoles

Barley (Hordeum vulgare L. cv Kustaa) grains were surface sterilized with 1% sodium hypochlorite. The grains were germinated aseptically on 0.75% agar for 6 d at 22 k 2OC with 16 h of light and 8 h of dark. Mesophyll protoplasts were prepared from the fírst leaves by a modification of the method described by Kaiser et al. (1982). The leaves were stripped of their abaxial epidermis and floated on the surface of a medium containing 10 mM Mes/KOH (pH 5.6), 500 mM sorbitol, and 1 mM CaC12. The leaves were vacuum infiltrated for 1 min and the medium was changed to the same medium containing 3.0% (w/v) cellulase onozuka R-10 (Yakult Hon- sha Co. Ltd, Tokyo, Japan) and 0.05% (w/v) Pectolyase (Sigma). After incubating for 2 h at room temperature, the

protoplasts were collected by centrifugation (1OOg for 10 min), purified according to Kaiser et al. (1982), and counted with a l-temocytometer. One-fourth of the protoplasts were changed to a medium containing 10 mM Mes/KOH (pH 5.6), 1 mM CiIClz, and 0.2 mM DTT, and frozen. For isolation of vacuoles, the remaining protoplasts were changed to a cold medium containing 0.45 M sorbitol, 30 mM sodium phosphate (pH 7.5), and 18% Percoll. Vacuoles were liberated from the protoplasts by pressing the suspension through a 22-gauge needle. The released vacuoles were purified as described by Martinoia et al. (1985). The purity of the vacuole fraction was assessed using phase-contrast microscopy and specific enzymic assays (see below).

Enzyme Assays and lnhibition Studies

For enzymic assays, the protoplasts were lysed osmotically followedl by a freeze-thaw cycle, whereas vacuoles were lysed by freeze-thaw followed by vortexing. Acid proteinase activity was measured as described by Sarkkinen e{: al. (1992) both in protoplast and vacuolar samples using bovine he- moglobiin as a substrate at pH 3.7. One unit of activity corresponded to the enzyme activity that 1ibe:rates TCA- soluble reaction products equivalent to 1 mg of BSA in 1 h at 3OoC (ISarkkinen et al., 1992). To inhibit the Asp proteolytic activity, pepstatin A (Sigma) was added to diluted enzyme solutions at a concentration of 0.01 mM and the mixtures were preincubated on ice for 10 min before assaying proteinase activity in a 4-fold larger volume. Activities for organellar enzymes were measured as.described (a-mannosidase, Glc- P isomerase, Cyt c oxidase by Boller and Kende [1979] and refs. therein; NADP-glyceraldehyde phosphate dehydrogen- ase by Lseegood and Walker [1983]). From four iridependent purifications, the average yield of vacuoles per isolation was 25%, basied on the total a-mannosidase activity in protoplasts (160.7 n:m min-', SE 4 24.7) and isolated vacuole fractions (40.3 nm min-', SE 3.0). Assuming that only one vacuole can be isolated per protoplast (Boudet et al., 1981; Kaiser et al., 1982), the number of vacuoles was routinely calculated by dividing the total mannosidase activity in the vacuolar fraction by the measured mannosidase activity per number of protoplasts.

Producti'on of Antibodies

Purified HvAP (Sarkkinen et al., 1992) containing both isoforms (36 pg) in PBS was mixed with Freund's complete adjuvant and injected into popliteal lymph nodes of a rabbit. Two boosts of identical samples were given 2 arid 6 weeks later and the rabbit was bled 1 week after the second boost. IgG fractions were separated on a protein A-Sepharose C1-4B column (Pharmacia) and antibodies reaciing to the carbohyclrate part of HvAP were removed by afj'inity chro- matography with bromelain and fetuin (Sigma), two highly glycosyla.ted proteins, coupled to cyanogen bromide-activated Sepharose. Monoclonal antibodies to the C-terminal amino acids of proBL were prepared using a peptide corresponding to the last 15 amino acids of proBL coupled to purified protein derivative of tuberculin (M.R. Schroeder and N.V. Ra:ikhel, unpublished results). Rabbit antibodies to



A Vacuolar Barley Aspartic Proteinase 323

wheat germ agglutinin, a homolog of BL, have been describedpreviously (Raikhel et al., 1984).

EM

All procedures were carried out at room temperature unlessotherwise indicated. Small pieces of barley root tissue werefixed in a mixture of 2% paraformaldehyde and 0.8% glutar-aldehyde in 10 HIM sodium-phosphate buffer (pH 7.2) for 2.5h. Following fixation, the tissue was washed in the bufferthree times for 10 min each. The tissue was postfixed in 1%OsO4 in water for 1 h and then rinsed three times in waterfor 10 min each. Barley leaf tissue was fixed in the samemanner as for tobacco leaves (Bednarek et al., 1990). Follow-ing dehydration in an ethanol series, the tissue was embeddedin London Resin White acrylic resin (Polysciences, Warring-ton, PA) and polymerized at 60°C under vacuum overnight.Thin sections were prepared on an Ultracut E microtome(Reichert-Jung, NuBloch, Germany) and mounted on form-var-coated nickel grids (EMS, Fort Washington, PA). Immu-nocytochemistry was performed essentially as described byHerman and Melroy (1990) with the antibody to HvAPdiluted 1 part in 10 and the antibody to BL diluted 1 part in50. Control sections were incubated with preimmune serumsimilarly diluted. Protein A-colloidal gold (EY Lab Inc., SanMateo, CA) was diluted 1 to 50. Thin sections were examinedon a JEOL 100CXII transmission microscope.

In Vitro Processing of Prolectin

proBL expressed in Escherichia coli (Schroeder and Raikhel,1992) was purified on GlcNAc columns (Selectin, Pierce) toassure that the protein was in its native conformation; 1.5 ^gof probarley lectin protein was incubated with 0.1 /xg ofaffinity-purified HvAP (Sarkkinen et al., 1992). The reactionwas carried out in 25 pL in a 50-mM lactate buffer (pH 4) at37°C; 20 fiM pepstatin A was used as a specific inhibitor. Forthe gel analysis of the lectin after hydrolysis, samples weretreated with iodoacetamide, loaded on SDS-poIyacrylamidegels, and either stained for protein or analyzed by westernblotting using the monoclonal antibody raised against the C-terminal amino acids of proBL. For the direct N-terminalsequencing as well as for the matrix-assisted laser desorptionmass spectrometry analysis, unprocessed and processed lectinwere purified on reverse-phase HPLC. The N-terminal se-quencing was performed by Dr. N. Kalkkinen (Institute ofBiotechnology, University of Helsinki, Finland), and the ma-trix-assisted laser desorption mass spectrometry analysis wascarried out by Dr. Anita Keane (Finnigan MAT ApplicationLaboratories, Hemel Hempstead, UK).

system. Thus, we wished to test whether HvAP was trans-located into the ER in vitro. Translation of the HvAP messagein vitro produced a protein of molecular mass approximately50 kD (Fig.l, lane 2), which is similar to that calculated fromthe cDNA sequence (54 kD; Runeberg-Roos et al., 1991).When the message was translated in the presence of micro-somes, the molecular mass of part of the translated HvAPincreased to approximately 52 kD (Fig. 1, lane 5). To verifythat this molecular mass change was coupled to translocationinto the microsomes, we tested whether this protein bandwas protected from proteolysis. Incubation of the proteinssynthesized in the presence of microsomes with proteinase Kshowed degradation of only the 50-kD band (Fig. 1, lane 6).When Triton X-100 was included, all proteins were degraded(Fig. 1, lane 7). This supported the hypothesis that the proteincan be translocated into the ER lumen.

To further characterize the ER-signal sequence, the N-terminal sequence of the protein translated in the presenceof microsomes was determined. [3H]Leu radioactivity wasreleased from the N terminus of the proenzyme in cycles 5,10, and 25 (Fig. 2). When this information was compared tothe cDNA sequence (Runeberg-Roos et al., 1991), it indicatedthat the ER signal sequence was cleaved between S25 and E26(Fig. 2). The cleavage site used is one of the four sites foundby computer analysis (PCGENE/psignal) of the cDNA se-quence (Runeberg-Roos et al., 1991). In the computer analy-sis, one site (P22-A23) gets higher scores. This site does not,however, conform to the (-1-3) rule (von Heijne, 1986),whereas the experimentally determined cleavage site does.Therefore, we concluded that the ER signal sequence was

1 2 3

52kD —50kD —

RESULTS

Translocation of HvAP into Microsomes in Vitro

The amino acid sequence as well as the hydrolytic specific-ity and low pH activity optimum of HvAP resemble those ofthe animal lysosomal and yeast vacuolar Asp proteinases(Runeberg-Roos et al., 1991; Sarkkinen et al., 1992; Kervinenet al., 1993). Vacuolar proteins generally are translocated intothe ER during translation as the entry step into the secretory

Figure 1. Proteinase protection of in vitro-translated HvAP. mRNAencoding HvAP was transcribed and translated as described in"Materials and Methods." 35S-labeled proteins that were synthesizedwere then separated on a 10% SDS-polyacrylamide gel and proc-essed for fluorography. Proteins were synthesized in the absence(lane 1) or presence of HvAP RNA (lanes 2-7), and in the absence(lanes 2-4) or presence of canine pancreatic microsomes (lanes 5-7). Proteinase K was added in the absence (lanes 3 and 6) orpresence of Triton X-100 (lanes 4 and 7).




V V VVHgntGLJUALLAAVLLLQTVLPAASEAEGLVRIALKKRPIDRMSRVATGLSGGEEQPLLSGANPLRSB

Figure 2. Determination of the N terminus of HvAP translocatedinto microsomes in vitro. The N-terminal sequence of HvAP de-duced from the cDNA (Runeberg-Roos et al., 1991) is shown ontop. Potential signal sequence processing sites are indicated withopen arrowheads and the N terminus of the mature proteinis indicated by a shaded arrowhead. Below is the tracing ofthe radioactive peaks eluting during the N-terminal sequencing of[3H]Leu-labeled HvAP synthesized in the presence of microsomes.From this information, the proposed signal sequence processingsite is indicated with a filled arrowhead. A short amino acid se-quence in the propeptide having homology to other N-terminalpropeptide vacuolar targeting sequences is underlined.

functional in vitro. The increased molecular mass, which wascoupled to translocation into the microsomes, may be due toglycosylation (Runeberg-Roos et al., 1991).

Biochemical Localization of HvAP in the Vacuoles

To determine if the HvAP was localized to vacuoles, wepurified this organelle from barley leaf protoplasts. To quan-titate the yield of vacuoles, we used a-mannosidase activitybecause the enzyme is known to be exclusively located inmesophyll cell vacuoles (Boiler and Kende, 1979; Martinoiaet al., 1981; Matile, 1987). The contamination of differentcellular organelles in the purified vacuolar fraction was as-sessed by analyzing marker enzyme activities (Table I). Thechloroplast marker NADP-glyceraldehyde phosphate dehy-drogenase and the cytosolic marker Glc-P isomerase were

Table I. Analysis of enzyme activities in protoplast and vacuolarfractions

The range of activities from four independently isolated vacuolefractions is given in parentheses.

Enzymes

CIc-P isomerase3

NADP-glyceralde-hyde-phosphate-dehydrogenasea

Cyt c oxidasea

Proteinase activity atpH 3.7"

per IO6

protoplasts

155307

13.289 x 1CT3

Activities

per 10"vacuoles

12

0.890 X 10~3

Percent invacuoles

1 (0-3)1 (0-1)

6(0-10)101 (97-109)

hardly detectable, but the mitochondria! marker Cyt c oxidasewas present in a minor amount in the vacuolar fraction (TableI). The proteinase activity at pH 3.7 was exclusively locatedin the vacuoles of mesophyll protoplasts based on the yieldof a-mannosidase activity in protoplast and vacuole fractions(see "Materials and Methods" and Table I). This is similar tothe results found by Martinoia et al. (1981) and Heck et al.(1981), who showed that vacuoles isolated from lysed barleymesophyll protoplasts contain 85% of the total proteinaseactivity at pH 3.9. Pepstatin, which is a highly selectiveinhibitor of Asp proteinases (Umezawa, 1976), inhibited 65%of the vacuolar proteinase activity at pH 3.7 (data not shown),indicating that most of the proteinase activity detected withour assay system was due to a vacuolar Asp proteinase. Fromthis we concluded that the Asp proteinase activity, andthus HvAP, was localized in the vacuoles of barley leafprotoplasts.

Localization of HvAP by Immunocytochemical EM

Antibodies made against the two forms of HvAP recog-nized all four subunits (11, 16, 29, and 32 kD) of the twopurified isoforms (Fig. 3A, lane 3), none of which weredetected by the preimmune serum (data not shown). Inaddition, some larger proteins (molecular mass approximately40 and 45 kD) were seen, the same as was sometimes detectedin Coomassie staining of the purified enzyme preparation(Sarkkinen et al., 1992). Whether the larger proteins corre-spond to precursor forms of HvAP remains to be elucidated.In extracts of developing embryos, the antibodies detected

B

1 2 3 4

45kD —

32kD —29kD —

16kD —11kD —

2OkD-18 kD-

' Units in nmol/min.rials and Methods."

b Proteinase units as defined in "Mate-

Figure 3. Western blot analysis. A, Proteins detected by the anti-body against HvAP. Extracts were purified as described (Sarkkinenet al., 1992) from developing embryos (22 DAF; 15 Mg/ lane 1),roots of barley (15 n%, lane 2), or pure HvAP from barley grains (200ng, lane 3) and separated on a 14% SDS-polyacrylamide gel. Pro-teins were transferred to polyvinylidene difluoride membranes andprobed with the antibody raised against HvAP. B, Staining of pro-teins with antibody against mature BL or the CTPP of proBL. Purewheat germ agglutinin (200 ng; lanes 1 and 3) or proBL purifiedfrom £. co// (400 ng; lanes 2 and 4) were separated on a 15% SDS-polyacrylamide gel and transferred to nitrocellulose. Lanes 1 and 2were probed with an antibody against BL and lanes 3 and 4 wereprobed with a monoclonal antibody raised against the CTPP ofproBL (M.R. Schroeder and N.V. Raikhel, unpublished data).




the same bands (Fig. 3A, lane I), whereas in extracts of roots,the detection was weaker and the 11-kD protein band wasmissing (Fig. 3A, lane 2), probably due to a lower amount ofthe enzyme. In root extracts, the 29-kD subunit seemed to beslightly smaller, possibly due to differential processing. Inleaves, the Asp proteinase activity was low and only thestrongest antigen, the 16-kD subunit, was detected in leafextracts (data not shown). Immunocytochemistry of barleyleaves and roots using these antibodies showed specific label-ing in vacuoles of cells (Fig. 4, A and C). No labeling wasdetected in the cytoplasm or the cell wall (Fig. 4, A and C).Only background labeling was observed with the preimmuneserum in both cell types (Fig. 4, B and D), thus confirmingthe biochemical fractionation of HvAP to the vacuoles inbarley leaves.

Digestion of proBL by HvAP

BL is a 36-kD homodimeric protein that is localized in thevacuoles/protein bodies of embryonic and root-tip cells of

barley (Mishkind et al., 1983; Lerner and Raikhel, 1989). It issynthesized as a preproprotein and translocated into the ERwith a concomitant removal of the signal sequence. Withinthe ER, the protein dimerizes and is modified by the additionof N-linked carbohydrate within the CTPP. The CTPP con-tains 15 amino acid residues, many of which are hydrophobic,and it has been shown to be necessary and sufficient forvacuolar targeting in tobacco (Bednarek et al., 1990; Wilkinset al., 1990; Bednarek and Raikhel, 1991). Since only themature form of BL is detected in the vacuole, the processingto remove the C-terminal 15 amino acid residues must occurwithin or just prior to delivery to the vacuole. Based on thevacuolar localization as well as the hydrolytic specificity ofHvAP with synthetic substrates, we speculated that the hy-drophobic CTPP of proBL might be processed by this pro-teinase. Incubation of proBL in its native form, synthesizedin £. co/i (Schroeder and Raikhel, 1992) with purified HvAP,resulted in the loss of a small peptide of about 2 kD fromproBL within 5 min (Fig. 5A, lane 4), whereas the rest of the

! X,V'H,

*

cw

Figure 4. Immunohistochemical localization of HvAP in barley. Barley root tip (A and B) and barley leaves (C and D)were prepared for histochemistry as described in "Materials and Methods." Sections were probed with antibody to HvAP(A and C) or preimmune serum (B and D). Bar = 0.5 ^m.




5 6

20kD18kD

B

20kD —

Figure 5. Processing of proBL with HvAP in vitro. proBL purifiedfrom E. coli (lane 1) was incubated without HvAP (lane 2) or in thepresence of HvAP (lanes 3-7). To inhibit the Asp proteinase activity,pepstatin A was added to some samples (lanes 3 and 5). Theincubation times were for 5 min (lanes 3 and 4), 15 min (lanes 5and 6), or overnight (lanes 2 and 7). The proBL was then separatedon a 17% SDS-polyacrylamide gel and either stained with Coo-massie blue (A) or transferred to polyvinylidene difluoride mem-branes and probed with antibodies to the CTPP of proBL (B).

lectin remained intact, even after incubating overnight (Fig.5A, lanes 4, 6, and 7). This hydrolysis was inhibited by thespecific inhibitor of HvAP, pepstatin A (Fig. 5A, lanes 3 and5), and occurred only when HvAP was added (Fig. 5A, lane2). Antibodies to a related lectin from wheat recognized themature lectin and proBL on western blots (Fig. 3B, lanes 1and 2), whereas monoclonal antibodies to the C-terminalamino acids of proBL recognized only the protein with theCTPP (Fig. 3B, lanes 3 and 4). Western blotting of the proBLwith the monoclonal antibody showed loss of the antigenduring incubation with HvAP (Fig. 5B, lanes 4, 6, and 7).This indicated that HvAP removed at least a part of the Cterminus of BL. Sequencing of the N terminus of the in vitro-processed lectin showed no change in the amino acid residuesin that region of proBL during the hydrolysis (data notshown), demonstrating that all of the changes in the molec-ular mass of the BL most likely were due to changes at the Cterminus. Molecular mass determination of the lectin beforeand after hydrolysis using matrix-assisted laser desorptionspectrometry analysis showed that the sizes of the proteinswere 18,832 ± 18.8 D and 17,556 ± 17.6 D, respectively.This indicated that 1276 ± 36.4 D were lost during hydrolysis,suggesting that 13 amino acid residues (corresponding to1241 D) from the C terminus of proBL were removed duringincubation with HvAP. A loss of 12 and 14 amino acidresidues would correspond to theoretical molecular masschanges of 1170 and 1389 D, respectively.

Immunocytochemical Co-Localization of HvAP and BL

If, indeed, HvAP processes the proBL, the two proteinsmust be localized in the same cells and in the same subcellular

compartment. Previous work with the two proteins separately(Lerner and Raikhel, 1989; this work) suggested that thisshould be the case. Double immunocytochemistry with theHvAP antibody and one directed against the mature barleylectin (Raikhel et al., 1984) confirmed that the proteins canbe localized to the same vacuoles in root-tip cells (Fig. 6A)and embryonic root cells (Fig. 6C).

DISCUSSION

We have shown here by biochemical means and by im-munocytochemistry that the Asp proteinase of barley, HvAP,is localized to the vacuoles in barley leaf and root cells. First,we have demonstrated that the HvAP can be translocated invitro into the lumen of microsomes with the concomitant lossof 25 amino acid residues at the N terminus, the putative ERsignal sequence. Second, we isolated vacuoles from barleyleaf protoplasts and were able to show that the majority ofthe proteinase activity at low pH present in this fraction wasdue to an Asp proteinase. Finally, we were able to detect theHvAP protein exclusively in vacuoles by immunocytochem-istry of barley root and leaf cells. These data, taken togetherwith the western blot results, show that HvAP, or at least aclose homolog to HvAP, is localized to the vacuole in leaves,root tips, and embryonic roots. The similar but not identicalprotein banding pattern in the western blot of these tissues(Fig. 3A) may reflect either differential processing or stabilityof the proteinase subunits. Expression of another related geneis also possible, although only a singly RNA species wasdetected in these different tissues (Runeberg-Roos et al.,1991).

Positive ci's-acting signals are required for soluble proteinsto be targeted to vacuoles. Vacuolar targeting signals in plantshave been found within N-terminal propeptides (Matsuokaand Nakamura, 1991; Holwerda et al., 1992), C-terminalpropeptides (Bednarek et al., 1990; Neuhaus et al., 1991;Melchers et al., 1993), and within the body of the matureprotein (Saalbach et al., 1991; Sonnewald et al., 1991; vonSchaewen and Chrispeels, 1992). The ER-signal sequencecleavage site (determined here) and the N terminus of themature HvAP in vacuoles (Runeberg-Roos et al., 1991; Sark-kinen et al., 1992) define an N-terminal propeptide of 41amino acid residues (see Fig. 2). Although there is no consen-sus vacuolar targeting signal known in plants, it is temptingto suggest that the N-terminal propeptide of HvAP containsa putative signal. In sweet potato sporamin, the sequenceNPIRLP is located in the N-terminal propeptide, and whendeleted (Matsuoka and Nakamura, 1991) or mutated (Naka-mura et al., 1993), can change the vacuolar localization ofsporamin in transgenic tobacco. In barley aleurain, the N-terminal amino acids in the proregion include SNPIR, which,when deleted, alters the vacuolar targeting of aleurain (Hol-werda et al., 1992). The HvAP prosequence contains a similaramino acid sequence, N38PLR, which may be the vacuolartargeting signal of this protein (underlined in Fig. 2). Testsconducted in vivo with mutagenized protein would addressthis hypothesis. The proregion may also play a role in regu-lating proteolytic activity, as has been shown with animalAsp proteinases, where the removal of the prosequence seems




Figure 6. Immunohistochemical co-localization of HvAP and BL in barley. Barley root tip (A and B) or barley embryonicroots (C and D) were prepared for histochemistry as described in "Materials and Methods." Sections in A and C wereprobed with antibody to mature BL, followed by secondary antibody linked to 10-nm gold particles, and then wereprobed with antibody to HvAP, followed by secondary antibody linked to 15-nm gold particles. Sections in B and Dwere probed with control antibodies. Bar = 0.5 Mm-

to be necessary for the activation of the enzyme (Tang andWong, 1987).

The primary structure, hydrolytic specificity, and vacuolarlocalization of HvAP are similar to those of the animallysosomal cathepsin D and the yeast vacuolar Asp proteinaseA. These latter enzymes have been proposed to be involvedin cellular protein turnover and processing and activation ofother enzymes (Zubenko and Jones, 1981; Ammerer et al.,1986; Woolford et al., 1986; Gottschalk et al., 1989; Nishi-mura et al., 1989; Teichert et al., 1989). Knowing the aminoacid bond specificity of the HvAP (Kervinen et al., 1993), wehypothesized that the proteinase would be able to processthe C-terminal proregion of BL. This region is 15 amino acidslong and contains many hydrophobic residues that couldtheoretically be processed by HvAP. The lectin is localized tovacuoles within cells and is found in root-tip cells of barley

seedlings, similar to the HvAP. Here we were able to showthat the proBL can be processed by HvAP in vitro to a sizesimilar to that found in the vacuoles. Both the N-terminalsequencing of the processed lectin and loss of antigenicity toa C-terminal-specific antibody after processing suggested lossof a C-terminal peptide during the incubation. Matrix-assisted laser desorption mass spectrometry suggested theloss of 13 C-terminal amino acids from the proBL during theprocessing. This is two amino acids less than the full lengthof the propeptide, based on the complete sequence of a highlyrelated lectin from wheat (Wright and Raikhel, 1989). Assum-ing the vacuolar form of BL is similar to wheat germ agglu-tinin, the additional amino acids would have to be removed,possibly by a carboxypeptidase.

But does HvAP process proBL in vivo? The work presentedhere provides an indication but not direct evidence for this.




We do not know if the Asp proteinase present in root tips with proBL has enzymic properties identical to the grain Asp proteinase used for the in vitro assay, although Asp proteinases generally cleave bonds between amino acid residues with large hydrophobic side chains, and the HvAP hydrolytic specificity is similar to the mammalian lysosomal cathepsin D and the yeast vacuolar proteinase A (Kervinen et al., 1993). Therefore, if there are similar but not identical Asp proteinases in the different tissues, their hydrolytic specificity is most probably very similar. The crystal structure of BL shows a very compact molecule with many disulfide bonds (Wright et al., 1993). Although only the preliminary structure of proBL has been determined (Wright et al., 1993), it is assumed to be similar for the bulk of the protein, since both forms can bind GlcNAc (Lerner and Raikhel, 1989; Schroeder and Raikhel, 1992). The CTPP may be exposed on the surface of the molecule, and thus could be susceptible to many types of proteinases. Indeed, the proBL can also be processed in vitro by yeast carboxypeptidase Y (data not shown). Thus, HvAP may participate in the processing of proBL in vivo, but other enzymes might also be involved. Extensive mutagenesis of the BL CTPP shows that many amino acids can be tolerated in that region and not disrupt processing of the BL in tobacco (Dombrowski et al., 1993). Thus, the proteinase(s) that processes the proBL must have a rather broad specificity, or severa1 different proteases may participate in the processing. But many of these specific changes do not alter potential processing sites of HvAP based on the hydrolytic specificity of this enzyme (Kervinen et al., 1993). As yet, none of these mutant proteins have been tested with HvAP in vitro.

The next question is in which subcellular compartment is proBL processed and is that the same compartment in which HvAP is active? Here we show that HvAP from vacuoles is active at pH 3.7, and that HvAP and BL are co-localized in vacuoles of barley embryonic and germinating roots. The site where the CTPP of proBL is removed is not known, but since the CTPP is required for correct vacuolar targeting (Bednarek et al., 1990), and since only the mature form of BL is found in the vacuoles of cells (Mishkind et al., 1983), processing must occur prior to or concomitant with the delivery to this organelle. Processing is probably not required for targeting, however, since a mutant CTPP containing a bulky glycan near the processing site is only slowly processed, but is nonetheless located in the vacuoles of tobacco (Dombrowski et al., 1993; J.E. Dombrowski and N.V. Raikhel, unpublished observations).

The data presented here are an indication that HvAP is involved in the processing of proBL in vivo. Whether other endo- and/or exoproteases are also involved remains to be elucidated. The effect of overexpression or reduced expression of the HvAP cDNA in vivo on the processing of BL would provide one way to test this directly.

ACKNOWLEDCMENTS

The authors would like to thank Nisse Kalkkinen for the N- terminal sequencing, Ossi Renkonen for suggesting. the matrix-assisted laser desorption mass spectrometry, and Anita Keane (Finnigan MAT Application Laboratories, UK) for performing the analysis. We also gratefully acknowledge Mart Saarma and Teemu Teeri for helpful discussions during different stages of the work. Part of the

studies were camed out at VTT, Biotechnical Laboratory, Espoo, Finland, and Tuomas Sopanen and Marjatta Salmenltallio-Marttila are acknowledged for their help there. We thank Liisa Pyhala and Leena Liesirova for the production of the antibodies to HvAP at the Nationa1 Public Health Institute, Helsinki. We thank Maior Bar-Peled, Jim Dombrowski, Tracey Reynolds, and Kirsi Tormakangas for critica1 reading of the manuscript.

Received November 22, 1993; accepted February 1, 1994. Copyright Clearance Center: 0032-0889/94/105/0321/09.

LITERATURE CITED

Ammerler G, Hunter CP, Rothman JH, Saari GC, Valls LA, Stevens TH (1986) PEP4 gene of Saccharomyces cerevisiae encodes proteinase A, a vacuolar enzyme required for processing of vacuolar precursors. Mo1 Cell Biol6: 2490-2499

Bednarek SY, Raikhel NV (1991) The barley lectin carboxyl-termi- na1 propeptide is a vacuolar protein sorting determinant in plants. Plant Cel l3 1195-1206

Bednarek SY, Wilkins TA, Dombrowski JE, Raikhel. NV (1990) A carboxyl-terminal propeptide is necessary for proper sorting of barley lectin to vacuoles of tobacco. Plant Cell 2: 1145-1155

Belozersky MA, Sarbakanova ST, Dunaevsky YE (1.989) Aspartic proteinase from wheat seeds: isolation, properties .and action on gliadin. Planta 177: 321-326

Biehl B, Heinrichs H, Ziegeler-Berghausen H, Srivastava S, Xiong Q, Passem D, Senyuk VI, Hammoor M (1993) The proteases of ungenminated cocoa seeds and their role in the fermentation process. Angew Bot 67: 59-65

Boller T, Kende H (1979) Hydrolytic enzymes in the central vacuole of plaint cells. Plant Physiol 6 3 1123-1132

Boudet AM, Canut H, Alibert G (1981) Isolation anc. characterization of vacuoles from Melilotus alba mesophyll. Plant Physiol 6 8 1354-1358

Bourgeolis J, Malek L (1991) Purification and characterization of an aspartyl proteinase from dry jack pine seeds. Seed Sci Res 1:

Campbcll DJ (1987) Circulating and tissue angiotensin systems. J Clin Invest 7 9 1-6

Davies DR (1990) The structure and function of the aspartic proteinases. Annu Rev Biophys Biophys Chem 19: 189-215

D’Hondt K, Bosch D, Van Damme J, Goethals M, Vanidekerckhove J, Krebbers E (1993) An aspartic proteinase present in seeds cleaves Arabidopsis 2 s albumin precursors in vitro. J Bicil Chem 2 6 8

Doi E, Shibata D, Matoba T, Yonezawa D (1980) Characterization of pepstatin-sensitive acid protease in resting rice secds. Agric Biol Chem 44: 741-747

Dombrowski JE, Schroeder MR, Bednarek SY, Raikhel NV (1993) Deterrnination of the functional elements within the vacuolar targeting signal of barley lectin. Plant Cell 5: 587-55’6

Dunaevsky YE, Sarbakanova ST, Belozersky MA ((1989) Wheat seed carboxypeptidase and joint action on gliadin of proteases from clry and germinating seeds. J Exp Bot 40: 1323--1329

Elpidinai EN, Dunaevsky YE, Belozersky MA (1990) l’rotein bodies from buckwheat seed cotyledons: isolation and characterization. J

Faro CJ, Moir AJG, Pires EV (1992) Specificity of a milk clotting enzyme extracted from the thistle Cynara cardunculu:; L.: action on oxidized insulin and K-casein. Biotechnol Lett 14 841-846

Foltmann B (1981) Gastric proteinases-structure, function, evolution arid mechanism of action. Essays Biochem 17: 52-84

Gottschalk S, Waheed A, Schmidt 8, Laidler P, von Figura K (1989) Sequential processing of lysosomal acid phosphatase by a cytopli~smic thiol proteinase and a lysosomal aspartyl proteinase.

Heck U, Martinoia E, Matile P (1981) Subcellular localization of acid proteinase in barley mesophyll protoplasts. Planta 151:

Heimgartner U, Pietrzak M, Geertsen R, Brodelius P, da Silva

139-147

20884-20891

EXP Bot 41: 969-977

EMBO J 8: 3215-3219

198-200




Figueiredo AC, Pais MSS (1990) Purification and partia1 characterization of milk clotting proteinases from flowers of Cynara cardunculus. Phytochemistry 2 9 1405-1410

Herman EM, Melroy DL (1990) Electron microscopic immunocytochemistry in plant molecular biology. Plant Mo1 Biol Manual B13:

Holwerda BC, Padgett HS, Rogers JC (1992) Proaleurain vacuolar targeting is mediated by short contiguous peptide interactions. Plant Cel l4 307-318

Kaiser G, Martinoia E, Wiemken A (1982) Rapid appearance of photosynthetic products in the vacuoles isolated from barley mesophyll protoplasts by a new fast method. 2 Pflanzenphysiol 107:

Kalkkinen N, Tilgmann C (1988) A gas-pulsed-liquid-phase sequencer constructed from a Beckman 890D instrument by using Applied Biosystems delivery and cartridge blocks. Joumal of Pro- tein Chemistry 7: 242-243

Kervinen J, Sarkkinen P, Kalkkinen N, Mikola L, Saarma M (1993) Hydrolytic specificity of the barley grain aspartic proteinase. Phy- tochemistry 32: 799-803

Leegood RC, Walker DA (1983) Chloroplasts. Zn JL Hall, AL Moore, eds, Isolation of Membranes and Organelles from Plant Cells. Academic Press, London, pp 185-210

Lerner DR, Raikhel NV (1989) Cloning and characterization of root-specific barley lectin. Plant Physiol91: 124-129

Martinoia E, Flügge UI, Kaiser G, Heber U, Heldt HW (1985) Energy-dependent uptake of malate into vacuoles isolated from barley mesophyll protoplasts. Biochim Biophys Acta 806 31 1-319

Martinoia E, Heck U, Wiemken A (1981) Vacuoles as storage compartments for nitrate in barley leaves. Nature 289 292-294

Matile P (1987) The sap of plant cells. New Phytol 105: 1-26 Matsuoka K, Nakamura K (1991) Propeptide of a precursor to a

plant vacuolar protein required for vacuolar targeting. Proc Natl Acad Sci USA 88: 834-838

Melchers LS, Sela-Buurlage MB, Vloemans SA, Woloshuk CP, Van Roekel JSC, Pen J, van den Elzen PJM, Cornelissen BJC (1993) Extracellular targeting of the vacuolar tobacco proteins AP24, chitinase and /3-1,3-glucanase in transgenic plants. Plant Mo1 Biol 21: 583-593

Mishkind ML, Palevitz BA, Raikhel NV, Keegstra K (1983) Local- ization of wheat germ agglutinin-like lectins in various species of the Gramineae. Science 2 2 0 1290-1292

Nakamura K, Matsuoka K, Mukumoto F, Watanabe N (1993) Processing and transport to the vacuole of a precursor to sweet potato sporamin in transformed tobacco cell line BY-2. J Exp Bot

Neuhaus J-M, Sticher L, Meins F Jr (1991) A short C-terminal sequence is necessary and sufficient for the targeting of chitinases to the plant vacuole. Proc Natl Acad Sci USA 88: 10362-10366

Nishimura Y, Kawabata T, Furuno K, Kato K (1989) Evidence that aspartic proteinase is involved in the proteolytic processing event of procathepsin L in lysosomes. Arch Biochem Biophys 271:

Polanowski A, Wilusz T, Kolaczkowska MK, Wieczorek M, Wil- imowska-Pelc A (1985) Purification and characterization of aspartic proteinases from Cucumis sativus and Cucurbita maxima seeds. Zn V Kostka, ed, Aspartic Proteinases and their Inhibitors. Walter de Gruyter, New York, pp 49-52

1-24

103-113

SUppl44 331-338

400-406

Raikhel NV, Mishkind ML, Palevitz BA (1984) Characterization of a wheat germ agglutinin-like lectin from adult wheat plants. Planta

Rodrigo I, Vera P, Van Loon LC (1991) Degradation of tobacco pathogenesis-related proteins. Plant Physiol 9 5 616-622

Runeberg-Roos P, Tormakangas K, Ostman A (1991) Primary structure of a barley-grain aspartic proteinase. A plant aspartic proteinase resembling mammalian cathepsin D. Eur J Biochem 202:

Saalbach G, Jung R, Kunze G, Saalbach I, Adler K, Muntz K (1991) Different legumin protein domains act as vacuolar targeting signals. Plant Cell 3: 695-708

Sarkkinen P, Kalkkinen N, Tilgmann C, Siuro J, Kervinen J, Mikola L (1992) Aspartic proteinase from barley grains is related to mammalian lysosomal cathepsin D. Planta 186: 317-323

Schroeder MR, Raikhel NV (1992) Isolation and characterization of pro-barley lectin expressed in Escherichia coli. Protein Expression and Purification 3: 508-511

Sonnewald U, Brauer M, von Schaewen A, Stitt M, Willmitzer L (199 1) Transgenic tobacco plants expressing yeast-derived inver- tase in either the cytosol, vacuole or apoplast: a powerful tool for studying sucrose metabolism and sink/source interactions. Plant J

St. Angelo AJ, Ory RL (1970) Properties of a purified proteinase from hempseed. Phytochemistry 9 1933-1938

Tang J, Wong RNS (1987) Evolution in the structure and function of aspartic proteinases. J Cell Biochem 3 3 53-63

Teichert U, Mechler B, Müller H, Wolf DH (1989) Lysosomal (vacuolar) proteinases of yeast are essential catalysts for protein degradation, differentiation, and cell survival. J Biol Chem 264

Tokes ZA, Woon WC, Chambers SM (1974) Digestive enzymes secreted by the carnivorous plant Nepenthes macferlanei L. Planta

Umezawa H (1976) Structures and activities of protease inhibitors of microbial origin. Methods Enzymol45 678-695

von Heijne G (1986) A new method for predicting signal sequence cleavage sites. Nucleic Acids Res 14 4683-4690

von Schaewen A, Chrispeels MJ (1992) Identification of vacuolar sorting information in phytohemagglutinin, an unprocessed vacuolar protein. J Exp Bot 4 4 339-342

Wilkins TA, Bednarek SY, Raikhel NV (1990) Role of propeptide glycan in post-translational processing and transport of barley lectin to vacuoles in transgenic tobacco. Plant Cell 2 301-313

Woolford CA, Daniels LB, Park FJ, Jones EW, Van Arsdell JN, Innis MA (1986) The PEP4 gene encodes an aspartyl protease implicated in the posttranslational regulation of Saccharomyces cerevisiae vacuolar hydrolases. Mo1 Cell Biol 6 2500-2510

Wright CS, Raikhel NV (1989) Sequence variability in three wheat germ agglutinin isolectins: products of multiple genes in polyploid wheat. J Mo1 Evol28: 327-336

Wright CS, Schroeder MR, Raikhel NV (1993) Crystallization and preliminary x-ray diffraction studies of recombinant barley lectin and pro-barley lectin. J Mo1 Biol 233 322-324

Zubenko GS, Jones EW (1981) Protein degradation, meiosis and sporulation in proteinase-deficient mutants of Saccharomyces cerevisiae. Genetics 97: 45-64

162 55-61

1021-1027

1: 95-106

16037-16045

119 39-46



Documents

The Aspartic Proteinase of Barley 1s a Vacuolar Enzyme That … · Plant Physiol. (1 994) 105: 321 -329 The Aspartic Proteinase of Barley 1s a Vacuolar Enzyme That Processes Probarley