Eur. J . Biochem. 248, 304-312 (1997) 0 FEBS 1997

Proteinase A, a storage-globulin-degrading endopeptidase of vetch (Vicia sativa L.) seeds, is not involved in early steps of storage-protein mobilization Claudia BECKER', Vitalyi I. SENYUK', Andrei D. SHUTOV', Van Hai NONG'.', Jurgen FISCHER', Christian HORSTMANN ' and Klaus MUNTZ' I Institut fur Pflanzengenetik und Kulturpflanzenforschung, Gatersleben, Germany ' Laboratory of Protein Chemistry, Moldavian State University, Kishinev, Republic of Moldova ' Institute of Biotechnology, National Center for Natural Science and Technology, Hanoi, Vietnam

(Received 28 AprilRO June 1997) - EJB 97 0604/1

Proteinase A is a papain-like cysteine endopeptidase of vetch (Vicia sativu L.) which was assumed to initiate storage-globulin breakdown just after the onset of seed germination. This enzyme was purified from cotyledons of vetch seedlings. On gelatin-containg SDS gels, active proteinase A migrated with an apparent molecular mass of 21 kDa, whereas after heat denaturation its molecular size on SDSPAGE was 29 kDa. Although proteinase A is capable of hydrolyzing storage globulins in vitro it could not be localized in the protein-body fraction of cotyledons from germinating seeds. cDNA clones encoding proteinase A precursor have been obtained by PCR. The precursor is composed of an N-terminal signal sequence followed by a propeptide, the region encoding mature proteinase A, and a C-terminal KDEL sequence. Mature proteinase A with a derived molecular mass of 25244 Da does not have the KDEL sequence. The derived amino acid sequence of the proteinase A precursor is 78.2% identical to sulfhydryl- endopeptidase (SH-EP), a cysteine endopeptidase from germinating Vigna tnungo seedlings. Northern blot analysis indicated that proteinase A mRNA appears de novo in cotyledons of I-day-germinated vetch seeds, where its amount increases up to day 6. No proteinase A mRNA was detected in other vetch organs, not even in the embryo axis, which contains stored globulins. By means of antibodies raised against the purified and against recombinantly produced proteinase A, the 29-kDa bands of mature pro- teinase A were detected in cotyledon extracts of 6-day-germinated seeds when globulin degradation has already far proceeded. The reported data do not agree with the proposed triggering role of proteinase A i n storage-globulin breakdown during germination.

Keywords: triggering cysteine endopeptidase ; seed germination; Vicia sativa L. ; endoplasmic reticulum retention signal.

The beginning of storage-protein degradation is indicated by slight mobility changes of globulins from cotyledon extracts of germinating dicotyledonous seeds on electrophoresis (Shutov and Vaintraub, 1973; Lichtenfeld et al., 1979, 1981; Boylan and Sussex, 1987; Dunaevsky and Belozersky, 1989). During this first stage of breakdown the holoproteins remain assembled, which is reflected by nearly unchanged positions in the sucrose- density gradient after centrifugation and in gels after electropho- resis under non-denaturing conditions. Only small trichloro- acetic-acid-soluble peptides are lost as the result of limited pro- teolysis (Shutov and Vaintraub, 1987). These first modifications are thought to be catalyzed either by stored endopeptidases that have to be activated just after imbibition, as in buckwheat (El- pidino et al., 1991), or by newly formed endopeptidases, as as- sumed for the majority of legumes (Chrispeels et al., 1976;

Correspondence to C. Becker, Institut fur Pflanzengenetik und Kulturpflanzenforschung (IPK), Corrensstrasse 3, D-06466 Gatersleben, Germany

Far: +49 39482 5 366. Abbrc.viution. SH-EP, sulthydryl-endopeptidase. Note. The novel nucleotide sequence data published here have been

submitted to the EMBL Data Library and are available under accession number 234859.

Lichtenfeld et al., 1979; Muntz et al., 1985; Shutov and Vaintraub, 1987). Only after 3-6 days depending on the species and germination conditions, a second stage with rapid storage- globulin degradation begins, accompanied by strong increases in proteolytic activity, which results from de novo synthesis of proteinases.

According to Shutov and Vaintraub (1987) a cysteine endo- peptidase, termed proteinase A (Shutov et al., 1984), is responsi- ble for the triggering of limited proteolytic cleavages during stage 1 of globulin degradation in cotyledons of germinating vetch seeds. As a consequence, conformational changes are ex- pected to occur that open the proteins to the combined attack by proteinase A with a second cysteine endopeptidase, called proteinase B (Shutov et al., 1982), and stored carboxy-pepti- dases leading to rapid complete breakdown during stage 2. We have shown that proteinase B is an Asn-specific cysteine endo- peptidase (Becker et al., 1995). It belongs to the recently discov- ered new group of legumain-like cysteine endopeptidases (Ishi, 1994) that includes the 1 1s-globulin-processing proteinases from developing seeds. Proteinase B exhibits similar processing activity but at least in vitro cannot attack unmodified globulins from mature seeds. Similar confirmation of the presumed func- tion of proteinase A is lacking.

Becker et ai. (Eur J . Biochem. 248) 305

Here we report on the purification of proteinase A from 8-day-germinated vetch seeds, and on the analysis of proteinase- A-specific cDNA synthesized from mRNA of 3-day-germinated vetch. Whereas the enzyme rapidly degrades globulins in vitro it is not present in protein bodies during the early stages when proteinase A was proposed to initiate globulin degradation in cotyledons of germinating vetch. Mature proteinase A could only be immunologically detected in cotyledon extracts after 6 days germination when the isolation of intact protein bodies has become impossible and globulin degradation has far pro- ceeded. Transcripts encoding proteinase A were only detected in cotyledons after 24 h germination and their amounts subse- quently increased. These results contradict the proposal of a trig- gering function of proteinase A in globulin degradation. Other candidates for this function have to be sought and might be found among cysteine proteinases that appear earlier during vetch germination (Becker et al., 1994).

MATERIALS AND METHODS Plant materials. Kcia sativa L. [cv. cosentini (Guss) Ar-

cang] plants were grown under greenhouse conditions of 18 h light at 20°C. For germination, vetch seeds were soaked in water for 2 h, spread on wet filter paper and germinated in the dark at 2.5 "C. Cotyledons just after imbibition were considered to repre- sent day zero of germination. Proteinase A was isolated from etiolated cotyledons of day 8 seedlings. For cDNA cloning, northern blot analysis and immunoblotting samples were taken after different times of seed development and germination. Har- vested tissues were frozen in liquid nitrogen and stored at -70°C until use.

Purification of proteinase A. All purification steps were carried out at 4°C. Lyophilized cotyledons (200 g) were homog- enized in 2 L 2 mM dithiothreitol. After centrifugation at 6000 rpm (RCSC, Sorvall Instruments) for 20 min, the superna- tant was titrated to pH 4.0 with 0.05 M HCl to precipitate the enzyme with other proteins. The pellet was dissolved in 0.2 M NaCl and adjusted to pH 6.5. Contaminating proteins were par- tially removed by a second isoelectric precipitation at pH 4.8. The supernatant was applied to an anion-exchange column (Fractogel TMAE-650 M, 2.6 cmX26 cm) equilibrated with 0.2 M NaCI, 37 mM sodium phosphate, pH 6.5. A linear gradi- ent from 0.2 M to 0.6 M NaCl was used for elution of the en- zyme. Active fractions were pooled, NaCl was added up to 4.0 M, and the sample was loaded on a phenyl-Sepharose CL-4B column (1.1 cmX8.0 cm) equilibrated with 4 M NaCl, 37 mM sodium phosphate, pH 6.5. The enzyme was eluted with a linear gradient from 4.0 M to 0 M NaCI. Gel filtration was used as the final purification step. The enzyme was retarded on Sephadex G- S O (1.2 cmX60 cm) as shown previously (Shutov et al., 1984). It was eluted by 37 mM sodium phosphate, pH 6.5. A relatively pure and concentrated proteinase A preparation tends to autolyze at pH below 6.0. Therefore, all the chromatographic procedures were performed in 37 mM sodium phosphate, pH 6.5, 0.5 mM EDTA, 2 mM dithiothreitol. Protein concentration was deter- mined using the Bio-Rad protein-assay system according to the standard protocol.

Protease assays. Hydrolysis of casein was assayed as de- scribed previously (Shutov et al., 1982). 1 U enzyme activity represented the release of 1 pM trichloroacetic acid-soluble amino groups/min at 30°C. Effects of various inhibitors on pro- teinase A activity were investigated after incubating aliquots of the purified enzyme for 30 min at 4°C with the inhibitors in the presence of 2 mM dithiothreitol.

SDS/PAGE. Analytical gel electrophoresis under denaturing conditions was performed in 12.5 % (mass/vol.) polyacrylamide

gels (Laemmli, 1970). Proteins were stained with Coomassie Blue R250. To identify active proteinase bands, gel slabs con- taining 0.05% gelatin were used according to Becker et al. (1995). Samples were not heated prior to loading for zymo- graphic analyses.

Cleavage of protein substrates by proteinase A. Siihsfrate preparation. Mature vetch seeds were extracted with 50% (mass/vol.) ammonium sulfate solution. Solid ammonium sulfate was added to the extract to obtain 90% saturation. After centrif- ugation the precipitated crude storage-globulin preparation was dissolved in and dialysed against distilled water overnight at 4°C. Vicilin was extracted from the resulting precipitate by 0.25 M sodium acetate, pH 4.8. The remaining pellet was dis- solved in 0.25 M NaCI, pH 7.0. Legumin was isoelectrically precipitated from this solution at pH 4.8. The preparations of legumin and vicilin were applied to phenyl-Sepharose CL4B columns to eliminate partially denatured storage globulins.

Cleavage ussay. Each globulin (1.5 mg in 150 pl McIlvaine buffer, pH 5.6, 0.18 M NaCl) was mixed with 5 pl (2.5 mu) en- zyme solution. After 60 min incubation at 30°C the reaction was stopped by the addition of 15% trichloroacetic acid to 3.75%. Digestion products were analyzed by SDWPAGE.

CNBr-cleavage and peptide sequencing. SO pg of the lyo- philized enzyme preparation were dissolved in 200 pl of 70% formic acid. 50 p1 CNBr stock solution (40 pg CNBr in 400 pl acetonitrile and 600 p1 formic acid) were added. The cleavage reaction was carried out for 15 h at room temperature in the dark. The cleavage products were separated by HPLC. Sequenc- ing was performed as described previously (Becker et al., 1995).

Preparation of total cotyledon extracts and immunoblot analysis. Frozen cotyledons (500 mg) from suitable develop- mental stages were ground in a mortar with a pestle in 3 ml 0.1 M Tris/HCI, pH 8.0, 0.15 M NaCl. The extracts were incu- bated with agitation at 4°C for 30 min, centrifuged at 13 000 rpm (Biofuge 13R, Heraeus Sepatech) for 30 min, and the superna- tants were used for SDS/PAGE. After electrophoresis the protein samples were electrophoretically transferred onto a cellulose ni- trate filter (Serva). The immunoreactions were carried out as described (Becker et al., 1995), except that proteinase A anti- bodies raised either against the purified plant enzyme or recom- binant proteinase A were used at a dilution of 1 : 10000.

Isolation of protein bodies. For preparation of protein bod- ies the method of Mader and Chrispeels (1984) was followed. Cotyledons were ground in 0.6 M mannitol, 100 mM Mes, pH 5.5, 1 mM EDTA (buffer A), using 2 m1/10 cotyledons. The homogenates were filtered through miracloth and layered on 8 ml5% (mass/vol.) Ficoll in buffer A and centrifuged at 100Xg for 10 min. After washing twice with buffer A the pellets con- taining protein bodies were lysed in I00 mM Tris/HCI, pH 7.8, 1 mM EDTA, and aliquots were prepared for SDS/PAGE. The cytoplasmic proteins from the supernatant of the Ficoll gradients were precipitated with trichloroacetic acid, washed with acetone and dissolved in 100 mM Tris/HCl, pH 7.8, 1 mM EDTA.

Nucleic acid isolation procedures. Total RNA was isolated from cotyledon material of different developmental stages and germination days and from leaves, roots and shoots essentially as described by Becker et al. (1995). Genomic DNA for South- ern blot analysis was isolated from 3-5-day-old seedlings of V sativa by the method of Miller et al. (1988), modified for plant material (Becker et al., 1995).

cDNA preparation and cloning. Extraction and preparation of poly(A)-rich RNA by paramagnetic particle technology (PolyATract mRNA system; Promega) followed the protocol recommended by the manufacturer. Single-stranded cDNA was synthesized from poly(A)-rich RNA from 3-day-germinated cotyledons by means of an oligo(dT) primer and reverse

306 Becker et al. (ELM J . Bioc,keni. 248)

trnnscriptase (First Strand cDNA Synthesis kit, Stratagene) and subjected to PCR with a degenerate oligonucleotide primer (5’-

T

AT$ GA:. TGG :;GI AAT, AA; GGI GCT GTI ACI GG-3’) that was synthesized on the basis of the N-terminal amino acid se- quence of the purified enzyme (IDWRNKGAVTG) in combina- tion with oligo(dT). The PCR assay conditions were the same as described previously (Becker et al., 1995), except that high- fidelity DNA polymerase (Boehringer Mannheim) was used. The amplified fragments (850 bp) were cloned into the pCRIl vector (Tnvitrogen) and analyzed by sequencing from both strands on an automated DNA sequencer (ALF, Pharmacia). After sequence analyses of several clones by means of the PC/ GENE program (IntelliGenetics), a sequence-specific antisense primer upstream of the poly(A) tail was used in combination with oligo(dG) to screen a second cDNA library (5’ RACE sys- tem, GIBCO BRL) for full-length cDNA. The full-length clone pSK19 was used for further analyses.

Production of proteinase A in Escherichia coli. Ncol frag- ments containing the coding sequences for either the mature pro- teinase A or pro-proteinase A were generated by PCR amplifica- tion of pSK19 and inserted into the corresponding site of the pET3d expression vector (Novagen). Single colonies of E. coli HMS 174(DE3), harbouring the expression plasmids with the inserts in correct orientation, were grown in Luria-Bertani me- dium overnight. The overnight culture was diluted with fresh Luria-Bertani to a starting absorbance (A,,,,,,,,,,) of 0.3. Isopropyl- thio-P-D-galactoside was added to 1 .O mM to start gene expres- sion. After 3-5 h the bacterial cells were harvested by centrifu- gation, suspended in 0.1 vol. 50 mM Tris/HCI, pH 8.0, 10 mM EDTA, 1 mg/ml lysozyme, and disrupted by sonication using a Vibra-Cell sonicator (BioBlock). The total lysate was separated into soluble and ‘inclusion body’ fractions by centrifugation at 13 000 rpm (Biofuge 13R, Heraeus Sepatech) for 30 min at 4 “C. The sedimented inclusion bodies were solubilized with 6 M urea and electrophoretically fractionated. The recombinant proteins were purified by elution from SDSIPAGE slabs. Recombinant proproteinase A eluate was used for immunization of a rabbit.

To determine whether purified plant proteinase A has a KDEL sequence, four oligonucleotide primers were designed ac- cording to different C-terminal parts of proteinase A and used for PCR amplification in combination with an oligonucleotide corresponding to the N-terminal sequence of mature proteinase A. Primer 43 was used to amplify the complete cDNA sequence of mature proteinase A including KDEL, primer 98 to generate mature proteinase A without KDEL, primer 111 to amplify pro- teinase A without n ine C-terminal amino acids, and primer 99 to amplify proteinase A without 12 C-terminal amino acids (Fig. 3). The PCR fragments were introduced in pET3d. The recombinant plasmids were transformed into E. coli HMSl74(DE3), and expression was performed as described above. Inclusion bodies were lysed with 50 mM Tris/HCI. pH 8.0. 10 mM EDTA, and the recombinant proteins were heat denatured and loaded on a gel with the purified plant enzyme.

Production of antibodies. To raise antibodies against re- combinant proproteinase A the eluates. which contained about 1 mg/nil recombinant proteinase A, were used for immunization of a rabbit. To raise antibodies against native proteinase A from vetch seeds the purified enLyme solution (about 500 pg) without any inactivation of the proteinase by heat or SDS treatment was used for immunization. Rabbit sera were purified using Econo- Pac Serum IgG Purification Columns (Bio-Rad).

Hybridization procedures. Northern blot analysis was per- formed a s dr.iriketi in Sambrook et al. (1989). 5 pg total RNA were denatured wi!h ylyoxal, electrophoresed i n a 1.2(!,’ ::garose gel, blotted ontc. yv!im membrane (Nytran-N+) by downward

Tahle 1. Purification of vetch proteinase A from 25 g cotyfedons. n.d.. not determined. 1 U emynie activity represented the release of 1 pM trichloroacetic-acid-soluble amino groupdmin at 30°C.

Purification Protein Activity Specific Yield Purifi step activity cation

mg U U/mg % -fold

Crude extract 569 11.6 0.0193 100 1 lsoelectric

precipitation 68.4 8.04 0.1180 73.1 6.1 1 Fractogel

TMAE-650 M 0.282 5.22 18.5 47.5 959 Phen yl-Sepharose

CL-4B n.d. 3.73 n.d. 33.9 n.d. Sephader G-50 0.073 1.83 25.1 16.6 1300

A B 1 2 1 2

kDa

--69.0-

-43.0-

-30.0 I +

II + -20.1

-14.4

- Heating 4- - + Fig. 1. SDSRAGE of purified proteinase A. SDS/PAGE without (A) and with (B) gelatin in the gels. Lane 1, heated sample: lane 2, unheated sample. The posit ions of molecular-mass markers are indicated. Position I of proteinax A corresponds to the unfolded polypeptide chains. With- out previous heating the isozymes migrated with an increased mobility (position 11).

transfer (Turboblotter, Schleicher & Schiill) and fixed by ultra- violet cross-linking. For Southern blot analysis 10 pg genomic DNA digested with an excess of EcoRI, EcoRV and BarnHl were separated in a 0.8% agarose gel and blotted as described above. Hybridizations were carried out in 6XNaCVCit (NaCI/ Cit is 0.15 M NaCI, 15 mM sodium citrate, pH7.0), 5XDen- hardt, 0.5% SIX, and 0.5% blocking solution (Boehringer) at 65°C overnight. The ”P-labelled insert of pSK19 was used as a probe. The filters were washed twice in 2xNaCl/Cit, 0.1 c / o SDS and twice in O.ilXNaCl/Cit, 0.1 % SDS at 65°C for 4 h. Mem- branes were exposed to Hyperfilm-MP X-ray films with an in- tensifying screen (Amersham) for 1-5 days at -70°C.

RESULTS

Preparation and characterization of proteinase A. The results of the purification procedure for proteinase A are summarized i n Table I . The final enzyme preparation obtained after gel chro- matography gave two closel:. migrating protein bands as shown on SDSfPAGE under reduc,rii i’ conditions (Fig. I ). Sequencing revealed that both polypeptides possess identical N-terminal

Becker et al. ( E M J. Biochem. 248) 307

Table 2. Effects of various protease inhibitors on the activity of pro- teinase A. All assays included 2 mM dithiothreitol.

Protease inhibitor Concentration Activity

None E-64 Iodoacetate Iodoacetamide N-Ethylmaleimide Leupeptin Phenylmethylsulfonyl fluoride Pepstatin A EDTA

mM

2 0.01

10 10 10 0.1 1 0.003 1

% of control

100 6 4 4 9 6

110 124 140

vicilin legumin

kDa -1 proteinaseA

94.0 - 67.0 -

43.0 -

30.0 - f 22-24 kDa legumin beta chains

20.1 -

14.4

1 2 3 4

Fig. 2. SDSIPAGE showing hydrolysis of legumin and vicilin by pro- teinase A. Lane 1, intact vicilin; lane 2, partially hydrolyzed vicilin; lane 3, intact legumin; lane 4, partially hydrolyzed legumin. Aliquots of the purified legumin and vicilin (1.5 mg each) were incubated with 2.5 mU purified proteinase A for 60 min at 30°C. Partial degradation products of legumin and vicilin appeared when the hydrolysis time was restricted. Otherwise, complete degradation of both the legumin a-chain and vicilin occurred.

amino acid sequences : VPSSIDWRNKGAVTG. The positions to which the proteins migrated in SDS gel slabs depended upon whether the sample had been heated or not (Fig. 1 A). After heat denaturation the enzyme migrated with a mobility corresponding to a molecular mass of 29 kDa. Presumably this position corre- sponds to the unfolded polypeptide chains. Without heating two protein bands with apparent molecular masses of about 21 kDa (Fig. 1 A) became detectable and showed gelatin-degrading ac- tivity (Fig. 1 B). If the 21-kDa bands were eluted from gel slabs and electrophoresed after heating, they migrated with a mobility corresponding to an apparent molecular mass of 29 kDa.

Similarly to papain-like cysteine proteinases, proteinase A was inhibited strongly by a number of cysteine proteinase inhibi- tors and activated by reducing agents (Table 2).

The purified enzyme digested vicilin and legumin from mature vetch seeds (Fig. 2). Legumin a chains were rapidly hydrolyzed during 60 min incubation. Two intermediates, of about 31 kDa and 33 kDa, and several smaller proteolysis prod- ucts appeared. The enzyme did not degrade legumin chains in a 60-min incubation. During hydrolysis of vicilin, intermediates with apparent molecular masses of about 47, 28 and 22kDa

were detectable if the incubation times were short (Fig. 2), whereas further degradation took place during longer experi- ments (not shown).

Molecular characterization of proteinase A. Six full-length cDNA clones obtained from germinating seeds were sequenced in both directions. They all exhibited the same nucleotide se- quences. The full-length cDNA sequence of pSK19 (1284 bp), the clone with the longest 5’ untranslated region, is shown in Fig. 3. It contains an ORF of 1079 bp that encodes a polypeptide of 359 amino acids (40.6 kDa). Three putative secretory signal sequences were detected by means of PC/GENE (subprogram PSIGNAL). They all meet the (-3;-1) rule (von Heijne, 1983). The putative cleavage site with the highest cut-off value is be- tween Thrl8 and Va119. A propeptide should be located between the putative C-terminus of the signal peptide and the N-terminal amino acid of the mature polypeptide which was determined to be Val. The estimated molecular mass of proproteinase A, that is the precursor without a signal peptide is 38520 Da, whereas mature proteinase A has a molecular mass of 25244 Da. The amino acid sequences obtained from N-terminal peptide se- quencing of cleavage products of the purified enzyme agree to the corresponding sequences in the cDNA-derived amino acid sequence (Fig. 3). Mature proteinase A is probably not glyco- sylated since a potential N-glycosylation site was only present in the putative propeptide.

The EMBL Data Library (release 49) and Swiss-Prot Data- base (release 34) revealed 78.2% sequence identity of proteinase A to cysteine endopeptidase SH-EP, which is one of the major proteinases formed in cotyledons of Kgna mungo seedlings (Akasofu et al., 1989). The cDNA-derived amino acid sequence of SH-EP contains the tetrapeptide Lys-Asp-Glu-Leu (KDEL) at its C terminus (Okamoto et al., 1994). The KDEL tetrapeptide was found in the derived amino acid sequences of all proteinase A cDNA clones.

Southern blot analysis of genomic DNA revealed, under stringent conditions, two restriction fragments of similar size that strongly hybridized with the proteinase A cDNA probe in all three restriction digests (Fig. 4). In the EcoRI digest two ad- ditional, smaller fragments appeared. According to cDNA analy- ses no BamHI and EcoRV cleavage sites are present in pSK19 cDNA, whereas it contains one EcoRI site. Therefore, in case of only one gene, the probe should recognize only one band in the BamHI and EcoRV lanes but the EcoRI digest should give an additional fragment provided that there are no recognition sites for these restriction enzymes in introns.

Activation of proteinase A produced in E. coli. cDNA se- quences of pSK19 encoding mature proteinase A (estimated mo- lecular mass -25 kDa) or proproteinase A (-38 kDa) were ex- pressed in E. coli HMS174(DE3) under the control of the induc- ible T7 promoter. Recombinant proproteinase A was located in inclusion bodies and in the soluble part of the E. coli cell. The soluble proproteinase A was processed into the mature protein- ase A either by E. coli proteinases or by autocatalysis. The 29-kDa and the 21-kDa forms of recombinant proteinase A dis- played gelatin-degrading activity after electrophoretic separation in SDS gel slabs (Fig. 5 ) . In an insoluble ‘inclusion body’ frac- tion, exclusively recombinant mature proteinase A was located, which could not be reactivated (data not shown). Similar to the plant enzyme it appeared as a double band in the ‘inclusion body’ fraction, suggesting that minimal differences in charge could occur (Fig. 6). Heat-treated recombinant mature proteinase A was detectable both at 29 kDa, which was observed only after

308 Becker et al. (EuI: J . Biochem. 248)

CCAATACTTCCCAACAAAATGGAAATGAAAAAGCTCTTGTTTATTTCCTTATCCCTTGCTTTAATTTTCACAGT~GCCAACACCT~TGAT 90 H E H K K L L F I S L S L A L I F T V A N T F O 24

t it) c t ) TTCAACGAACATGATTTAGAGTCAGAGAAAAGTTTGTGGAATTTATACGAAAGATGGAGAAGTCACCACACAGTTACTCGAAATCTTGAT 180

F N E H D L E S E K S L W N L Y E R W R S H H T V T R N L D 54

GAGAAACATAATCGCTTCAACGTGTTCAAAGCCAATGTTATGCATGTTCACAATACCAATAAATTGGATAAGCCTTATAAGCTTAAGTTG 270 E K H N R F N V F K A N V M H V H N T N K L D K P Y K L K L 84

AACAAGTTTGGTGACATGACTAATTATGAATTTAGAAGAATATATGCTGATTCGAAGATTAGTCATCATAGAATGTTTCGTGGTATGTCA 360 N K F G D H T N V E F R R I V A D S K I S H H R M F R G M S 114

CATGAGAATGGAACATTCATGTACGAGAATGCGGTAGATGTTCCTTCTTCGATCGATTGGAGAAATAAAGGTGCTGTTACTGGCGTAAAA 450 H E N G T F H Y E N A V D V P S S I D W R N K G A V T G V K 144

GATCAAGGCCAATGTGGTAGTTGTTGGGCGTTTTCAACTATTGCAGCCGTGGAAGGTATTAACCAAATAAAAACACAAAAGCTAGTTTCA 540 D O G O C G S C W A F S T I A A V E G I N O I K T O K L V S 174

TTATCTGAACAACAACTTGTTGATTGTGACACTGAAGAAAATGAAGGATGCAATGGTGGGTTGATGGAATATGCGTTTGAGTTCATCAAA 630 L S E O O L V D C D T E E N E G C N G G L M E Y A F E F I K 204

CAAAATGGTATAACAACTGAAAGCAATTATCCTTATGCTGCAAAAGATGGAACTTGTGATGTAGAAAAGGAGGATAAGGCAGTCTCAATT 720 O N G I T T E 5 N Y P Y A A K D G T C D V E K E D K A V 5 I 234

GATGGGCATGAAAATGTGCCTATAAATAATGAAGCTGCATTGCTGAAAGCTGCTGCTAAACAACCTGTTTCTGTAGCAATTGATGCTGGT 810 D G H E N V P I N N E A A L L K A A A K O P V S Y A I D A G 264

GGATATAATTTCCAGTTCTATTCTGAGGGAGTATTTACTGGACACTGTGACACTGATCTGAATCATGGGGTAGCAATTGTTGGCTATGGT 900 G Y N E O F Y S E G V F T G H C D T D L N H G V A I V G Y G 294

GTAACTCAAGATAGAACAAAATATTGGATAGTGAAGAATTCATGGGGAAGTGAATGGGGAGAACAAGGTTACATTAGAATGCAAAG~GGC 990

t

V T O D R T K Y W I V K N S W G S E W G E O G Y I R H I I R G 324 99 It1 98

ATATCTTCTAGGGAAGGTCTA~GTGGCATAGCTATGGAAGCTTCCTATCCAATCAAAAAATCATCTACCAAGCCAACAGAAAGTTCTATT 1080 I S S R E G L C G I A H E A S Y P I K K S S I K P T E S S I 354

< t .. ..... ..43.. ..... .

C~~AAAGATGAACTTTAATGATGCCATTGCCACACAAAAAGTTGTATTGTATAGTGCTTCATAATCATTTGTTTATGTAACCAGGAGGTT 1170 L K D E L 359

CTTATGCTAGGATTATGTCCAAAAAAAAAAAAAAAATAATGTGTATCTCAATTCGAAGTTATAAGTTGCTTGTATTTTTCTTTTTCCATTG 1260 GGAATAGAAATTTATGCTTATTTTpolyA. 1284

Fig. 3. Nucleotide and deduced amino acid sequences of the full-length proteinase A cl1NA clone pSK19. Horizontal arrows indicate primer positions for primary amplification of proteinase A cDNA. Dashed lines indicate primer positions for amplification of C-terminal variants of proteinase A. 43, 98, 99, 11 1 indicate oligonucleotide primers used for amplification of proteinase A variants. Sequenced peptide fragments of the purified protein are underlined. The putative cleavage site between signal peptide and propeptide, the cleavage site between propeptide and the mature polypeptide, and the estimated C-terminal cleavage site are indicated by vertical arrows. The less likely cleavage sites for elimination of a signal peptide are indicated in brackets. The potential glycosylated site in the propeptide is labeled by a dot.

Fig.4. Southern blot analysis of vetch genomic DNA. DNA (10 pg/ Fig.5. SDSPAGE of proproteinase A synthesized in E. coli well) was digested with EcaRI, EcoRV or BanzHI and then fractionated HMS174(DE3). The gel contained 0.05% gelatin and was incubated on 0.8% agarose gels. Separated DNA was transferred onto a nylone after electrophoresis in Mcllvaine buffer, pH 5.6, 180 mM NaCI, 2 mM membrane (Nytran-N+, Schleicher B Schiill) and probed with the "P- dithiothreitol. Lanes 1 and 2, total E. cnli cell extracts; lanes 3 and 4, labelled insert of pSK19. The exposure time was 4 days at -70°C with soluble components of the E. coli extracts. Lane 1, uninduced control; an intensifying screen. The positions of Eo,RIIHindlIl-digested lambda lane 2, proproteinase A ; lane 3, uninduced control ; lane 4, soluble DNA size markers are indicated. proproteinase A.

1 2 3 4 5 6 kDa

Becker et al. (ELM J. Biochem. 248)

E E E E Z $

309

(I) c 0

46.0-

30.0-

21.5-

-4 -9 -12 amino acids

Fig. 6. Immunohlot of different proteinase A variants (lanes 1-4) synthesized in E. coli HMS174(DE3) in comparison with purified plant proteinase A (lanes 5, 6). The insoluble 'inclusion bodies' frac- tions were separated from the soluble E. coli parts by centrifugation and lysed in 50 mM Tris/HCI, pH 8.0, 10 mM EDTA. After electrophoresis the proteins were blotted onto nitrocellulose membrane and immuno- stained as described in Fig. 8. -4, -9 and -12, proteinase A without 4, 9 and 12 C-terminal amino acids, respectively.

heat treatment, and at 21 kDa, which represents the enzymati- cally active form of proteinase A from plants (Fig. 6).

Does mature proteinase A have a KDEL sequence? Four pro- teinase A variants encoding the mature enzyme but differing in the length of their C-terminal regions were produced in E. coli to determine whether plant proteinase A has a KDEL sequence. Fig. 6 shows an immunoblot of the recombinant proteins in com- parison with purified proteinase A. The purified plant enzyme migrated with the same mobility in SDS gel slabs as recombi- nant proteinase A that lacks 12 C-terminal amino acids. This result was confirmed by N-terminal peptide sequencing of a CNBr-cleavage product of purified plant proteinase A generated at Met336. It ends C-terminally with Thr347 (Fig. 3).

Developmental regulation and compartmentation of protein- ase A gene expression. To analyze the expression pattern of the proteinase A genes in different organs of the seed during development and germination and in other parts of the plant, the cDNA insert of pSK19 was used as a probe in Northern blot experiments. Fig. 7 shows that specific mRNA was found only in cotyledons of germinating seeds. There, the mRNA became

A proteinase A

germination

cotyledons , , embryo, % f r

Oh 12 24 72 6d Oh12 24

Fig. 7. Northern blot analysis of the proteinase A transcript during different stages of seed development and in other parts of the plant. Total RNA (10 @well) was denatured with glyoxal (Sambrook et al., 1989) and separated on 1 % agarose gels. The RNA was transferred onto a nylon membrane and hybridized with the "P-labelled insert of pSK19. The exposure time was 3 days at -70°C with an intensifying screen.

develop- rnent germination

kDa 7 1 7 1

97.4-

69.0-

46.0-

30.0-

21.7-

4 precursor

4 mature proteinase A

Fig. 8. Immunoblot analysis of vetch cotyledon extracts. Cotyledons (500 mg/sample) were homogenized in 100 mM Tris/HCI, pH 7.8, 150 mM NaCI, centrifuged, and equal amounts of the total extracts were loaded on SDS/PAGE. After electrophoresis the proteins were blotted electrophoretically onto a nitrocellulose membrane. An antibody raised against purified proteinase A was used at a dilution of 1 :10000. 3- 5 mm, cotyledon length; DAI, days after germination.

B proteinase 6

1 2 3 4 5 6 7 8 9 kDa 1 2 3 4 5 6 7 8 9 kDa

97.4 --

69.0 - 97.4-

69.0-

46.0 - 46.0 -

30.0 - 30.0-

21.5 - 21.5-

I 1 2 u L~__.- I 5mm Od I d 46 5mm Od I d 46

supernatant protein bodies

5mm Od I d 46 5mm Od I d 4d

supernatant protein bodies

Fig. 9. Immunoblot analysis of fractions obtained after Ficoll-density-gradient centrifugation. (A) Reacted with anti-(proteinase A) Ig raised against the recombinant enzyme, diluted 1: 10000; (B) reacted with anti-(proteinase B) Ig, diluted, 1 :300000. Cotyledons from developing (5 mm) or mature seeds (Od), or after germination for 1 day (Id) or 4 days (4d) were used as starting material. Lanes 1-4, supernatants after centrifugation through 5% Ficoll; lanes 5-8, protein-body fractions. Lane 9 was a cotyledon extract 6 days after germination.

310 Becker et al. ( E M J. Biochem. 248)

first detectable at day 1 after imbibition and reached a maximum at 6-8 days after imbibition (data not shown).

Immunoblot analysis of total cotyledon extracts of different developmental and germination stages revealed that mature pro- teinase A (29 kDa) appeared in the cotyledons at day 6 after imbibition (Fig. 8). The antibody detected additional protein bands of aproximately 41, 3.5 and 33 kDa that might represent precursor molecules and processing products. The 41-kDa bands, presumably representing precursor molecules, were de- tectable already from 3 days after imbibition. The data indicate that proteinase A is synthesized de novo during germination. It does not belong to the early-expressed cysteine-proteinase genes, such as cysteine proteinases 1 and 2 (Becker et al., 1994; Fischer et a]., unpublished results).

Ficoll-density-gradient centrifugation was used to isolate protein bodies from cotyledons of different developmental stages. The quality of the isolation procedure was checked by protein determination, SDS/PAGE and immunoblotting with an anti-vicilin Ig (not shown). More than 80% of the total protein could be detected in the protein-body fraction. No mature pro- teinase A (29 kDa) was found in the protein-body fractions dur- ing all developmental stages investigated (Fig. 9). Only an un- known cross-reacting 33-kDa band appeared in the supernatant of all developmental stages. In contrast, proteinase B, an Asn- specific cysteine proteinase (Becker et al., 1995) was located exclusively inside the protein bodies.

DISCUSSION

Purified proteinase A and proteinase-A-specific cDNA were used as tools to verify the function that Shutov and Vaintraub (1987) attributed to this enzyme for globulin degradation in cotyledons of germinating vetch and related legume seeds.

Proteinase A is the most abundant proteinase in cotyledons of day-8 vetch seedlings. For purification of the enzyme the iso- lation procedure described by Shutov et al. (1984) was used as a guide. The first three steps effectively removed other protein- ases from the proteinase A extract. For further purification, chro- matography on another ion-exchange material followed by hy- drophobic-interaction chromatography were used. These steps efficiently removed contaminating proteins. The final purifica- tion step was gel filtration on Sephadex G-50, as used earlier (Shutov et al., 1984). Similar to the enzyme prepared previously, proteinase A prepared here was retarded on the column material and eluted by 37 mM sodium phosphate, pH 6.5. The applica- tion of a very similar purification scheme and that both enzymes show identical characteristics and activity patterns when acting on casein or storage proteins allowed us to conclude that we purified essentially the same enzyme as Shutov et al. (1984).

A small family with one or two genes encodes proteinase A. We found only one cDNA. From the one complete proteinase- A-specific cDNA that we analyzed in detail, the amino acid se- quence of 40600-Da preproproteinase A was derived. Its N ter- minus corresponds to an 18-amino-acid signal sequence, which indicates that proteinase A is nuclear encoded and that during its biosynthesis the precursor enters cotranslationally the secretory pathway. Proproteinase A which should be generated by signal- sequence detachment i n the endoplasmic reticulum is composed of an N-terminal 109-amino-acid propeptide followed by the re- gion that encodes mature proteinase A. The derived amino acid scquence of the proteinase A precursor ends C-terminally with the endoplasmic reticulum-retention signal KDEL. Proteinase A can be classified as a papain-like cysteine endopeptidase accord-

ing to sequence similarity and to the conserved positions of amino acids forming the active center in papain. The sequence of proteinase A is most similar to that of the pa- pain-like cysteine endopeptidase SH-EP from germinating mung bean, V mungo L. (Akasofu et al., 1989), including the presence of the KDEL signal at the C terminus of its precursor.

From the cDNA-derived amino acid sequence, a molecular mass of 2.5244 Da was calculated for mature proteinase A. As indicated by partial sequencing of the N-terminal and several internal polypeptide fragments, the proteinase A preparation can be assigned to the sequenced cDNA clone. On SDS gels, bands of the purified enzyme preparation occupied positions that cor- responded to relative molecular masses of 29 kDa and 21 kDa, depending on whether the protein underwent heat denaturation or not. Since the proteolytically active 21-kDa bands of non- heated samples could be transformed into the inactive 29-kDa bands by heat treatment before re-electrophoresis, the 29-kDa polypeptides should represent a denatured conformation, whereas the 21-kDa bands should at least have maintained a conformation that permits enzyme activity. Contrary to that, the 29-kDa proteinase A polypeptide generated in the soluble frac- tion of E. coli cells that expressed the proproteinase A cDNA should represent a processing intermediate and/or molecular conformation of proteinase A that permits enzymatic activity. Its conformation should differ from that of the 21-kDa band, which seems to represent active mature proteinase A similar to that prepared from cotyledons of germinating seeds. Recombinant proteinase A formed a double band in SDS gel slabs, indicating that the double band that forms in plant proteinase A prepara- tions on SDS gels cannot be explained by different glycosylation or expression of different genes. Since both chains agree in the sequenced N-terminal fragments they might differ in chain length at the C terminus or in charge. Different C termini can be excluded since we expressed four proteinase A variants in E. coli that had different C termini, and they all formed double bands.

According to publications from Minamikawa’s group (Mit- suhashi and Minamikawa, 1989; Okamoto and Minamikawa, 199.5), in germinating mung bean the SH-EP propolypeptide is step-wisely transformed into the mature enzyme by proteolytic processing accompanied by the generation of several polypep- tides of intermediate length. Some of these processing steps were attributed to the action of a proteinase-B-like processing enzyme. At least the conversion of a 36 kDa intermediate into the mature 33 kDa enzyme is assumed to proceed autocatalyt- ically. In the soluble fraction of E. coli cells proproteinase A is converted into active mature proteinase A due to the action of endogenous 1;. coli proteinases or to autocatalytic processing. If it is autocatalyis then proproteinase A can activate itself without participation of proteinase-B-like enzymes. On western blots of total cotyledon extracts, proteinase-A-reactive bands with rela- tive molecular masses in the range of precursor molecules and processing intermediates were found between day 3 and 8 of germination. Although the proteinase A precursor could be de- tected 3 days after imbibition immunoreactive bands corre- sponding to the mature enzyme appear only 6-7 days after in- bibition. This suggests an inhibition of precursor protein matura- tion or/and a delayed transport of the precursor into the vacuolar compartment where final activation of proteinase A should take place. Among the fragments of the proteinase A polypeptide that were partially sequenced was a CNBr cleavage product gener- ated at Met24, upstream from the C terminus. Its sequence ends at the Thr 13 residues from the C terminus. To confirm this result we produced proteinase A variants in E. coli that differed only in the length of their C termini. Purified plant proteinase A migrated with a mobility in SDS gels similar to that of recornbi-

Becker et al. (Eur: J. Biochem. 248) 31 1

nant proteinase A that lacks 12 C-terminal amino acids. This strongly suggests that the KDEL residues and eight additional amino acids are detached posttranslationally by a C-terminal processing step. The C terminus of mature SH-EP has been shown to be 10 amino acids shorter than the corresponding cDNA-derived presursor sequence (Okamoto et al., 1994). This justifies to some degree the assumption that the two enzyme precursors undergo similar N-terminal and C-terminal process- ing. Mature proteinase A appeared only at day 6 of germination, although the proteinase A precursor was found in total cotyledon extracts already at day 3 and the corresponding mRNA was al- ready detected on day l . This time gap might be explained by delayed starting of translation or initial instability of translation products or by retarded precursor processing. We suppose that KDEL may function in the retardation of proteinase A transport to the protein bodies to guarantee a controlled degradation of storage proteins and, consequently, a controlled flux of amino acids into the growing embryo axis.

During germination, protein bodies are transformed into vac- uoles. Therefore, the original protein-body fraction can only be isolated from these cotyledons until day 4 of germination, which is before mature proteinase can be detected. This explains the difficulty to localize the mature enzyme by cell fractionation in the vacuolar compartment where it is expected due to the pres- ence of a signal peptide in its precursor.

According to its molecular mass, isoelectric point, inhibitor sensitivity and ability to degrade globulins in vitro, proteinase A is not only similar to SH-EP from mung bean but also to cysteine proteinases that have been found previously in other legumes (Baumgartner and Chrispeels, 1977 ; Boylan and Sus- sex, 1987; Khembavi et al., 1993; Yu and Greenwood, 1994). Some of them have been shown to be synthesized during germi- nation (Chrispeels et al., 1976; Mitsuhashi and Minamikawa, 1989), whereas in moth bean the proteinase vignain was detect- able in mature seeds (Khembavi et al., 1993). But in nearly all cases the corresponding increases in proteolytic activity oc- curred after globulin degradation had started. Therefore, no conclusive evidence is available that globulin degradation is generally initiated by de novo-formed cysteine endopeptidases (Muntz, 1996).

Shutov and Vaintraub (1987) listed three criteria that protein- ases have to meet before they can be taken into account for playing an essential role in storage globulin degradation : they have to be present with their natural substrates in similar cellular compartments ; the time-course of enzyme activity must coincide with that of storage-protein degradation; and they have to de- grade their natural substrates in vitro. Only the last criterion was met by proteinase A. Although this enzyme could not be local- ized in the protein bodies, our results strongly suggest a partici- pitation of proteinase A in storage-protein hydrolysis during later stages of germination. However, the enzyme cannot be looked upon as the proteinase that initiates storage-globulin de- gradation since it is not present during early stages of germina- tion when globulins already are degraded. In addition, one has to expect that a triggering proteinase is not very active on stor- age proteins and has narrow rather than broad substrate specific- ity. New candidates have to be sought among those cysteine proteinases known to appear much earlier during vetch seed ger- mination (Becker et al., 1994) or among those that still remain to be discovered.

This work was supported by Deutsche Forschiing.sgemeinschaft grants Mu92515-1, 436 Mol-17/4/93, 436 Mol-17/3/93 and SFB 363. We thank Monika Wiesner, for skilled technical assistance, Susanne Kiinig, for performing sequence analyses on the automated DNA sequencer, and Rose-Marie Gillandt, for synthesis of oligonucleotides.

REFERENCES Akasofu, H., Yamauchi, D., Mitsuhasi, W. & Minamikawa, T. (1989)

Nucleotide sequence of cDNA for sulfl-rydryl-endopeptidase (SH-EP) from cotyledons of germinating Wgna miingo seeds, Nucleic Acids Rex 17, 6733.

Baumgartner, B. & Chrispeels, M. J. (1977) Purification and character- ization of vicilin peptidohydrolase, the major endopeptidaae in the cotyledons of mung-bean seedlings, Eur. J. Bioclwni. 77, 223- 233.

Becker, C., Fischer, J., Nong, V. H. & Muntz, K. (1994) PCR cloning and expression analyses of cDNAs encoding cysteine proteinases from germinating seeds of Wcili sufiva L., Plunt Mol. B id . 26, 1207 - 1212.

Becker, C., Shutov. A. D., Nong, V. H,. Senyuk, V. 1., Jung, R., Horst- mann, C., Fischer, J., Nielsen, N. C. & Muntz, K. (19%) Purification, cDNA cloning and characterization of proteinase B, an asparagine- specific endopeptidase from germinating vetch (Viciu sutivcr L.) seeds, Eur. J. Biochem. 228, 456-462.

Boylan, M. T. & Sussex, I. M.(1987) Purification of an endopeptidase involved with storage-protein degradation in Phuseotus vulgaris L. cotyledons, Pluntu (Heidelb.) 170, 343-352.

Chrispeels, M. J., Baumgartner, B. & Harris, N. (1976) Regulatioii of reserve protein metabolism in the cotyledons of mung bean seed- lings, Pmc. Naft Acud. Sci. USA 73, 3168-3172.

Dunaevsky, Y. E. & Belozersky, M. A.(l989) Proteolysis of the main storage protein of buckwheat seeds at the early stage of germination, Physiol. Plant. 75, 424-428.

Elpidina, E. N.: Voskoboynikova, N. E., Belozersky, M. A. & Dunaev- sky, Y. E. (1991) Localisation of a metalloproteinase and its inhibitor in the protein bodies of buckwheat seeds, Plurzta (Heide/b.) Ih 'S , 46- 52.

Laemmli, U. K. (1970) Cleavage of structural proteins during the assem- bly of the head of bacteriophage T4, Nature 227, 680-685.

Lichtenfeld, C., Manteuffel, R., Muntz, K., Neumann, D., Scholz, G. & Weber, E. (1979) Protein degradation and proteolytic activities in germinating field beans, Bioclzern. Physiol. Pjanz. (BPP) 174, 255 - 274.

Lichtenfeld, C. (1981) Aktivitutswechsel, Fiinktion urzd nutiirliche Sub- .strate proteolytischer Enzyme in keimenden Ackerbohnen.sameri (Ificiajuhu L.), PhD Thesis, University of Halle.

Mader, M. & Chrispeels, M. J. (1984) Synthesis of an integral protein of the protein-body membrane in Phaseolus vulgaris cotyledons, Plunta

Miller, S. H., Dykes, D. D. & Polesky, H. F. (1988) A simple salting out procedure for extracting DNA from human nucleated cells, Nu- cleic Acids Res. 16, 1215.

Mitsuhasi, W. & Minamikawa, T. (1989) Synthesis and posttranslational activation of sulphydryl-endopeptidase in cotyledons of germinahg Vigna mungo seeds, Plant Physiol. (Rockv.) 89. 274-279.

Muntz, K., Bassuner, R., Lichtenfeld, C., Scholz, G. & Weber, E. (1985) Proteolytic cleavage of storage proteins during embryogenesis and germination of legume seeds, Physiol. Veg. 23, 75-79.

Muntz, K. (1996) Proteases and proteolytic cleavage of storage proteins in developing and germinating dicotyledonous seeds. J . Exp. Boi. 47,

Okamoto, T., Nakayama, H., Seta, K., Isobe, T. & Minamikawa, T. (1 994) Posttranslational processing of a carboxy-terminal propeptide containing a KDEL sequence of plant vacuolar cysteine endopepti- dase (SH-EP), FEBS Left. 351. 31 -34.

Okamoto, T. & Minamikawa, T. (1995) Purification of a processing en- zyme (VMPE-1) that is involved in post-translational processing of a plant cysteine endopeptidase (SH-EP), Eur. J. Biochenz. 231, 300- 305.

Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989) Molec,~ilur cloning: n laboratory manual, 2nd edn, Cold Spring Harbor Laboratory, Cold Spring Harbor NY.

Shutov, A. D. & Vaintraub, 1. A. (1973) Primary changes of reserve proteins during germination of vetch seeds, Fiziol. Rust. (Moscow) 20, 504-509.

Shutov, A. D., Lanh, Do N. & Vaintraub, I. A. (1982) Purification and partial characterization of protease B froin germinating vetch seeds. Biokhirniya 475, 814-821.

(Heidelb.) 160, 330-340.

605 -622.

312 Becker et al. (EuI: J . Biochem. 248)

Shutov, A. D., Bulmaga, V. P. & Vaintraub, I. A. (1984) Protease A from germinating vetch seeds: purification by affinity chromatography

179, 191-196.

in germinating seeds, Phyrochernistry (O$) 26, 1557- 1566.

Van Heijne, G. (1983) Patterns of amino acids near signal sequence cleavage sites, EUI: J. Biochem. 133, 17-21.

of a cysteine proteinase involved in globulin hydrolysation in germi- nated xcja,faba L,, J, E ~ ~ , B ~ ) ~ , 45, 261 -268,

2nd substrate specificity studies, Binche/?t Pkysiol. ff2an:. (BPP) yu, W. J. & Greenwood, J. S. (1994) Purification and characterjzation

Shutov. A. D. & Vaintraub, I. A. (1987) Degradation of storage proteins