Proteinase lnhibitors in Nicotiana alata Stigmas Are Derived from a ...

The Plant Cell, Vol. 5,- 203-213, February 1993 O 1993 American Society of Plant Physiologists

Proteinase lnhibitors in Nicotiana alata Stigmas Are Derived from a Precursor Protein Which 1s Processed into Five Homologous lnhibitors

Angela H. Atkinson,a Robyn L. Heath,a9b Richard J. Simpson,c Adrienne E. Clarke,a and Marilyn A. Andersonai’ a Plant Cell Biology Research Centre, School of Botany, University of Melbourne, Parkville, Victoria 3052, Australia

lnstitute of Plant Sciences, Department of Agriculture, Burnley, Victoria 3121, Australia Joint Protein Structure Laboratory, Ludwig lnstitute for Cancer Research and the Walter and Eliza Hall lnstitute for

Medical Research, P.0. Royal Melbourne Hospital, Parkville, Victoria 3050, Australia

A cDNA clone, NA-PI-II, encoding a protein with partia1 identity to proteinase inhibitor (Pl) II of potato and tomato has been isolated from a cDNA library constructed from Nicotiana alata stigma and style mRNA. The cDNA encodes a polypeptide of 397 amino acids with a putative signal peptide of 29 amino acids and six repeated domains, each with a potential reactive site. Domains 1 and 2 have chymotrypsin-specific sites and domains 3,4,5, and 6 have sites specific for trypsin. In situ hybridization experiments demonstrated that expression of the gene is restricted to the stigma of both immature and mature pistils. Peptides with inhibitory activity toward chymotrypsin and trypsin have been isolated from stigmas of N. alata. The N-terminal amino acid sequence obtained from this protein preparation corresponds to six regions in the cDNA clone NA-PI-II. The purified PI protein preparation is likely to be composed of a mixture of up to five similar peptides of -6 kD, produced in vivo by proteolytic processing of a 42-kD precursor. The PI may function to protect the reproductive tissue against potential pathogens.

INTRODUCTION

Severa1 members of the families Solanaceae and Fabaceae accumulate serine proteinase inhibitors (Pls) in their storage organs and in leaves in response to wounding (Richardson, 1977; Brown and Ryan, 1984). The inhibitory activities of these proteins are directed against a wide range of proteinases of microbial and animal origin but rarely against plant proteinases (Richardson, 1977). It is believed that these inhibitors are in- volved in protection of the plants against predators. In potato tubers and legume seeds, the inhibitors can comprise 10% or more of the stored proteins (Richardson, 1977), whereas in leaves of tomato and potato (Green and Ryan, 1972) and alfalfa (Brown and Ryan, 1984), Pls can accumulate to levels of 2% of the soluble protein within 48 hr of insect attack or other types of wounding (Brown and Ryan, 1984; Graham et al., 1986). High levels of these inhibitors (up to 50% of total soluble protein) are also present in unripe fruits of the wild tomato Lycopersicon peruvianum (Pearce et al., 1988).

There are two families of serine Pls in tomato and potato (Ryan, 1984). Type I inhibitors are small proteins (monomer Mr 8100) that inhibit chymotrypsin at a single reactive site (Melville and Ryan, 1970; Plunkett et al., 1982). lnhibitors of

To whom correspondence should be addressed.

the type II family contain two reactive sites, one of which in- hibits chymotrypsin and the other trypsin (Bryant et al., 1976; Plunkett et al., 1982). The type II inhibitors have a monomer Mr of 12,300 (Plunkett et al., 1982). PI I accumulates in etiolated tobacco (Nicofiana fabacum) leaves (Kuo et al., 1984), and elicitors from Phytophfhora parasifica var nicofianae were found to induce PI I accumulation in tobacco cell suspension cultures (Rickauer et al., 1989).

Here we report the isolation of a cDNA clone encoding a protein with sequence similarity to a PI II of potato and tomato. The cDNA is derived from mRNA expressed in the stigma of an ornamental tobacco, N. alata. We have also isolated peptide inhibitors with homology to six regions of the amino acid sequence deduced from the cDNA clone.

RESULTS

lsolation and Characterization of the PI cDNA Clone

A cDNA library prepared from mRNA isolated from the stigmas and styles of mature flowers of N. alata was screened for clones of highly expressed genes that were not associated with the

204 The Plant Cell

self-incompatibility genotype. The most abundant clones (comprising 3% of the library) encoded a protein with sequence identity to the type II Pls from potato and tomato (Graham et al., 1985; Thornburg et al., 1987). The largest clone, NA-PI-II, is 1359 bp long with an open reading frame of 1191 nucleotides. Comparison of the predicted amino acid sequence of the N. alata clone NA-PI-II with the type II inhibitors of potato (Sanchez-Serrano et al., 1986; Thornburg et al., 1987) and tomato (Graham et al., 1985) is shown in Figure 1. Of the first

o

90 100 120 Na PI I N D R I m ” m A l ~ T l ~ K Tom PI s L I Y P T G c T T c c T o I r G c r POt PI IIR 5 L I r P T c c T T c c T G x R c c I POt PI 11 5 L I Y P T c c T T c c T G Y K c c Y

130 140 150 160 Na PI C O P R I ~ Y G I C P L I E E r r N D R ~ C ~ ” ~ ~ ~ ~ ~ ~ G C ~ ~ ~ S Tom PI POZ PI IIR eot PI II

170 180 190 200 Na PI D D c T F ” c E c E 5 D P R N P K * c P C D C R I I P C I C P L S E E K K ”

2 1 0 220 230

2 5 0 2 6 0 270 280 N a P I C O G I I ~ I C I C P L S E E ~ ~ N D R ~ ~ ~ * = = ~ = ~ ~ ~ = ~ ~ ~ ~ ~ ~ ~ ~

N a s 1 D i l C T N C C I C I ~ C C r r F S D D C T I V C E O E S D P i N P I I C P

290 300 3 1 0 120 C D G I I I P C I C P L S E E I I N -

n a e I T N C C I G R R C C R ~ F S D D G T F ~ C T C E S E Y A S R Y D ~ I Y G E V E N

370 N S i I O L Y R S R V I V S

Na PI F Y c E G E 5 D P R N P K A c P

330 3 4 0 350 360

Figure 1. Comparison of Deduced Amino Acid Sequence of the A! alata PI Protein with the PI II Proteins of Tomato and Potato, and the Small Pls from Potato Tubers, PCI and PTI, and Eggplant

The amino acid sequence is numbered beginning with 1 for the first amino acid of the mature protein (based on the potato and tomato sequences) The signal sequence is indicated by negative numbers Common sequences are boxed Gaps were introduced at positions -26, -20,86, and 87 in the NA-PI-II sequence, at positions -5 to -10 in the tomato sequence, position -4 in the potato IIK sequence, and at position 10 in the tomato and both potato PI II sequences to max- imize similarity. The sequence of the tomato PI was from Graham et al (1985), the potato PI IIK was from Thornburg et al. (1987), and the potato PI II was from Sanchez-Serrano et al (1986) The sequences of the peptide inhibitors PCI-I and PTI were from Hass et al. (1982), and the eggplant inhibitor sequence was from Richardson (1979). The residues underlined in the N alata sequence correspond to the N-terminal sequence of the 6-kD PI preparation, purified from N alata stigmas The reactive site residues of the inhibitors are shaded

The N alata PI sequence contains six similar domains domain 1, residues 1 to 58, domain 2, residues 59 to 118, domain 3, residues 119 to 176; domain 4, residues 177 to 234; domain 5, residues 235 to 292; domain six, residues 293 to 350 The Pls from tomato and potato are composed of two domains domain 1, residues 1 to 58, domain 2, residues 59 to 118.

151 amino acids of the N. alata PI, there is 53% identity with the potato PI IIK, 56% identity with potato PI II, and 50% identity with the tomato PI. All of the cysteine residues present in the potato and tomato sequences are also conserved in the N. alata sequence. There are no potential N-glycosylation sites.

In contrast to the potato and tomato Pls, the N. alara cDNA clone encodes a protein with six repeated domains that have high, but not perfect, sequence identity (Figure 1). Each of these domains contains a potential reactive site which is highlighted in Figure 1 and was identified by analogy with the reactive site residues of tomato and potato Pls (Figure 1). The residues at the putative reactive sites of the N. alata PI are consistent with the inhibitor having two sites that would specifically inhibit chymotrypsin (Leud-Asn-6 and Leu-63-Asn-64) and four sites specific for trypsin (Arg-123-Asn-124, Arg-181-Asn-182, Arg- 239-Asn-240, and Arg-297-Asn-298).

To ensure that the repeat structure of NA-PI-II was not due to a cloning artifact, three additional cDNA clones were sequenced and found to be identical to NA-PI-II, although none of the other clones was full length at the 5‘end of the sequence.

Table 1 is a comparison of the percentage amino acid identity of the two domains of the potato and tomato Pls with the six N. alata PI domains. The six domains of the N. alata PI are less divergent from each other than the domains of the other Pls. For example, domains 1 and 2 of the tomato sequence are 50% identical at the amino acid leve1 and domains 1 and 2 of the potato IIK sequence are 52% identical. In N. alata, however, the domains share between 79 and 100% identity (Table 1).

Temporal and Spatial Expression of the PI mRNA

Total RNA isolated from various tissues of N. alata was probed with the PI cDNA clone in the RNA gel blot analyses shown in Figure 2. Two hybridizing messages of 1.0 and 1.4 kb were present in total RNA isolated from styles (including stigmas). Only the larger message, which was predominant in this tissue, is of sufficient size to encode the cDNA clone NA-PI-II (1.4 kb). The smaller message is not detected with the cDNA probe at higher stringency. An homologous message of 4 . 4 kb was also present in RNA isolated from the styles of N. tabacum and N. sylvestris (Figure 2).

In the other floral organs (except pollen), both messages were detectable at low levels; however, the smaller RNA species appeared more abundant. There was no hybridization to pollen RNA. No hybridizing species were evident in leaf RNA, but two species, 1.0 and 1.4 kb, were detected 24 hr after me- chanical wounding. The smaller message (1.0 kb) was more abundant in this case.

In situ hybridization of radiolabeled N. alata PI cDNA to longitudinal sections of styles from immature (1-cm long) buds is shown in Figure 3. RNA homologous to the cDNA clone bound strongly to cells of the stigma and weakly to vascular bundles. No hybridization was detected in the cortical tissue, transmitting tract tissue, or epidermis of the style. The same

Proteinase Inhibitor in N. alata Stigmas 205

Table 1. Amino Acid Identity between the Six Domains of the N. alata PI and the Two Domains of the Inhibitors from Potato and Tomato

N. alata

N. alata

Tomato

Potato UK

Potato II

PI

1

123456

12

12

12

2 3 4 5 6

100 88 88 90 7988 88 90 79

97 95 8698 90

90

TomatoPI II

1

535353555350

2

585853535350

50

PotatoPI UK

1

555557595753

8248

Potato

2

606057575753

5088

52

PI II

1

606057595753

8147

9552

2

606057575753

5088

52100

50

The figures in the table are the percentage of amino acid identity between each PI domain. The amino acids and domains are numbered asin Figure 1 and are illustrated in Figure 7. The six domains in the N. alata PI, encoded by the NA-PI-II clone, share between 79 and 100%sequence identity. In contrast, domains 1 and 2 of the Pis from tomato and potato have less identity (50 to 52%).

pattern of hybridization was observed in mature receptiveflowers (data not shown). Control sections treated withribonuclease A prior to hybridization were not labeled.

Genomic DMA Gel Blot Analysis

The cDNA clone NA-PI-II was used as a probe on the DNAgel blot shown in Figure 4, which contained genomic DNAdigested with either EcoRI or Hindlll. EcoRI produced twohybridizing fragments (11 and 7.8 kb) and Hindlll producedthree large hybridizing fragments (16.6, 13.5, and 10.5 kb).

Distribution of Proteinase Activity in Various Tissuesof N. alata

The inhibition of trypsin and chymotrypsin by crude extractsof various organs of N. alata is shown in Figure 5. Stigma ex-tract was the most effective inhibitor of both trypsin andchymotrypsin. The stigma extracts had up to eight times moreinhibitory activity than sepal extracts and more than 20 timesmore activity than extracts from styles, petals, leaves, andwounded leaves.

Purification of PI from N. alata Stigmas

Stigmas of N. alata were extracted in buffer and the inhibitoryactivity was concentrated by precipitation with 80% w/v am-monium sulphate. The precipitate was redissolved andfractionated by gel filtration on Sephadex G-50. Most of the

protein in the extract eluted early in the profile illustrated inFigures 6A and 6B relative to the PI. Fractions with PI activitywere pooled and applied to an affinity column of chymotrypsin-Sepharose. The PI activity co-eluted with a protein of ~6 kD,which appeared to migrate as a single band on the 20%

10 pg

St St Ov Po Pe Se L L4 L24 Nt Ms Ma10(ig 5pg hs

20

05-IfT

f

Figure 2. Gel Blot Analysis of RNA from Various Organs of N. alata.

Gel blot of RNA isolated from organs of N. alata and from stigmas andstyles of N. tabacum and N. sylvestris, hybridized with the cDNA cloneNA-PI-II. St, stigma and style; Ov, ovaries; Po, pollen; Pe, petals; Se,sepals; L, nonwounded leaves; L4, leaves 4 hr after wounding; L24,leaves 24 hr after wounding; Nt, N. tabacum stigma and style; Ns,N. sylvestris stigma and style; Na hs, W. alata stigma and style washedat higher stringency. The size markers are Hindlll restriction fragmentsof X-DNA.

The NA-PI-II clone hybridized to two mRNA species (1.0 and 1.4 kb).The larger mRNA was predominant in style, whereas the smaller mRNAspecies was more dominant in other tissues. After high stringencywashes, the 1.0-kb mRNA from stigma and style no longer hybridizedto the Na-PI-ll probe.

206 The Plant Cell

Figure 3. In Situ Localization of RNA Homologous to NA-PI-II in Stigmaand Style.

(A) Autoradiograph of a longitudinal cryosection through the stigmaand style of a 1-cm long bud after hybridization with the 32P-labeledNA-PI-II cDNA probe.(B) The same section as (A) stained with toluidine blue, c, cortex; v,vascular bundles; tt, transmitting tract; s, stigmatic tissue.The cDNA probe labeled the cells of the stigma heavily, and somehybridization to the vascular bundles can be seen. There was no hy-bridization to the epidermis, cortical tissue, or transmitting tissue.Bar = 200 urn.

SDS-polyacrylamide gel shown in Figure 6C. The purity ofthe PI at various stages was assessed by SDS-PAGE (Figure6C). The purified inhibitor represented ~50°/o of the inhibitoryactivity present in the crude extract (data not shown).

Amino Acid Sequence of the N Terminus of the6-kD PI Protein

at this position and the other four peptides contain lysine. Theposition of these peptides relative to the six repeated domainsin the predicted precursor protein is illustrated in Figure 7A.Five of the six predicted 6-kD peptides contain reactive sitesfor either chymotrypsin or trypsin (Figures 1 and 7). The sixthpotential peptide is four amino acids shorter than the otherfive peptides (58 amino acids) and may not be active becauseit does not contain an inhibitory site. The peptide from the Nterminus (x in Figure 7A) has a potential chymotrypsin reac-tive site but is much shorter (24 amino acids). Figure 1compares the predicted N. alata PI sequence with the se-quences of the potato and tomato type II PI proteins, two smallPis isolated from potato tubers, and a small PI from eggplant.The sequences of the small Pis from potato and eggplant be-gin in the same region as the first DRICTNCCAG(T/K)KGsequence and terminate just before the second such sequence.

Distribution of the PI Protein in N. alata

A polyclonal antiserum was raised to the purified PI proteinconjugated to keyhole limpet hemocyanin. The antibodyreacted strongly with the purified 6-kD PI protein in immuno-blot analyses and bound only to a 6- and a 32-kD protein, whichappears as a doublet, in total stigma and style extracts frommature flowers. Figure 8C shows an immunoblot containingprotein extracts of stigmas from flowers at different stages ofdevelopment (1-cm-long buds to mature flowers) probed withthe anti-Pi antiserum. Larger cross-reacting proteins of ~18and 42 kD were detected in buds from 1 to 5 cm in length inaddition to the 6- and 32-kD proteins. The 18- and 42-kD pro-teins decreased in concentration with maturity, whereas the

EcoRI Hindlll

94—

6-5—

4-3—

2-3—20—

I

The N-terminal amino acid sequence DRICTNCCAG(T/K)KGwas obtained from the purified PI protein. This sequence ofamino acids corresponds to six regions in the deduced se-quence of the cDNA clone, starting at positions 25, 83, 143,201,259, and 317 in Figure 1. At position 11 of the N-terminalsequence, both threonine and lysine were detected. This isconsistent with the purified inhibitor comprising a mixture offive or six peptides beginning with the sequences underlii edin Figure 1, because the first two peptides contain threonine

Figure 4. Gel Blot Analysis of Genomic DNA of N. alata

Gel blot analysis of N. alata genomic DNA digested with the restric-tion enzymes EcoRI or Hindlll and probed with radiolabeled NA-PI-II.Size markers (kb) are Hindlll restriction fragments of X-DNA.

EcoRI produced two hybridizing fragments (11 and 7.8 kb), whereasHindlll gave three large hybridizing fragments (16.6,13.5, and 10.5 kb).The NA-PI-II clone appears to belong to a small multigene family con-sisting of at least two members.

Proteinase Inhibitor in N. a/afa Stigmas 207

t ryps in

B

stigmastylepetalsepalOh leaf24h leaf

chymotrypsin

ug protein

Figure 5. Proteinase Inhibitor Activity in Various Organs of W. a/afa.

Buffer soluble extracts from various organs were tested for their abil-ity to inhibit trypsin and chymotrypsin.(A) Graphic representation of typsin inhibition.(B) Graphic representation of chymotrypsin inhibition.

domains. The first two domains are identical and containchymotrypsin-specific sites (Leu-Asn). The third, fourth, andfifth share 95 to 98% identity and have sites specific for tryp-sin (Arg-Asn). The sixth domain, which also has a trypsinreactive site, has less identity to the other domains (79 to 90%)

OD 280nmPI activity

Fraction no.

B

-16-9-14-4

-25

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

6-kD protein reached a peak concentration just before anthe-sis. The concentration of the 32-kD protein remained relativelyconstant during flower maturation.

DISCUSSION

Characterization of the N. alata PI cDNA Clone

N. alata contains a small gene family (two or three members)encoding proteins with some sequence identity to the serinePis of the potato type II family. The cDNA clone described cor-responds to one member of this family, which is expressedat high levels in stigmas at all stages of development from im-mature buds to mature, receptive flowers. The cDNA clonecorresponds to a 1.4-kb transcript and cross-hybridizes to asmaller transcript (1 kb). In the stigma, the larger messageis more abundant, whereas in the other floral tissues andwounded leaf tissue, the smaller message is more abundant.The protein encoded by the cDNA clone has six conserved

-16-9

-6-2-2-5

Figure 6. Purification of PI from N. alata Stigmas.

(A) Sephadex G-50 gel filtration chromatography of ammonium sul-phate precipitated proteins from stigma extracts. The PI activity elutedlate in the profile.(B) Twenty percent SDS-polyacrylamide gel (Laemmli, 1970) of frac-tions across the gel filtration column. The gel was silver stained andmolecular mass markers (Pharmacia peptide markers) are in kD. Aprotein of ~6 kD (arrow) co-elutes with the PI activity.(C) Analysis of Pi-containing fractions at different stages of the purifi-cation procedure by SDS-PAGE. Lane 1, crude stigma extract (5 ng);lane 2, stigma proteins precipitated by 80% (w/v) ammonium sulphate(5 ug); lane 3, PI protein eluted from the chymotrypsin affinity column0 ug).The PI is a 6-kD protein and is a major component in unfractionatedbuffer soluble extracts from stigmas.

208 The Plant Cell

-38 - -48 -

j c

'a 1

due mainly to a divergent 3'sequence. The related type II Pls of potato and tomato consist of two domains, each with a reactive site for trypsin or chymotrypsin. The two domains share 50 to 52% amino acid identity and are thought to have evolved through duplication of an ancestral gene (Graham et al., 1985; Sanchez-Serrano et al., 1986; Thornburg et al., 1987). This gene duplication may have occurred before speciation within the Solanaceae because in tomato and potato the first domains have more identity with each other (81 to 82%; Table 1) and the second domains have more identity with each other (88%) than the first and second domains within the same protein (50 to 52%). On this basis, an analogous PI gene with two domains might be expected in N. alata. The smaller mRNA of 1 kb seen in RNA gel blot experiments may encode such an inhibitor. Larger Pls such as the N. alata PI described here may also be present in tomato and potato. Indeed, a gene encoding an inhibitor with four domains from tomato has been described (Ryan and An, 1988). Both the potato and tomato Pls contain N-terminal sequences (Figure 1) that direct their cotranslational export into the endoplasmic reticulum, from where they are directed to their final destination in the vacu- ole (Wingate et al., 1991). The deduced sequence of the N. alata cDNA clone encodes an N-terminal sequence of 29 amino acids, similar to the potato signal sequence. The open reading frame of the cDNA clone encodes a protein of 43.4 kD; however, after cleavage of the putative signal sequence, the mature N. alata PI would have a molecular mass of -41.6 kD.

Localization and Properties of the PI Protein Encoded by the PI cDNA Clone

The gene corresponding to the cDNA clone is expressed at high levels in the cells of the stigma relative to the rest of the pistil, other floral organs, and wounded leaves. This pattern of gene expression is consistent with the high levels of PI activity in stigmas where the PI protein comprises m20 to 30% of the total soluble protein (data not shown). The expression of a PI at high levels in the stigma has not been reported previously, although Pena-Cortes et al., (1991) reported the occurrence of PI II mRNA in whole floral buds of potato and tomato and in ovaries, sepals, and petals of tomato.

A 6-kD protein with inhibitory activity toward trypsin and chymotrypsin was purified from stigma extracts rather than the 41.6-kD protein expected from the cDNA clone. This might have been due to degradation during the purification process, either enzymatically on the chymotrypsin column or during

(C) Hydropathy profile of the polypeptide encoded by the tomato PI II cDNA (Graham et al., 1985). The two domains are labeled I and II and the residues that would be potential processing sites (by analogy with the N. alata and potato Pls; Figure 1) are indicated with arrows. These sites are not present in regions predicted to be hydrophilic; con- sequently, a cleavage product is not marked.

Proteinase Inhibitor in N. alata Stigmas 209

—30

20-1

—14-4_8-2—6-2

—2-5

—94—67

43

—30

— 20-1

— 14-4— 8-2— 6-2— 2-5

Figure 8. Immunoblot Analysis of the PI Protein in Stigmas ofDeveloping Flowers.(A) Developing flowers of N. alata.(B) SDS-PAGE of stigma proteins at the stages of development shownin (A). Five micrograms of each extract was loaded. The peptide gelwas silver stained and molecular mass markers (LKB low molecularweight and Pharmacia peptide markers) are in kD.(C) Immunoblot of a gel identical to (B), probed with anti-Pi antiserum.

elution in 7 M urea at pH 3. However, because the 6-kD spe-cies is present at all stages of purification and is a majorcomponent in unfractionated extract from stigmas (Figure 6C),it is likely that processing of the 41.6- to the 6-kD componentoccurs in the plant. An antibody raised to the 6-kD compo-nent cross-reacts with a protein of ~42 kD in stigmas fromimmature flowers (Figure 8), which may be the precursor pro-tein encoded by the cDNA clone. The 18-kD component thatcross-reacts with the antibody may be a processing intermedi-ate composed of three domains. As the flower matures, the42- and 18-kD proteins decrease in concentration as the 6-kDprotein becomes more abundant. This may reflect the develop-mental regulation of the processing enzymes or possibly spatialconstraints to substrate-enzyme contact in the immature tis-sue. The 32-kD component persists at apparently similar levelsthroughout development and is not obviously related to pro-cessing. It may be the product of the 1.0-kb mRNA speciesfrom stigmas, which cross-hybridizes with the NA-PI-II cDNAclone.

Processing of the PI Protein Encoded by the PIcDNA Clone

The N-terminal sequence of the purified 6-kD PI is representedin each of the six repeated domains in the predicted sequenceof the 41.6-kD protein. Thus, it is likely that the 41.6-kD proteinis cleaved at six sites to produce seven peptides. Six of theseven peptides, peptides 2,3,4,5,6, and 7 (Figure 1; residues25 to 82, 83 to 142, 143 to 200, 201 to 258, 259 to 316, and317 to 370, respectively), would be in the same molecular massrange as the purified PI (~6 kD) and would have the sameN-terminal sequence. Indeed, the presence of both Lys andThr at position 11 in the N-terminal sequence of the purified6-kD PI indicates that it is composed of a mixture of peptides2 and/or 3, which have Thr at position 11, and peptides 4, 5,6, and 7, which have Lys at position 11. The detection of bothchymotrypsin and trypsin inhibitory activity in the stigma 6-kDPI preparation provides further evidence for a mixture of pep-tides, including peptide 2 (chymotrypsin inhibitor) and peptides3, 4, 5, and 6 (trypsin inhibitors). Peptide 7 does not containa consensus site for trypsin or chymotrypsin. Peptide 1(residues 1 to 24; Figure 1) is smaller than 6 kD, would havea different N terminus, and was not detected in the purifiedPI preparation. It could be envisaged that peptide 1 and pep-tide 7 would form a functional PI with the inhibitory site onpeptide 1 held in the correct conformation by disulphide bonds

Stigmas from developing flowers contain four proteins of ~42, 32, 18,and 6 kD that bind to the anti-Pi antibody. The 42-kD and the 18-kDcomponents decrease in concentration as the flowers mature, whereasthe 6-kD PI protein reaches a maximum concentration just before anthe-sis. The level of the 32-kD component, which runs as a doublet, doesnot alter significantly during flower development.

210 The Plant Cell

formed between the two peptides (see Greenblatt et al., 1989). However, there is no evidence to support this conjecture.

The processing site has not been determined but may be located at the Asn-Asp sequence at positions 24,82,142,200, 258, and 316 in Figure 1. Recently, a protease responsible for cleavage of an Asn-Gly linkage was isolated from the vacu- Oles of immature soybean seeds and is implicated in the posttranslational processing of proglycinin to produce mature glycinin (Scott et al., 1992). Another vacuolar protease isolated from pumpkin cotyledons also cleaves at the COOH-side of Asn but appears to have no specific requirement for the second amino acid (Hara-Nishimura et al., 1991). An enzyme with this specificity could be responsible for processing of the NA PI protein because cleavage at the Asn-Asp site could produce both the N and C termini of the small Pls from N. alara.

In the case of the N. alata 42-kD PI, processing analogous to that of the peptide hormones in animal systems is also possible. The peptide hormones are synthesized as large precursor proteins that undergo proteolytic cleavage to produce mature bioactive peptides (Kreiger, 1983; Douglas et al., 1984). The peptides in the precursor proteins are typically flanked by pairs of basic amino acids (usually Lys-Lys or LysArg). During maturation of the bioactive peptide, the paired basic sequence is cleaved by a trypsin-like endopeptidase and any remaining flanking amino acids are removed by a carboxypeptidase- or aminopeptidase-like enzyme (Steiner et al., 1980). A system such as this has not been described in plants, but there are analogous putative processing sites in the N. alata sequence. For example, each of the possible 6-kD proteins is flanked by dibasic residues (Lys-Lys; Figure 1).

A small peptide inhibitor of chymotrypsin (PCI-I) from potato tubers, completely homologous to a region of PI II, has been isolated and is believed to be produced by processing of a larger protein (Pearce et al., 1982; Figure 7B). Alignment of the predicted amino acid sequence of NA-PI-II with PCI-I from potato and another small inhibitor from eggplant (Figure 1) shows that they have similar N termini. The N terminus of the PCI-I protein could also be produced by an Asn-specific protease (Greenblatt et al., 1989; Figure l). However, a different mechanism is required for production of the C terminusof PCI-I because the second Asn-X processing site is absent in potato PI II. Greenblatt et al. (1989) proposed a mechanism in which a trypsin-like enzyme cleaves between Lys-80 and Ser-81 in potato PI II (Figure l), followed by trimming of the N terminus by a nonspecific aminopeptidase.

The three-dimensional structure of PCI-I, determined by x-ray crystallography, was used as a basis for computer modeling of a predicted structure of PI II. The potential processing sites in PI II for production of PCI-I are exposed in loops on the surface of the molecule (Greenblatt et al., 1989). The hydropathy plot of the polypeptide encoded by NA-PI-II (Figure 7A) shows that the potential processing sites on either side of the 6-kD PI sequence are also in predicted hydrophilic regions, which are likely to be on the surface of the molecule. Indeed, the processing sites in the N. alara 42-kD PI are predicted to be more hydrophilic than the corresponding sites in potato PI II.

Together with a potentially simpler processing mechanism, this may explain why the N. alata PI (42 kD) is efficiently processed to the 6-kD form, whereas most of the potato PI II protein (12 kD) is unprocessed. There are no reports of small inhibitors derived from the type II Pls from tomato. The first potential processing site in the tomato PI II protein is predicted to be in a hydrophobic region (Figure 7C) and is less likely to be exposed to processing enzymes.

Possible Function of the PI in Pistils

In the pistil, expression of the gene corresponding to the cDNA clone is restricted to the cells of the stigma. These are special- ized secretory cells that release large amounts of exudate onto the stigma surface as the flower matures (Kandasamy and Kristen, 1987). This exudate, which is rich in proteins, free amino acids, lipids, and carbohydrates (Kandasamy and Kristen, 1987), would provide a favorable environment for the growth of pests and pathogens. It is known, however, that infection of pistils is rare (Jung, 1956). The PI may have a defense role in the stigma because similar Pls are effective against pro- teases of fungal, bacterial, and insectorigin (Ryan, 1990). There is also limited evidence in other plants for the presence of defense-related molecules in the stigma. For example, in petu- nia, an active chitinase is localized in the stigmas of healthy, nonsenescent plants (Leung, 1992). In addition, the promoter of a bean hydroxyproline-rich glycoprotein is active in the stigma of transgenic tobacco (Wycoff et al., 1990). The Pls may be part of a group of defense-related molecules that, together, provide an effective barrier to entry into the stigma of microbial and insect pathogens and predators.

METHODS

Plant Material

Nicofiana alata (Link et Otto) plants of self-incompatibility genotype S1S3, S3S3, and SSSS were maintained under standard glasshouse conditions, as described previously (Anderson et al., 1989). Organs were collected directly into liquid nitrogen to avoid induction of a wound response and stored at -7OOC until required. For experiments on the effect of wounding on gene expression, leaves were wounded bycrush- ing across the midvein with a dialysis clip. Leaves were collected 4 and 24 hr after wounding.

ldentification and Sequencing of a cDNA Clone Encoding PI

Polyadenylated RNA was prepared from stigmas and styles isolated from mature flowers of N. alata (genotype S3S3) and used to construct a cDNA library in Lgt10 (Anderson et al., 1989). Single-stranded 32P- labeled cDNA was prepared from mRNA from stigmas and styles of N. alara (genotype S3S3 and s&) and was used to screen the library for highly expressed clones that were not S-genotype specific (Anderson et al., 1989). Plaques that hybridized strongly to cDNA probes from

Proteinase lnhibitor in N. alata Stigmas 21 1

both S-genotypes were selected and assembled into groups on the basis of cross-hybridization. The longest clone of each group was subcloned into M13mp18 and pGEM3zf+ and sequenced using an automated sequencer (model 373A; Applied Biosystems). 60th dye primer and dye terminator cycle sequencing chemistries were performed according to standard protocols (Applied Biosystems). Consensus sequences were generated using sequence editing software (SeqEd; Applied Biosystems). The GenBank data base was searched for sequences homologous to these clones. Because of the high degree of sequence similarity between the six domains of the N. alata PI clone, sequencing primers were made to nonrepeated 3' sequences (nucleotides 1117 to 1137,1188 to 1203, and 1247 to 1267) and to a5'sequence before the start of the repetitive regions (nucleotides 74 to 98). In addition, the pNA-PI-II insert was restricted with endonuclease Haelll, which cut at nucleotides 622 and 970 to produce three fragments. The fragments were subcloned into pGEMhf+ and sequenced in both directions using the M13 forward and reverse primers. The repetitive nature of the pNA-PI-II insert rendered it un- stable in both phagemid and plasmid vectors when cultures were grown longer than 6 h i

RNA Gel Blot Analysis

Total RNA was isolated and separated on a 1.2% agarose/formalde- hyde gel, as described previously (Anderson et al., 1989). The RNA was transferred to nylon membrane (Hybond-N; Amersham) and probed with the insert from pNA-PI-II labeled with Phosphorus-32 using random hexanucleotides (108 cpm pg-l; 107 cpm mL-'; Feinberg and Vogelstein, 1983). Prehybridization and hybridization at 68OC were as described by Anderson et al. (1989). The filters were washed in 2 x SSC (1 x SSC is 0.15 M NaCI, 0.015 M sodium citrate), 0.1% SDS or 0.2 x SSC, lO/o SDS at 68OC.

In Situ Hybridization

In situ hybridization was performed as described by Cornish et al. (1987). The probe was prepared by labeling the insert from pNA-PI-II (100 ng) to a specific activity of 108 cpm pg-l by random hexanucleotide prim- ing (Feinberg and Vogelstein, 1983). The labeled probe was precipitated and resuspended in hybridization buffer (50 pL), and 5 pL was applied to the sections. The sections were covered with coverslips and incubated overnight at 4OoC in a closed box containing 50% forma- mide. After incubation, sections were washed sequentially in 4 x SSC at room temperature, 2 x SSC at room temperature, and 1 x SSC at 4OoC for 30 min. The slides were dried and exposed directly to x-ray film (Cronex MRF 32; DuPont) overnight at room temperature. Hybrid- ized sections were counterstained with 0.025% (w/v) toluidine blue in H20 and permanent mounts were made using Eukitt mounting medium (Carl Zeiss, Freiburg, FRG) and a glass coverslip. Autoradio- graphs were transposed over sections to give the composites shown.

DNA Gel Blot Analysis

Genomic DNA was isolated from young leaves of N. alata using the procedure of Bernatzky and Tanksley (1986). DNA (10 pg) was digested to completion with the restriction endonucleases EcoRl or Hindlll, separated by electrophoresis on a 0.9% agarose gel, and transferred to nylon membrane (Hybond-N; Amersham) by wet blotting in 20 x SSC. Filters were probed and washed, as described for RNA blot analysis.

Preparation of Protein Extracts

Soluble proteins were extracted from plant material by freezing the tissue in liquid N2 and grinding to a fine powder in a mortar with a pestle. The powdered tissue was extracted in a buffer consisting of 100 mM Tris-HCI, pH 8.5,lO mM EDTA, 2 mM CaCI2, 14 pM P-mercaptoethanol. lnsoluble material was removed by centrifugation at 10,OOOg for 15 min. Protein concentrations were estimated by the method of Bradford (1976) with BSA as a standard.

Proteinase lnhibition Assays

Protein extracts and purified protein were assayed for inhibitory activity against trypsin and chymotrypsin, as described by Rickauer et al. (1989). lnhibitoryactivity was measured against 1 pg of trypsin (TPCK treated; Sigma) or 3 pg of chymotrypsin (TLCK treated; Sigma). The rate of hydrolysis of synthetic substrates N-a-P-tosykarginine methyl ester and N-benzoyl-L-tyrosine ethyl ester by trypsin and chymotryp sin, respectively, were taken as the uninhibited activity of the enzymes. lnhibitory activity of the extract was expressed as the percentage of control protease activity remaining after the protease had been prein- cubated with the extract.

Purification of the N. alata PI Protein

Stigmas (1000; 10 g) were ground to a fine powder in liquid N2 and extracted in buffer (100 mM Tris-HCI, pH 8.5, 10 mM EDTA, 2 mM CaCI2, 14 pM P-mercaptoethanol, 4 mL/g tissue). To concentrate the extract prior to the first purification step, gel filtration, the inhibitory activity was precipitated with 80% w/v ammonium sulphate, the concentration required to precipitate all the PI activity.

The ammonium sulphate pellet was resuspended in 5 mL of 0.15 M KCI, 10 mM Tris-HCI, pH 8.1, and loaded onto a Sephadex G-50 column (2 cm x 100 cm) equilibrated with the same buffer. The fractions (10 mL) eluted from this column and containing PI activity were pooled and applied to an affinity column of Chymotrypsin-Sepharose CL4B [I00 mg TLCK-treated a-chymotrypsin (Sigma) cross-linked to 15 mL Sepharose CL4B (Pharmacia) according to the manufacturers' instructions]. The column was washed with 10 volumes of 0.15 M KCl/lO mM Tris-HCI, pH 8.1, prior to elution of bound proteins with 7 M urea, pH 3 (5-mL fractions). The eluate was neutralized immediately with 200 pL 1 M Tris-HCI, pH 8, and dialyzed extensively against deionized HzO.

Amino Acid Sequence Analysis

Purified PI protein was chromatographed on a reverse phase HPLC microbore column prior to automated Edman degradation on a gas phase sequencer (Mau et al., 1986). Phenylthiohydantoin amino acids were analyzed by HPLC, as described by Grego et al. (1985).

Production of a Polyclonal Antiserum in Rabbits to the N. alata PI

The purified PI (Figure 6C, lane 3) was conjugated to a carrier protein, keyhole limpet hemocyanin (KLH; Sigma), using glutaraldehyde, as follows: one milligram of PI protein was dissolved in 1.5-mL H20 and mixed with 0.3 mg KLH in 0.5 mL of 0.4 M phosphate buffer, pH 7.5.

212 The Plant Cell

One milliliter of 20 mM glutaraldehyde was added dropwise over 5 min at room temperature while stirring. The mixture was stirred for an additional 30 min at room temperature, 0.25 mL of glycine was added, and the mixture was stirred for another 30 min. The conjugated protein was then dialyzed extensively against normal saline (0.8% NaCI). The equivalent of 100 kg of PI protein was used for each injection. Freunds complete adjuvant was used for the first injection and incom- plete adjuvant was used for two subsequent booster injections. The IgG fraction of the antiserum was separated on protein A Sepharose (Pharmacia) according to manufacturets instructions.

Protein Gel Blot Analysis

Protein extracts were electrophoresed in 15% SDS-polyacrylamide gels (Laemmli, 1970) and transferred to nitrocellulose in 25 mM Tris-HCI, 192 mM glycine, 20% v/v methanol, using a semi-dry electrophoretic transfer cell at 12 V for 20 min (Trans-Blot; BioRad). Loading and protein transfer were checked by staining the proteins on the membranes with Ponceau S (Harlow and Lane, 1988). Membranes were blocked in 3% (wlv) BSA for 1 hr and incubated with the anti-PI antibody (2 pglmL in 1% wlv BSA, Tris-buffered saline) overnight at room temperature. Bound antibody was detected using biotinylated donkey anti-rabbit IgG (11500 dilution; Amersham lnternational) and the Biotin- Streptavidin system according to procedures recommended by the manufacturer (Amersham International).

ACKNOWLEDGMENTS

We thank lngrid Bonig for assistance with the in situ hybridization experiments, Bruce McGinness and Susan Mau for maintenance of plant material, and Tim Orpin for instruction on the use of the SeqEd sequence editing software. A.H.A. was supported by an Australian Postgraduate Research Award and R.L.H. by the Department of Agricul- ture and a studentship from the Victorian Education Foundation.

Received November 30, 1992; accepted December 28, 1992.

REFERENCES

Anderson, M.A., McFadden, G.I., Bernatzky, R., Atkinson, A., Orpin, T., Dedman, H., Tregear, G., Fernley, R., and Clarke, A.E. (1989). Sequence variability of three alleles of the self-incompatibility gene of Nicotiana alata. Plant Cell 1, 483-491.

Bernatzky, R., and Tanksley, S.D. (1986). Genetics of actin-related sequences in tomato. Theor. Appl. Genet. 72, 314-321.

Bradford, M.M. (1976). A rapid and sensitive method for the quantita- tion of microgram quantities of protein using the principle of protein-dye binding. Anal. Biochem. 72, 248-254.

Brown, W.E., and Ryan, C.A. (1984). lsolation and characterization of a woundinduced trypsin inhibitor from alfalfa leaves. Biochem- istry 23, 3418-3422.

Bryant, J., Green, T.R., Gurusaddaiah, T., and Ryan, C.A. (1976). Proteinase inhibitor II from potatoes: lsolation and characterization of its protomer components. Biochemistry 15, 3418-3424.

Cornish, E.C., Pettitt, J.M., Bonig, I., and Clarke, A.E. (1987). De- velopmentally controlled expression of a gene associated with self-incompatibility in Nicotiana alata. Nature 326, 99-102.

Douglas, J., Civelli, O., and Herbert, E. (1984). Polyprotein gens expression: Generation of diversity of neuroendocrine peptides. Annu. Rev. Biochem. 53, 665-715.

Feinberg, A.P., and Vogelstein, 8. (1983). A technique for radiolabeling DNA restriction endonuclease fragments to high specific activity. Anal. Biochem. 132, 6-13.

Graham, J.S., Pearce, G., Merryweather, J., Titani, K., Ericsson, L.H., and Ryan, C.A. (1985). Wound induced proteinase inhibitors from tomato leaves. II. The cDNA-deduced primary structure of pre- inhibitor II. J. Biol. Chem. 260, 6561-6564.

Graham, J.S., Hall, G., Pearce, G., and Ryan, C.A. (1986). Regula- tion of synthesis of proteinase inhibitors I and II mRNAs in leaves of wounded tomato plants. Planta 169, 399-405.

Green, T.R., and Ryan, C.A. (1972). Wound-induced proteinase inhibitor in leaves: A possible defense mechanism against insects. Science 175, 776-777.

Greenblatt, H.M., Ryan, C.A., and James, M.N.G. (1989). Structure of the complex of Streptomyces griseus proteinase B and polypeptide chymotrypsin inhibitor-1 from Russet Burbank potato tubers at 2.1A resolution. J. MOI. Biol. 205, 201-228.

Grego, B., van Driel, I.R., Stearne, P.A., Goding, J.W., Nice, E.G., and Simpson, R.J. (1985). A microbore high-performance liquid chromatography strategy for the purification of polypeptides for gas-phase sequence analysis. Structural studies on the murine trans- ferrin receptor. Eur. J. Biochem. 148, 485-491.

Hara-Nishimura, I., Inoue, K., and Nishimura, M. (1991). A unique vacuolar processing enzyme responsible for conversion of severa1 proprotein precursors into the mature forms. FEBS Lett. 294,89-93.

Harlow, E., and Lane, D. (1988). Antibodies. A Laboratory Manual. (Cold Spring Harbor, NY Cold Spring Harbor Laboratory).

Hass, G.M., Hermodson, M.A., Ryan, C.A., and Gentry, L. (1982). Primary structures of two low molecular weight proteinase inhibitors from potatoes. Biochemistry 21, 752-756.

Jung, J. (1956). Sind narbe und griffel eintrittspforten für pilzinfek- tionen? Phytopathol. Z. 27, 405-426.

Kandasamy, M.K., and Kristen, U. (1987). Developmental aspects of ultrastructure, histochemistry and receptivity of the stigma of Nico- tiana sylvestris. Ann. Bot. 60, 427-437.

Kreiger, D.T. (1983). Brain peptides: What, where and why? Science 222, 975-985.

Kuo, T., Pearce, G., and Ryan, C.A. (1984). lsolation and characterization of proteinase inhibitor I from etiolated tobacco leaves. Arch. Biochem. Biophys. 230, 504-510.

Kyte, J., and Doolitte, R.F. (1982). A simple method for displaying the hydropathic character of a protein. J. MOI. Biol. 157, 680-685.

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

Leung, D.W.M. (1992). lnvolvement of plant chitinase in sexual repro- duction of higher plants. Phytochemistry 31, 1899-1900.

Mau, S.-L., Williams, E.G., Atkinson, A., Anderson, M.A., Cornish, E.C., Grego, E., Simpson, R.J., Kheyr-Pour, A., and Clarke, A.E. (1986). Style proteins of a wild tomato (Lycopersicon peruvianum) associated with expression of self-incompatibility. Planta 169, 184-191.

Proteinase lnhibitor in N. alata Stigrnas 213

Melville, J.C., and Ryan, C.A. (1970). Chymotrypsin inhibitor I from potatoes: A multi-site inhibitor composed of subunits. Arch. Biochem. Biophys. 138, 700-702.

Pearce, G., Sy, L., Russell, C., Ryan, C.A., and Hass, G.M. (1982). lsolation and characterization of two polypeptide inhibitors of serine proteinases. Arch. Biochem. Biophys. 213, 456-462.

Pearce, G., Ryan, C.A., and Liljegren, D. (1988). Proteinase inhibitors I and II in fruit of wild tomato species: Transient components of a mechanism for defense and seed dispersal. Planta 175,527-531.

Pena-Cortes, H., Willmitzer, L., and Sanchez-Serrano, J.J. (1991). Abscisic acid mediates wound induction but not developmental- specific expression of the proteinase inhibitor I I gene family. Plant Cell 3, 963-972.

Plunkett, G., Senear, D.F., Zuroske, G., and Ryan, C.A. (1982). Pro- teinase inhibitors l and II from leaves of wounded tomato plants: Purification and properties. Arch. Biochem. Biophys. 213,463-472.

Richardson, M. (1977). The proteinase inhibitors of plants and micro- organisms. Phytochemistry 16, 159-169.

Richardson, M. (1979). The complete amino acid sequence and the trypsin reactive (inhibitory) site of the major proteinase inhibitor from the fruits of aubergine (Solanum melongena L.). FEBS Lett. 104,

Rickauer, M., Fournier, J., and Esquerre-Tugaye, M. (1989). Induc- tion of proteinase inhibitors in tobacco cell suspension culture by elicitors of Phytophthora parasitica var. nicotianae. Plant Physiol. 9,

Ryan, C.A. (1984). Defense responses of plants. In Plant Gene Re- search, E.S. Dennis, B. Hohn, T. Hohn, P. King, J. Schell, and D.P.S. Verma, eds (New York: Springer-Verlag), pp. 375-386.

322-326.

1065-1070.

Ryan, C.A. (1990). Proteinase inhibitors in plants: Genes for improv- ing defenses against insects and pathogens. Annu. Rev. Phytopathol.

Ryan, C.A., and An, G. (1988). Molecular biology of wound-inducible proteinase inhibitors in plants. Plant Cell Environ. 11, 345-349.

Sanchez-Serrano, J.J., Schmidt, R., Schell, J., and Willmitzer, L. (1986). Nucleotide sequence of proteinase inhibitor II encoding cDNA of potato (Solanum tubemum) and its mode of expression. MOI. Gen. Genet. 203, 15-20.

Scott, M.P., Jung, R., Muntz, K., and Nielsin, N.C. (1992). A protease responsible for post-translational cleavage of a conserved Asn-Gly linkage in glycinin, the major seed storage protein of soybean. Proc. Natl. Acad. Sci. USA 89, 658-662.

Steiner, D.F., Quinn, P.S., Chan, S.J., Manh, J., and Tager, H.S. (1980). Processing mechanisms in the biosynthesis of proteins. Ann. NY Acad. Sci. 343, 1-16.

Thornburg, R.W., An, G., Cleveland, T.E., Johnson, R., and Ryan, C.A. (1987). Wound-inducible expression of a potato inhibitor II- chloramphenical acetyltransferase gene fusion in transgenic tobacco plants. Proc. Natl. Acad. Sci. USA 84, 744-748.

Wingate, V.P.M., Franceschi, V.R., and Ryan, C.A. (1991). Tissue and cellular localization of proteinase inhibitors I and II in the fruit of the wild tomato, Lycopersiconperuvianum (L.) Mill. Plant Physiol.

Wycoff, K.L., Powell, RA., Dixon, R.A., and Lamb, C.J. (1990). Tissue- specific and wound responsive expression of a hydroxyproline-rich glycoprotein gene promotor in transgenic tobacco. J. Cell. Biochem. 14E (suppl.), 344 (abstr.).

28, 425-449.

97, 490-495.

Proteinase lnhibitors in Nicotiana alata Stigmas Are Derived from a ...

Documents

Transcript of Proteinase lnhibitors in Nicotiana alata Stigmas Are Derived from a ...