Wheat ribosome-inactivating proteins: Seed and leaf forms with different specificities and cofactor requirements

Planta (1995)197:633 640 P l a n t ~

�9 Springer-Verlag 1995

Wheat ribosome-inactivating proteins: Seed and leaf forms with different specificities and cofactor requirements

Andrea J. Massiah, Martin R. Hartley

Department of Biological Sciences, University of Warwick, Coventry, West Midlands, CV4 7AL, UK

Received: 3 February 1995/Accepted: 27 March 1995

Abstract. Distinct forms of ribosome-inactivating proteins were purified from wheat (Triticum aestivum L.) germ and leaves and termed tritin-S and tritin-L, respectively. These differ in size and charge and are antigenically un- related. They are both RNA N-glycosidases which act on 26S rRNA in native yeast (Saccharomyces cerevisiae) ribosomes by the removal of A3024 located in a universally conserved sequence in domain VII which has previously been identified as the site of action of ricin A-chain. Tritin-S and tritin-L differ in both their ribosome substrate specificities and cofactor requirements. Tritin-S shows only barely detectable activity on ribosomes from the endosperm, its tissue of synthesis, whereas tritin-L is highly active on leaf ribosomes. Additionally, tritin-S is inactive on wheat germ, tobacco leaf and Escherichia coli ribosomes but active on rabbit reticulocyte and yeast ribosomes. Tritin-L is active on ribosomes from all of the above sources. Tritin-S, unlike tritin-L shows a marked requirement for ATP in its action.

Key words: Ribosome-inactivating protein (isoenzymes, specificity) - RNA N-glycosidase - Triticum

Introduction

Ribosome-inactivating proteins (RIPs) are widely distrib- uted in higher plants, having being detected in some 140 species (Barbieri et al. 1993). They most commonly exist as single-chain polypeptides of approximately 30 kDa and inhibit the translocation step in protein synthesis through the removal of a single adenine residue from a universally conserved stem-loop structure in domain VII (domain VI in eubacteria) in the large-subunit large rRNA (Endo and

Abbreviations: CM-= carboxymethyl-; FPLC = fast protein liquid chromatography; NEPHGE = non-equilibrium pH gradient gel electrophoresis; PAP = pokeweed antiviral protein; RIP = ribosome-inactivating protein Correspondence to: M.R. Hartley; FAX: 44 (1203) 523701; E-mail: [email protected]

Tsurugi 1987). Such single-chain RIPs are relatively non toxic to cells and organisms since they are unable to pen- etrate cells. In a few plants, the catalytic polypeptide (in this case the A-chain) has become fused to a lectin B-chain with galactose-binding properties. Such heterodimeric RIPs (toxic lectins or type II RIPs) can bind to and enter target cells and are amongst the most toxic substances known. An example of the latter is ricin from Ricinus communis seeds.

Although the physiological role of single-chain RIPs is not known with certainty, it is widely believed that they have a defensive function, protecting the plants from at- tack by pathogens and predators. The most widely studied defensive property is the antiviral activity shown towards infections by diverse plant viruses (Chen et al. 1991; Lodge et al. 1993; Taylor et al. 1994). An explanation for the antiviral activity of RIPs was first proposed by Ready et al. (1986). They performed immunocytochemical localisation studies using electron microscopy on pokeweed antiviral protein (PAP), a single-chain RIP, in the leaves of the pokeweed plant (Phytolacca americana) and found it to be localised largely in the cell wall matrix. They proposed that local cellular damage caused by a viral vector (e.g. an aphid) would result in release of the RIP into the cytosol where it could inactivate sensitive ribosomes, thereby pre- venting virus replication. The susceptibility of pokeweed ribosomes to PAP has recently been confirmed (Taylor and Irvin 1990; Bonness et al. 1994) and susceptibility to the homologous RIP has been reported for four other dicotyledonous species (Prestle et al. 1992). The extracellular localisation of a single-chain RIP has also been reported for saporin in the seeds and leaves of Saponaria officinalis (Carzaniga et al. 1994). The antiviral hypothesis is also supported by recent studies in which a strong correlation was demonstrated between the activities of five RIPs in the depurination of tobacco leaf ribosomes and inhibition of tobacco mosaic virus (TMV) (Taylor et al. 1994).

The evidence at present indicates that the majority of single-chain RIPs produced by dicotyledonous plants are active on their homologous ribosomes (Hartley and Lord 1993). However, this is not the case for cereal seed RIPs.

634

Tri t in , a single-chain RIP in wheat seeds, was first identified as a componen t of a wheat-germ protein-synthesising system that caused inh ib i t ion of protein synthesis in an ascites cell-free system (Stewart et al. 1977). Whea t and maize r ibosomes are resistant to their seed RIPs (Taylor and Irvin 1990; Bass et al. 1992) and the e D N A genes encoding the seed RIPs from barley and maize do not encode N- te rmina l signal sequences, sugges- t ing that these RIPs are cytosolic (Leah et al. 1991; Walsh et al. 1991). There is some evidence that cereal seed RIPs have an ant i fungal role (Roberts and Selitrennikoff 1986a; Leah et al. 1991; L o g e m a n n et al. 1992).

A n u m b e r of RIPs have been shown to have a pronounced requi rement for A T P for activity whereas others show little or no dependency (Carnicelli et al. 1992). The funct ion of A T P is uncer ta in , a l though it is not used to phosphoryla te the R I P and is not needed directly for catalysis (Stewart et al. 1977; Sperti et al. 1991).

Another element of diversity in single-chain RIPs is the f inding that m a n y plants produce several RIP isoforms of varying relatedness, either present within the same organ or in different organs (reviewed in Barbieri et al. 1993). In all reported cases the RIP isoforms in a given species share ant igenic determinants , and a l though their relative activities may differ on a given r ibosome substrate, there are no reports of different r ibosome specificities (Hartley and Lord 1993).

In this paper we report that wheat seed and leaf contain immunological ly distinct RIP isoforms which differ in their substrate specificity and requirement for ATP. To our knowledge this is the first report that RIP isoforms of such diversity are produced in different organs of the same plant.

Materials and methods

Plant material. Winter wheat (Triticum aestivum L. cv. Haven) was used for the work reported, with the exception of the source material for tritin-S purification, which was commercial wheat germ. Ears of field-grown wheat (Horticultural Research International, Welles- bourne, Warwickshire, UK) were harvested at 27 d post-anthesis and the immature seeds stored in liquid nitrogen. Leaf material, the source of tritin-L, was harvested from 10-d-old plants grown in Levingtons compost in the dark at 22 ~

Purification oftritin-S. The seed form of tritin was purified from 200 g commercial wheat germ following the procedure described by Roberts and Stewart (1979) with the exception that the diethylaminoethyl (DEAE)-cellulose and phosphocellulose chromatography steps were substituted by a carboxymethyl (CM)-cellulose chromatography step (column dimensions 1.6 i.d., 23 cm long). In this step, elution was carried out with a linear gradient of 0-0.3 M NaC1 in 5 mM sodium phosphate (pH 6.5) at a flow rate of 0.6 ml. min- 1.

Purification oftritin-L. Leaves (200 g) were homogenised in 600 ml of 100mM Tris-HC1 (pH 8.5), 100mM KC1, 4mM MgCI v 5 mM dithiothreitol by ten, 20-s bursts at full power using a Polytron homogeniser (Kinematica, Lucerne, Switzerland). The homogenate was filtered through Miracloth (Calbiochem, La Jolla, Calif., USA) and the filtrate centrifuged at 10000 rpm and 4~ for 30 min in a GSA rotor (Sorvall, Stevenage, Herts., UK). The resulting supernatant was centrifuged at 50000 rpm and 4~ for 3 h in a 55.2 Ti rotor (Beckman, High Wycombe, Bucks., UK). Supernatant proteins were precipitated by the addition of ammonium sulphate to 100% saturation and dialysed against column buffer (5 mM sodium phos-

A.J. Massiah and M.R. Hartley: Wheat ribosome-inactivating proteins

phate, pH 6.5). The protein solution was applied to a CM-Sepharose (Pharmacia, Milton Keynes, Bucks, UK) column (2.6 i.d., 24 cm long) and the unbound proteins removed by washing the column with column buffer. Elution of bound proteins was carried out by the application of a 1000-ml linear gradient of 0 0.5 M NaCI in column buffer at a flow rate of 0.75 ml 'min -1. The protein fractions containing tritin-L were pooled, dialysed against column buffer and applied to a Mono S column (1.2 i.d., 3.7 cm long) fitted to a fast protein liquid chromatography (FPLC) system (Pharmacia-LKB, Milton Keynes, Bucks., UK). After washing with column buffer to remove unbound proteins, bound proteins were eluted by the se- quential application of two linear salt gradients (10 ml of 0-0.1 M NaC1 and 120ml of 0.1-0.25 M NaCI in column buffer) at a flow rate of 5.0 ml. min- 1. Fractions containing tritin-L were pooled and concentrated using a Centriprep-10 concentrator (Amicon, Stone- house, Gloucs., UK).

The concentration of tritin-L and tritin-S were determined by the method of Bradford (1976) using a BioRad (Watford, UK) protein assay kit and recombinant ricin A-chain (Zeneca, Alderley Edge, Cheshire, UK) as a protein standard.

Isolation of ribosomes. Tobacco leaf ribosomes were isolated from expanding leaves (ca. 5 cm in length) of Nicotiana tabacum cv. Sam- son NN by the method of Jackson and Larkins (1976) as described previously (Taylor et al. 1994). Yeast ribosomes were prepared from the vacuolar protease-deficient Saccharomyces cerevisiae ABYS 1 (Rothblatt and Meyer 1986) as described previously (Hartley et al. 1991). Escherichia coli ribosomes were prepared from mid-exponen- tial cultures of E. coli PR-7 as described by Traub et al. (1971). Rabbit reticulocyte ribosomes were prepared from non-nuclease- treated reticulocyte lysate (Promega, Southampton, UK). Re- ticulocyte lysate (1 ml) was layered over a 1-ml cushion of 1.0 M sucrose in 25 mM Tris-HC1 (pH 7.6), 25 mM KC1, 5 mM MgCl 2 and the ribosomes pelleted by centrifugation at 100 000 rpm and 4 ~ for 40 min in a Beckman TL-100.3 rotor. The ribosomes were resuspended in 2 ml 25 mM Tris-HC1 (pH 7.6), 25 mM KCI, 5 mM MgC12, pelleted as before and resuspended in the same buffer. Wheat-germ ribosomes were prepared from a wheat-germ lysate, prepared by the method of Anderson et al. (1983), using the same procedure as for reticulocyte ribosomes. For the preparation of endosperm ribosomes, wheat seeds at 27 d post-anthesis were cut transversely, whilst still frozen, so as to produce embryo-containing and embryo-less portions. Embryo-less portions were used to isolate ribosomes by the procedure of Taylor et al. (1994). Isolated ribosomes were stored at -70 ~ in 25 mM Tris-HC1 (pH 7.6), 25 mM KCI, 5 mM MgCI 2 at 15-20 ~tg. ~t1-1.

In vitro assay for the detection of RNA N-glycosidase activity. In the standard assay, isolated ribosomes (40 ~tg) were incubated with the RIP at 30 ~ for 30 n'fin, in a total volume of 20 ~tl 25 mM Tris-HC1 (pH 7.6), 25 mM KCI, 5 mM MgCI z. Control reactions lacking RIPs were similarly incubated. ATP, at a final concentration of 1 mM, was added to certain assays as indicated. Following incubation RNA was extracted (Hartley et al. 1975) and dissolved in sterile distilled water at 1 ~tg.~tl 1. Aniline treatment of RNA and electrophoresis in agarose/formamide gels were carried out as described previously (May et al. 1989). Gels were stained in ethidium bromide and photo- graphed on a UV transilluminator. Measurement of the extent of depurination of rRNA was calculated as described previously (Taylor et al. 1994) and expressed as the DC 50 value (the concentration of RIP giving 50% depurination of rRNA in the standard assay).

Detection, by Northern analysis, of the rRNA fragment released upon aniline-cleavage at the depurination site. RNA was transferred from agarose/formamide gels to Hybond-N (Amersham, Bucks., UK) as described in Sambrook et al. (1989). The probe used for detection of the RNA fragment released from the 3' end of wheat 25S rRNA following aniline treatment was made by RT-PCR from tobacco 25S rRNA as described by Taylor et al. (1994). The DNA product was labelled with 3zp by random priming (Sambrook et al. 1989). Filter hybridisation was performed as described previously (Taylor et al. 1994).


Determination of the RIP depurination site by primer extension analysis. The primer used (5' CATAATCCAGCGGATGG 3') is complementary to Saccharomyces cerevisiae 26S rRNA at positions 3088 to 3104 (Georgiev et al. 1981). Annealing of the primer, extension reactions, dideoxynucleotide sequencing reactions and analysis of products in sequencing gels were carried out by modifications of the method described by Moazed et al. (1986) as detailed by May et al. (1989).

Polyacrylamide gel electrophoresis of proteins. Proteins were analysed by SDS-PAGE on gels consisting of a 7.5% (w/v) acrylamide stacker and a 15% (w/v) acrylamide resolving gel as described by Laemmli (1970). Two-dimensional gel electrophoresis was performed as described by O'Farrell (1975) with modifications to the procedure made for analysing basic proteins by non-equilibrium pH gradient electrophoresis (NEPHGE) (O'Farrell et al. 1977). First dimension gels, containing carrier ampholytes in the pH range 6.0-10.5, were cast to a height of 100 mm in glass tubes of 3 mm internal diameter.

Preparation of anti-tritin-S and anti-tritin-L polyclonal antibodies. Purified tritin-S and tritin-L (50 ~tg each) were injected subcu- taneously into Sandy half-lop rabbits at two-weekly intervals. The first injection was carried out in the presence of Freund's complete adjuvant (Gibco BRL, Paisley, Scotland, UK) and all subsequent injections in Freund's incomplete adjuvant (Gibco BRL). Serum was prepared from blood obtained two weeks after the fourth injection and stored at - 20 ~

Western analysis. Transfer of proteins from SDS-polyacrylamide gels to Hybond-C (Amersham) was carried out as described by Towbin et al. 0979). Antibody probing of filters and visualisation of proteins were carried out as described by Chaddock and Roberts (1993). Tritin-S and tritin-L polyclonal antibodies were used at a 1 : 250 dilution and the secondary antibody, conjugated to alkaline phosphatase (Promega), at a 1 : 7500 dilution.

635

the eluate fractions showing absorbance at 280 nm were analysed directly by S D S - P A G E and assayed for R N A N-glycosidase activity on yeast r ibosomes (Fig. 1). The silver-stained protein gel (Fig. 1A) shows that co lumn fractions 9-15 inclusive contain two prominent polypeptides of approximate Mr 37 000 and 37 860. These fractions also contain R N A N-glycosidase activity (Fig. 1B), as judged by the release of a fragment of approximate ly 370 nucleotides f rom the 26S r R N A (Georgiev et al. 1981; Hart ley et al. 1991). It is concluded that the activity resides in either one or both of these polypeptides.

Comparison of tritin-S and tritin-L by 2-D gel electrophoresis. Most single chain RIPs are known to be basic (Hartley and Lord 1993) and H a b u k a et al. (1993) esti- mated an isoelectric point (pI) of 10.13 for an amino-acid sequence deduced from a genomic clone believed to encode the seed form of tritin. A mixture of purified tritin-S and tritin-L was separated in the first dimension by N E P H G E , which is the preferred technique for the separ- at ion of highly basic proteins (O'Farrel l 1977), and in the second dimension by S D S - P A G E (Fig. 2). Tritin-S separates into two partially resolved bands (arrowed) of approximate Mr 32 100 and 32 800 which are of similar basic charge. In contrast , trit in-L is approximate ly neutral, and separates into two major charge variants of Mr 37 860 and 37 000 (arrowed).

Immunological dissimilarity. Polyclonal antisera against tritin-S and tritin-L were raised in rabbit and used to probe Western blots of samples of the two R I P s (Fig. 3).

Results

In preliminary experiments it was observed that the r R N A prepared from wheat leaf r ibosomes showed the modifica- t ion characteristic of that caused by RIP action, i.e. a fragment of ca. 370 nucleotides was released from 25S r R N A following aniline t reatment (Endo and Tsurugi 1987). It is k n o w n that the r ibosomes from many plants producing single-chain RIPs are modified during their extraction th rough contact with a R I P that is sequestered in situ (Taylor and Irvin 1990; Prestle et al. 1992). Previous work on the wheat seed RIP, tritin, had shown it to be inactive on wheat seed r ibosomes (Taylor and Irvin 1990). This raises the following possibilities: either the seed and leaf R I P s have similar r ibosome-inact ivat ing specificities but seed and leaf r ibosomes differ in their sensitivity towards the RIPs or the seed and leaf RIPs show different specificities towards similar r ibosomes in the leaf and seed.

It was therefore decided to purify the leaf R I P and compare its properties with those of the seed RIP. We propose to name the wheat seed and leaf RIPs tritin-S and tritin-L respectively.

Purification of tritin-L. C h r o m a t o g r a p h y of a crude soluble leaf extract on CM-Sepharose yielded a fraction, eluting at 70-250 m M NaC1, consisting primarily of a polypept ide of 37 k D a and showing R N A N-glycosidase activity on yeast r ibosomes (data not shown). This frac- t ion was ch roma tog raphed on a M O N O - S column and

Fig. 1A, B. Purification of tritin-L from wheat by MONO-S FPLC and detection of RNA N-glycosidase activity in column fractions. A Silver-stained SDS-polyacrylamide gel of the fractions obtained by elution with a 0.1=0.25 M NaC1 gradient. B Detection of RNA N-glycosidase activity on S. cerevisiae ribosomes. The rRNA in all tracks has been treated with aniline. The arrow indicates the fragment released upon aniline cleavage at the depurination site

636 A.J. Massiah and M.R. Hartley: Wheat ribosome-inactivating proteins

Fig. 2. Two-dimensional gel analysis of tritin-S and tritin-L. Silver- stained second dimension SDS-polyacrylamide gel containing 2.5 lag each of tritin-S and tritin-L. Samples were electrophoresed in the first dimension by NEPHGE for a total of 1500 Vh in the presence of ampholytes in the pH range 6.0-10.5. Arrows labelled L and S indicate isoenzymes of tritin-L and tritin-S, respectively

Fig. 4. Determination of the depurination site in S. cerevisiae 26S rRNA caused by ricin-A chain (RTA), tritin-S and tritin-L. The S. cerevisiae ribosomes were incubated in the absence of RIP (unmodified sample) or in the presence of 0.2 laM RTA, tritin-S or tritin-L. The rRNAs were extracted and used as templates for primer extension. The unmodified rRNA was also used as a template for dideoxynucleotide sequencing reactions. GATC refer to sequencing reactions carried out on unmodified rRNA template incorporating the appropriate ddNTP. The numbers next to the sequence repres- ent the position of the bases in S. cerevisiae 26S rRNA according to the sequence of Georgiev et al. (1981). The arrowhead indicates the major primer extension product. The depurination site is shown by �9

Fig. 3A, B. Immunological dissimilarity between tritin-S and tritin- L. A An immunoblot containing 0.5 lag tritin-S and 5.0 lag tritin-L probed with anti-tritin-S polyclonal antibodies. B An immunoblot containing 5.0 lag tritin-S and 0.5 lag tritin-L probed with anti-tritin- L polyclonal antibodies. M denotes BioRad pre-stained broad- range protein standards

Tritin-S ant iserum reacts only with tritin-S and likewise tritin-L ant iserum reacts only with tritin-L. There is no visible cross-reaction between the heterologous combina- tions of ant iserum and RIP. It is concluded that tritin-S and tri t in-L are distinct RIPs.

Determination of the depurination site in 26S rRNA result- in9 from the action of tritin-S and tritin-L on yeast ribosomes. The precise position(s) of the modificat ion to yeast

26S r R N A caused by tritin-S and tritin-L was determined by primer extension. Ricin A-chain was also used, as the modification site in yeast 26S r R N A had previously been determined as A3024 (Stirpe et al. 1988; Hart ley et al. 1991). The rationale for this method is that reverse tran- scriptase is unable to read th rough the depurinat ion site, generating a band corresponding to the length of the c D N A from the 5' end of the primer to the nucleotide immediately preceding the modified position. The primer used is complementary to nucleotides 3088-3104 in Sac- charomycescerevisiae 26S r R N A (Georgiev et al. 1981). The 3' end of this lies 64 nucleotides to the 3' side of the putative depurinat ion site in the rRNA. Figure 4 shows a sequencing gel of the products of primer extension from unmodified 26S rRNA, ricin A-chain, tritin-S and tritin-L modified rRNAs, together with dideoxy sequencing lanes from unmodified r R N A extended from the same primer. Specific termination products, all of which corresponds to depurinat ion at A3024, are produced when the templates used have been modified by all of the three RIPs. This terminat ion product is absent in the extension products of unmodified rRNA. It is concluded that tritin-S and tritin- L depurinate yeast 26S r R N A at A3024, the same posit ion as ricin A-chain.

AJ. Massiah and M.R. Hartley: Wheat ribosome-inactivating proteins 637

Fig. 5A, B. Effect of ATP on the RNA N-glycosidase activities of tritin-S and tritin-L on rabbit reticulocyte ribosomes. Serial dilutions of tritin-S (A) and tritin-L (B) were assayed for RNA N- glycosidase activity on rabbit reticulocyte ribosomes in the absence or presence of 1 mM ATP. A Tritin-S was used at the following concentrations: lanes 1, 2, 7 and 8, 2 gM; lanes 3, 4, 9 and 10, 0.2 I,tM; lanes 5, 6, 11 and 12, 0.02 gM; lanes 13 and 14, 2 nM; lanes 15 and 16, 0.2 nM; lanes 17 and 18, no added RIP (negative control). B Tritin-L was used at the following concentrations: lanes 1, 2, 11 and 12, 1.3 IxM; lanes 3, 4, 13 and 14, 0.13 gM; lanes 5, 6, 15 and 16, 13 nM; lanes 7, 8, 17 and 18, 1.3 nM; lanes 9, 10, 19 and 20, no added RIP (negative control). + and - below the tracks indicate aniline-treated and untreated rRNA, respectively. The arrowhead indicates the fragment released upon aniline cleavage at the depurination site

The A T P requirements f o r tritin-S and tri t in-L activity. It is known that tritin-S has a pronounced requirement for ATP for the inhibition of poly(U)-directed polyphenyl- alanine synthesis in cell-free systems from ascites cells and Artemia salina (Roberts and Stewart 1979; Carnicelli et al. 1992). To investigate whether this ATP requirement could be demonstrated directly for the RNA N-glycosidase activity of tritin-S, rRNA depurination assays were performed on rabbit reticulocyte ribosomes in reaction mix- tures containing or lacking ATP. The ATP was included at a final concentration of 1 m M which had been shown by Roberts and Stewart (1979) to be saturating for the inhibition of protein synthesis in ascites cell extracts. Figure 5A shows the effects of ATP on the RNA N- glycosidase activity shown by a range of concentrations of tritin-S. In order to quantitate the activities the DCso values (the concentration of RIP giving 50% rRNA depurination under standard assay conditions) were determined as described in Materials and methods. In the absence of ATP the DCso value for tritin-S activity is 0.58 gM and in the presence of 1 m M ATP is 4.4 nM. It is concluded that ATP causes a 132-fold stimulation of the activity of tritin-S on reticulocyte ribosomes. Assays were

similarly performed to determine whether or not tritin-L has an ATP requirement (Fig. 5B). In this case the DCso values are 23 nM in the absence of ATP and 30 nM in the presence of ATP. It is concluded that there is a marked difference between the ATP requirements of tritin-S and tritin-L.

Ribosome substrate specificities o f tritin-S and tritin-L. The fact that wheat-germ lysates contain high concentrations of tritin-S (Roberts and Stewart 1979) yet are active in protein synthesis suggests that tritin-S is inactive on wheat-germ ribosomes. This was subsequently confirmed by Taylor and Irvin (1990) using the depurination assay. Our preliminary observation that wheat leaf ribosomes could serve as a substrate for tritin-L prompted a more detailed investigation of the r ibosome substrate specificities of tritin-S and tritin-L. Figure 6A shows depurination assays on rRNA extracted from ribosomes from rabbit reticulocyte lysate, S. cerevisiae, tobacco leaf, wheat germ and E. coli which had been incubated with tritin-S. ATP (1 mM) was included in all assays and the tri t in-S:ribosome molar ratio was the high value of approximately 4: 1. As judged by the appearance of the

638

diagnostic rRNA fragment released by aniline, only the rabbit reticulocyte and S. cerevisiae ribosomes act as substrates for tritin-S. Tritin-S was also assayed on the ribosomes from the leaves of Lotusjaponicus and Arabidop- sis thaliana, both of which were resistant (data not shown). Figure 6B shows the results of similar assays to those above but using tritin-L. In contrast to tritin-S, tritin-L is active on all of these ribosomes. It is clear that tritin-L has a much broader substrate range in terms of ribosome source than tritin-S.

The finding that tritin-S is inactive on wheat-germ ribosomes is consistent with the observation that the genomic sequence does not encode an N-terminal signal sequence (Habuka et al. 1993), from which it can be inferred that the RIP accumulates in the cytosol in contact with ribosomes. It was therefore surprising that Kataoka et al. (1992) reported, using a sensitive assay based on primer extension, that 25S rRNA extracted from a wheat- germ homogenate was specifically depurinated but to a low level. However a possible source of misinterpreta- tion arises in this work since it is known for maize and barley seeds (Di Fonzo et al. 1986; Leah et al. 1991) and also wheat (data not shown) that the respective RIPs are located exclusively in the endosperm. In the work of Kataoka et al. (1992), wheat-germ ribosomes, derived largely from the embryo, were used and these ribosomes would not be the physiological substrate for tritin-S. The


tritin-S present in wheat germ is most likely to be derived from contaminating endosperm fragments (Morch et al. 1988).

In order to clarify this possible source of error, ribosomes were prepared from embryo-less halves of seeds at 27 d post-anthesis (designated endosperm ribosomes) and used as a substrate for tritin-S action. Initially, the rRNA from such assays was analysed by fractionation in de- naturing agarose gels and ethidium bromide staining. Even at the high RIP:ribosome molar ratios of 4:1 no activity was detectable (data not shown). As a more sensitive assay for the detection of the 'aniline fragment', gels were Northern-blotted and the blots hybridised with a probe specific for the 3' end of 25S rRNA as described by Taylor et al. (1994). An example of such a Northern blot assay is shown in Fig. 7A. Incubation of endosperm ribosomes with pokeweed antiviral protein (PAP), a highly active single-chain RIP with broad substrate specificity (reviewed in Hartley and Lord 1993) purified from the leaves of Phytolacca americana, results in the almost complete depurination of 25S rRNA (Fig. 7A). Incubation of endosperm ribosomes with tritin-S at a RIP:r ibosome ratio of ca. 40:1 also results in specific depurination but to a very small extent (Fig. 7A). However, rRNA extracted from a control incubation of endosperm ribosomes, without added RIP, shows no detectable depurination (Fig. 7A). It has been demonstrated that the endosperm tissue of seeds at 27 d post-anthesis is actively accumulating tritin-S (data not shown) which would suggest that tritin-S is not active in situ. In contrast to this situation, 25S rRNA extracted from ribosomes prepared from leaf homogenates is specifically depurinated without the need for in vitro incubation with RIP (Fig. 7B). Similar obser- vations have been reported for the ribosomes of several other species of higher plants that produce single-chain leaf RIPs (Taylor and Irvin 1990; Prestle et al. 1992).

Fig. 6A, B. Ribosome substrate specificity of tritin-S and tritin-L. Purified tritin-S at 2 ~tM (A) and tritin-L at 2 laM (B) were assayed for RNA N-glycosidase activity on the ribosome substrates shown. All assays were carried out in the presence of tmM ATP. + and -- indicate aniline-treated and untreated rRNA, respectively

Discussion

The occurrence of RIP isoenzymes in the same or different organs of the same plant is a common feature. For a particular tissue, isoenzymes which usually differ in isoelectric point, are commonly separated by chromatography over cation-exchange resins and elution by shallow salt gradients.

In this study, the seed and leaf forms of wheat tritin have been purified, both forms of which comprise of at least two isoenzymes. Attempts were made to separate these isoenzymes but these were unsuccessful. Reisbig and Bruland (1983) have previously resolved the tritin-S isoenzymes and have shown that theses are antigenically similar and have indistinguishable enzymatic activities.

Recently it has become evident that ATP is required by RIPs to a varying degree for optimal RNA N- glycosidase activity on ribosomes. Tritin-L activity on reticulocyte ribosomes was shown to be unaffected by the presence of ATP whilst a strong dependency was observed for tritin-S. The fact that some RIPs exhibit no dependency on ATP whilst the activity of others is enhanced in its presence raises the question as to the involvement of ATP. With regard to RIP-ribosome interaction, Sperti et al.


Fig. 7A, B. Differences in sensitivity of endosperm and leaf ribosomes to the action of endogenous RIPs during their extraction. A Ribosomes were prepared from embryo-less halves of wheat seeds at 27 d post-anthesis. Ribosomes were incubated in the absence of added RIP (track labelled No RIP), with 20 ~tM tritin-S or with 1.6 IxM PAP under standard assay conditions in the presence of 1 mM ATP. The agarose/formamide gel was Northern blotted and probed with a 32p-labelled DNA fragment complementary to the 3' region of 25S rRNA. B Ribosomes, prepared from 10-d-old wheat leaves, were incubated in the absence of added RIP and in the presence of 1 mM ATP. RNA species were visualised by ethidium bromide staining of the agarose/formamide gel. The arrow indicates the fragment released upon aniline cleavage at the depurination site. + and

- indicate aniline-treated and untreated rRNA, respectively

(1991) have studied the kinetics of adenine release caused by gelonin (a single-chain RIP from Gelonium multiflorum seeds) inactivation of Artemia salina ribosomes in the presence and absence of ATP. They showed that the apparent Km is essentially the same for both conditions whilst the kc,t is reduced by the order of 1000-fold in the absence of ATP. It therefore seems unlikely that ATP affects the binding of the RIP to the ribosome. The most probable explanation for the ATP requirement is that it affects ribosome structure in such a way that it renders the rRNA more sensitive to depurination. A possible way in which this could be mediated is through the reversible phosphorylation of a ribosomal protein(s) through a ribosome-associated kinase. Hydrolysis of the 7-phosphate bond of ATP seems necessary for its function as a cofactor since adenylyl ([3,7-methylene) diphosphonate (AMP- PCP), a non hydrolysable analogue of ATP, was shown not to stimulate tritin-S activity on reticulocyte ribosomes (data not shown).

The molecular basis for the differing ribosomal specificity of tritin-S and tritin-L is unknown. Although the primary structure of tritin-L is unknown, the postulated active-site residues in RNA N-glycosidase RIPs sequenced to date are similar and key residues thought to be involved in catalysis are absolutely conserved (Katzin

639

et al. 1991; Robertus 1991). The rRNA sequence around the RIP target site is totally conserved in eukaryotes (Rau6 et al. 1988) which makes this an unlikely candidate for the specificity differences observed. It is known for ricin that native E. coli ribosomes do not serve as a substrate for its action whereas the deproteinised rRNA is a substrate. This suggests that ribosomal protein(s) affect the conformation of ribosomes in ways that render the rRNA either sensitive or refractory to a given RIP.

It is clear that wheat leaf cytoplasmic ribosomes are sensitive to tritin-L (Fig. 7B), a finding consistent with the antiviral hypothesis working through a local suicide mechanism. Contrary to what is observed with tritin-L action on endogenous cell ribosomes, wheat endosperm rRNA is not depurinated by endogenous levels of tritin-S. However, wheat endosperm ribosomes are not totally refractory to the action of tritin-S since the incubation in vitro of endosperm ribosomes with very high concentrations of tritin-S does result in a very low level of rRNA depurination (Fig. 7A). The relative insensitivity of wheat endosperm ribosomes to tritin-S and the inferred cytosolic location of the protein would suggest that the seed form of tritin could not function with an antiviral role as speculated for many other single-chain RIPs, including tritin-L.

Cereal seed RIPs accumulate late during seed develop- ment (Soave et a1.1981; Leah et al. 1991) and are relatively abundant proteins. Tritin-S, for example, represents approximately 2% of the total soluble protein in mature wheat seeds (Coleman and Roberts 1982). The RIPs of several cereal seeds have been shown to retain their activities and remain in the seed until at least 7 d after germination. However, more-recent work on the maize kernel RIP has shown that it is synthesised as an inactive proenzyme which is proteolytically activated during seed germination by the removal of an internal peptide (Walsh et al. 1991). A possible physiological role served by cereal seed RIPs is that of antifungal agents which protect the seed during dormancy and early after germination. Indeed, cereal seed RIPs have been shown to inhibit fungal growth (Roberts and Selitrennikoff 1986b; Leah et al. 1991) and for the barley seed RIP in particular, fungal ribosomes have been shown to be extremely sensitive to RIP action (Roberts and Selitrennikoff 1986a). Logemann et al. (1992) have expressed the barley seed RIP in tobacco and these transgenic plants were observed to have increased resistance to the soil-borne phytopathogen Rhizoctoniasolani. The antifungal properties of the barley seed RIP have been shown to be increased when in the presence of a chitinase and (1-3)-[3-glucanase (Leah et al. 1991). The synergistic inhibition of growth of Trichoderma reesi and Fusariumsporotrichiodes has been attributed to the chitinase and (1-3)-~-glucanase breaking down the cell wall, allowing entry of the barley RIP into the cell where it can then inactivate the ribosomes.

Leah et al. (1991) have speculated that the accumula- tion of an RIP in the endosperm of the barley seed could be responsible for the programmed senescence of this tissue at seed maturity. In view of the finding that rRNA depurination is undetectable in ribosomes isolated from endosperm tissue actively accumulating tritin-S, this pro- posal seems unlikely, but cannot be entirely excluded.

640

A.J.M. was the recipient of a U.K. Science and Engineering Research Council CASE studentship sponsored by Agricultural Genetics Company Ltd., Cambridge CB4 4GG, UK.

References

Anderson CW, Strauss JW, Duduck BS (1983) Preparation of a cell-free protein-synthesising system from wheat germ. Methods Enzymol 101:635-644

Barbieri L, Battelli MG, Stirpe F (1993) Ribosome-inactivating proteins from plants. Biochim Biophys Acta 1154:237-282

Bass HW, Webster C, O'Brian GR, Roberts JKM, Boston RS (1992) A maize ribosome-inactivating protein is controlled by the transcrip- tional activator opaque-2. Plant Cell 4:225-234

Bonness MS, Ready MP, Irvin JD, Mabry TJ (1994) Pokeweed antiviral protein inactivates pokeweed ribosomes: Implications for the antiviral mechanism. Plant J 5:173-183

Bradford M (1976) A rapid and sensitive method for the quantitation of microgram quantities of proteins utilising the principle of protein- dye binding. Anal Biochem 72:248-254

Carnicelli D, Brigotti M, Montanaro L, Sperti S (1992) Differential requirement of ATP and extra-ribosomal proteins for ribosome inactivation by eight RNA N-glycosidases. Biochem Biophys Res Com- mun 182:579-582

Carzaniga R, Sinclair L, Fordham-Skelton AP, Harris N, Croy RRD (1994) Cellular and subcellular distribution of saporins, type I ribosome-inactivating proteins in soapwort (Saponaria officinaIis L). Planta 194:461 470

Chaddock JA, Roberts LM (1993) Mutagenesis and kinetic analysis of the active site Glu177 of ricin A-chain. Protein Eng 6:425-431

Chen ZC, White RF, Antoniw JF, Lin Q (1991) Effect of pokeweed antiviral protein (PAP) on the infection of plant viruses. Plant Patho140: 612420

Coleman WH, Roberts WK (1982} Inhibitors of animal cell-free protein synthesis from grains. Biochim Biophys Acta 696:239-241

Di Fonzo N, Manzocchi L, Salamini F, Soave C (1986) Purification and properties of an endospermic protein of maize associated with the Opaque-2 and Opaque-6 genes. Planta 617:587-594

Endo Y, Tsurugi K (1987) RNA N-glycosidase activity of ricin A-chain: Mechanism of action of the toxic lectin ricin on eukaryotic ribosomes. J Biol Chem 262:8128-8130

Georgiev OI, Nikolaev N, Hadjiolov AA, Skryabin KG, Zaharyev VM, Bayer AA (1981) The structure of the yeast ribosomal ribonucleic acid genes. 4. Complete sequence of the 25S ribosomal ribonucleic acid gene from Saccharomyces cerevisiae. Nucleic Acids Res 9:6953-6958

Habuka N, Kataoka J, Miyano M, Tsuge H, Aga H, Noma M (1993) Nucleotide sequence of a genomic clone encoding tritin, a ribosome inactivating-protein from Triticum aestivum. Plant Mol Biol 22: 171-176

Hartley MR, Lord JM (1993) Structure, function and applications of ricin and related toxins. In: Grierson D (ed) Biosynthesis and manip- ulation of plant products. Chapman and Hall, Glasgow, pp 210-239

Hartley MR, Wheeler A, Ellis RJ (1975) Translation of the messenger RNA for the large subunit of fraction I protein in a heterologous cell-free system. J Mol Biol 91:67-77

Hartley MR, Legname G, Osborn R, Chen Z, Lord JM (1991) Single- chain ribosome-inactivating proteins from plants depurinate Escherichia coli 23S ribosomal RNA. FEBS Lett 290:65~8

Jackson AO, Larkins BA (1976) Influence of ionic strength, pH and chelation of divalent metals on isolation of polyribosomes from tobacco leaves. Plant Physiol 57:5 10

Kataoka J, Habuka N, Miyano M, Matsuta C, Koiwa A (1992) Adenine depurination and inactivation of plant ribosomes by an antiviral protein of Mirabilis jalapa (MAP). Plant Mol Biol 20:1111-1119

Katzin B J, Collins E J, Robertus JD (1991) Structure of ricin A-chain at 2.5 A Proteins 10:251-259

Laemmli UK (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680~585

Leah R, Tommerup H, Svendsen I, Mundy J (1991) Biochemical and molecular characterisation of three barley seed proteins with antifungal properties. J Biol Chem 226:1564-1573

Lodge JK, Kaniewski WK, Tumer NE (1993) Broad-spectrum virus resistence in transgenic plants expressing pokeweed antiviral protein. Proc Natl Acad Sci USA 90:7089-7093


Logemann J, Jach G, Tommerup H, Mundy J, Schell J (1992) Expression of a barley ribosome-inactivating protein leads to increased fungal protection in transgenic tobacco plants. Bio/Techno- logy 10:305-308

May M J, Hartley MR, Roberts LM, Krieg PA, Osborn RW, Lord JM (1989) Ribosome inactivation by ricin A-chain: a sensitive method to assess the activity of wild-type and mutant polypeptides. EMBO J 8: 301 308

Moazed D, Stern S, Noller HF (1986) Rapid chemical probing of conformation in 16S rRNA and 30S ribosomal subunits using primer extension. J Mol Biol 187:399-416

Morch MD, Drugeon G, Zagorski W, Haenni AL (1988) The synthesis of high molecular weight proteins in the wheat germ translation system. In: Weissbach A, Weissbach H (eds) Methods for plant molecular biology. Academic Press, San Diego, pp 91-101

O'Farrell PH (1975) High resolution two-dimensional electrophoresis of proteins. J Biol Chem 250:4007-4021

O'Farrell PZ, Goodman HM, O'Farrell PH (1977) High resolution two-dimensional electrophoresis of basic as well as acidic proteins. Cell 12:1113-1142

Prestle J, Sch6nfelder M, Adam G, Mundry K (1992) Type I ribosome- inactivating proteins depurinate plant 25S rRNA without species specificity. Nucleic Acids Res 20:3179-3182

Rau6 HA, Klootwijk J, Musters W (1988) Evolutionary conservation of structure and function of higher molecular weight ribosomal RNA. Prog Biophys Mol Biol 51:77-129

Ready MP, Brown DT, Robertus JD (1986) Extracellular localisation of pokeweed antiviral protein. Proc Natl Acad Sci USA 83:5053 5056

Reisbig RR, Bruland O (1983) The protein synthesis inhibitors from wheat, barley and rye have identical antigenic determinants. Biochem Biophys Res Commun 114:190-196

Roberts WK, SelitrennikoffCP (1986a) Isolation and characterisation of two antifungal proteins from barley. Biochim Biophys Acta 880: 161-170

Roberts WK, Selitrennikoff CP (1986b) Barley, rye, wheat and maize RIPs inhibit growth of protoplasts of Neurospora crassa. Biosci Rep 6:19 29

Roberts WK, Stewart TS (1979) Purification and properties of a translation inhibitor from wheat germ. Biochemistry 18:2615 2621

Robertus J (1991) The structure and action of ricin, a cytotoxic N- glycosidase. In: Lord JM (ed) Redirecting nature's toxins (Seminars in cell biology, vol 2), Saunders Company, Philadelphia, pp 23-30

Rothblatt JA, Meyer DI (1986) Secretion in yeast: reconstruction of the glycosylation of alpha-factor and invertase in a homologous cell-free system. Cell 44:619-628

Sambrook J, Fitsch EF, Maniatis T (1989) Molecular cloning: A laborat- ory manual, 2nd edn, Cold Spring Harbor, New York

Soave C, Tardani L, Di Fonzo N, Salamini F (1981) Zein level in maize endosperm depends on a protein under control of the opaque-2 and opaque-6 loci. Cell 27:403-410

Sperti S, Brigotti M, Zamboni M, Carnicelli D, Montanaro L (1991) Requirements for the inactivation of ribosomes by gelonin. Biochem J 277:281-284

Stewart TS, Hruby DE, Sharma OK, Roberts WK (1977) An ATP- dependent inhibition of protein synthesis in ascites cell extracts by wheat germ protein. Biochim Biophys Acta 479:31-38

Stirpe F, Bailey S, Miller SP, Bodley JW (1988) Modification of ribosomal RNA by ribosome-inactivating proteins from plants. Nucleic Acids Res 16:1349 1357

Taylor BE, Irvin JD (1990) Depurination of plant ribosomes by pokeweed antiviral protein. FEBS Lett 273:144-146

Taylor S, Massiah A, Lomonossoff G, Roberts LM, Lord JM, Hartley MR (1994) Correlation between the activities of five ribosome-inactivating proteins in depurination of tobacco ribosomes and inhibition of tobacco mosaic virus infection. Plant J 5:827-835

Towbin H, Staehelin T, Gordon J (1979) Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: Procedure and some applications. Proc Natl Acad Sci USA 76:4350-4354

Traub P, Mizushima S, Lowry CV, Nomura M (1971) Reconstitution of ribosomes from subribosomal components. Methods Enzymol 20 part C: 391-407

Walsh TA, Morgan AE, Hey TD (1991) Characterisation and molecular cloning of a proenzyme form of a ribosome-inactivating protein from maize. J Biol Chem 266:23422-23427

Documents

Wheat ribosome-inactivating proteins: Seed and leaf forms with different specificities and cofactor requirements