Ribosome-inactivating and Adenine Polynucleotide Glycosylase Activities in Mirabilis jalapa L. Tissues*

Several tissues of Mirabilis jalapa L. (Nyctaginaceae) were assayed for inhibition of translation by a rabbit reticulocyte lysate (as a signal of ribosome-inactivating activity) and for adenine DNA glycosylase activity, activities that are both due to the presence of a class of enzymes called ribosome-inactivating proteins (RIPs), currently classified as rRNA N -glycosylases (EC 3.2.2.22). These activities were highest in seed; interme-diate in flower bud, immature seed, sepal (cid:1) gynoecium, leaf, and root; and very low in all other tissues. By cat-ion-exchange chromatography, four protein peaks with inhibitory activity on cell-free translation were identified in extracts from seeds, and two proteins were isolated from peaks 1 and 4, all of which have the properties of single-chain type 1 RIP. One is Mirabilis antiviral protein (MAP), so far purified only from roots. The second is a new protein that we propose to call MAP-4. The distribution of MAP and MAP-4 in several tissues was determined with a novel experimental approach based on liquid chromatography/mass spectrometry. The direct enzymatic activity of MAP Experimental details are given under “Experimental Procedures.” One unit of RIP activity is defined as the amount of protein necessary to reduce translation by 50% in 1 ml of a rabbit reticulocyte lysate system under the present experimental conditions. One unit of APG activity is defined as the amount of protein necessary to release 1 (cid:1) mol of adenine from DNA under the present experimental conditions. Values have been normalized to activity in roots, taken as 100%. Values for root tissue were as follows: for crude extract, 9261 (cid:7) 1177 (mean (cid:7) S.D.) and 1.55 (cid:7) 0.18 units/g of starting tissue for RIP and APG activities, respectively; and for basic protein fraction, 12,400 and 1.06 for RIP and APG activities, respectively. Values refer to the mean of the results of three different crude extracts or to a single basic protein fraction. Samples were all run in duplicate, and controls with all reagents but no incubation were run in each experiment.

Ribosome-inactivating proteins (RIPs) 1 from plants may be classified as type 1 or 2 according to their single-or doublechain structure (reviewed in Refs. [1][2][3][4]. Besides the classical type 1 and 2 RIPs, a 60-kDa RIP (called JIP60) has been identified in barley (Hordeum vulgare) that consists of an ami-no-terminal domain closely related to the RIP enzymatic chain linked to an unrelated carboxyl-terminal domain with unknown function (5); this protein may be classified as type 3 RIP (3). The mechanism of action of this class of proteins became clearer when it was found that ricin and subsequently all RIPs tested release a single adenine residue from ribosomes in a precise position (A 4324 in the case of rat liver ribosomes) of a universally conserved GAGA sequence in a peculiar stem-loop structure (review in Ref. 2). They were thus classified as rRNA N-glycosidases (EC 3.2.2.22). Subsequently, it was observed that some RIPs release more than one adenine residue from ribosomes, that others act on RNA species apart from ribosomal RNA and on poly(A), and that all RIPs release adenine from DNA. Thus, the enzymatic activity of RIP was defined as polynucleotide:adenosine glycosidase (6), which we propose to change to adenine polynucleotide glycosylase (APG) in analogy with the EC nomenclature of nucleic acid glycosylases.
Two main biological properties of RIPs, viz. (i) inhibition of multiplication of plant viruses (reviewed in Ref. 4) and (ii) extremely potent cytotoxicity (reviewed in Ref. 1), upon entry into eukaryotic cells, led to several applications. In agriculture, plants were transfected with RIP genes, viz. barley RIP, pokeweed antiviral protein (PAP), trichosanthin, and dianthin, to confer resistance to viruses and fungi (reviewed in Ref. 4). In traditional Chinese medicine, trichosanthin and momordin have been used as abortifacient agents (reviewed in Ref. 1). RIPs are currently under study as therapeutic agents against cancer (review in Ref. 7) and possibly HIV infection (8) after linkage to antibodies (immunotoxins) or other specific carrier molecules to make them selectively toxic to a given type of target cells. One of the unsolved problems in the clinical use of immunotoxins is the immune response elicited against both the mouse monoclonal antibody and the toxic moiety that prevents repeated administrations. This problem can be partially circumvented by the use of immunotoxins prepared with human or humanized antibodies and different RIPs that do not crossreact with each other. Thus, the availability of a set of noncross-reacting RIPs should be highly valuable in this kind of therapeutic strategy.
A potent antiviral activity was found in extracts from a yellow flower cultivar of Mirabilis jalapa L. (Nyctaginaceae) in root, leaf, and stem tissues and in in vitro cultured cells (9,10). From the roots of M. jalapa, a protein was then purified that was highly effective in preventing viral infection caused by contact-transmitted virus (11). This protein, named Mirabilis antiviral protein (MAP), was later identified as a RIP (12,13) for its activity on the major rRNA in intact ribosomes.
Little is known about the distribution of RIPs in plant organs and tissues (14,15) and very little in M. jalapa (9). Here, we describe the distribution of both translation inhibitory activity (RIP activity) and, for the first time in any plant, APG activity in the different organs of M. jalapa. From the seeds of this plant, which contain the highest levels of both activities, two RIP isoforms were purified, and the most abundant one was characterized. This protein is identical to the isoform purified from root tissue, MAP (12,16), and has never been characterized for APG activity. In plants, RIPs may be present with many isoforms in several tissues (PAPs in Phytolacca americana (reviewed in Ref. 17), saporins in Saponaria officinalis (14), and luffins in Luffa cylindrica (15)), often with different yields and biological properties. In this study, the distribution of the two major isoforms of MAP in six tissues was determined, and the hitherto unknown APG activity and immunological properties of MAP are described.

EXPERIMENTAL PROCEDURES
Materials-M. jalapa (red flower cultivar) tissues were collected from plants grown in the garden of the Dipartimento di Patologia Sperimentale, Università di Bologna. L-[U-14 C]leucine and L- [4, H]leucine were from Amersham Biosciences (Buckinghamshire, United Kingdom). Materials for low-pressure chromatography, including calibrating substances, were from Amersham Biosciences (Uppsala, Sweden). Adenine, tRNA, and electrophoresis markers were from Sigma. Poly(A), genomic RNA from tobacco mosaic virus (TMV), and rRNA from Escherichia coli (16 S ϩ 23 S, M r ϭ 1.75 ϫ 10 6 ) were from Roche Molecular Biochemicals (Mannheim, Germany). Cell culture medium and supplements and all other chemicals were as described in previous work (18). Sera against various RIPs were a gift from Dr. P. Strocchi (University of Bologna). Chloroacetaldehyde was prepared according to McCann et al. (19). All other reagents were of analytical or molecular biology grade and, when possible, RNase-free. Milli-Q water (Millipore Corp., Milford, MA) was used when applicable. DNA from herring sperm (hsDNA) (Sigma) was mechanically sheared and made RNA-free by treatment with DNasefree RNase A (Roche Molecular Biochemicals) for 2.5 h at 37°C. DNA was then repeatedly precipitated in ethanol to remove the enzyme. Genomic DNA from M. jalapa leaves, prepared following the general procedure described by Ausubel et al. (20), was precipitated with isopropyl alcohol, resuspended in 10 mM Tris-HCl (pH 8.0) containing 1 mM EDTA, twice phenol-extracted, precipitated with ethanol, and resuspended in the same buffer. DNA was then subjected to mechanical shearing and RNase treatment as described for hsDNA. Poly(A) ϩ RNA from Bryonia dioica leaves was obtained by extraction on oligo(dT)cellulose as described by Ausubel et al. (20).
Preparation of Crude Extracts and Basic Protein Fractions-Fresh plant materials were frozen in liquid nitrogen, ground in a mortar, and homogenized with an Ultraturrax apparatus (Ika, Staufen, Germany) in cold phosphate-buffered saline (0.14 M NaCl and 5 mM sodium phosphate buffer (pH 7.5)). Mature seeds and sepals ϩ gynoecia were homogenized in 8 ml and all other tissues in 5 ml of cold phosphatebuffered saline/g of starting material. The slurries were extracted overnight at 4°C with magnetic stirring, filtered through cheesecloth, and clarified by centrifugation at 10,000 ϫ g for 30 min at 4°C (crude extracts). When appropriate, crude extracts were adjusted to pH 4.0 with glacial acetic acid and centrifuged again at 10,000 ϫ g for 30 min at 4°C. The supernatant was applied to an SP-Sepharose Fast Flow column (15 ϫ 2.5 cm) equilibrated with 10 mM sodium acetate (pH 4.5) at room temperature. The column was washed with 1 volume of the equilibration buffer and then extensively with 5 mM sodium phosphate buffer (pH 7.0), and bound proteins were eluted with 1 M NaCl in the same buffer. Eluted protein was either dialyzed exhaustively against water at 4°C (for further preparative processing) or concentrated at 4°C in an Amicon concentrator equipped with a PM-10 membrane under nitrogen pressure (4 bars) and magnetic stirring and then desalted by chromatography on Sephadex G-25. These preparations are referred to as basic protein fractions and contained most, if not all, RIPs present in the starting material in all plant species and tissues assayed (21).
Purification of Ribosome-inactivating Proteins-The basic protein fraction from 100 g of M. jalapa seeds was adjusted to 5 mM phosphate buffer (pH 7.0) and applied to a CM-Sepharose Fast Flow column (42 ϫ 1.6 cm) equilibrated with the same buffer at room temperature. The column was washed with the equilibration buffer until the absorbance at 280 nm was lowered to the base line and was eluted with 1 liter of a linear (0 -300 mM) gradient of NaCl in the same buffer (see Fig. 1). Fractions appropriately diluted in phosphate-buffered saline were as-sayed for translation inhibitory activity by a rabbit reticulocyte lysate. Active fractions from peaks indicated as MAP, MAP-2, MAP-3, and MAP-4 were pooled, dialyzed extensively against water at 4°C, and stored frozen at Ϫ80°C.
Analytical Methods-Proteins obtained by cation-exchange chromatography were analyzed for purity and molecular mass by gel filtration, SDS-PAGE under reducing conditions, and reverse-phase HPLC as described (22). Chromatographic conditions were as follows: protein C 4 reverse-phase column (250 ϫ 4.6 mm; Vydac) equipped with the appropriate pre-column and equilibrated and eluted at 20°C at 1 ml/min; solvent A, 0.1% trifluoroacetic acid in water; and solvent B, 0.1% trifluoroacetic acid in acetonitrile. The column was equilibrated with 90:10 solvent A/solvent B and loaded with 100 l of protein sample in water. Bound material was eluted with a linear gradient of solvent B up to 40% in 10 min and then to 60% in 50 min. The effluent was split; 5% was analyzed by electrospray ionization mass spectrometry (single quadrupole ZMD, Micromass, Manchester, United Kingdom); and molecular mass was calculated by the maximum entropy algorithm (MaxEnt, Micromass). The remaining effluent was analyzed at 214 nm in a Kontron spectrophotometric monitor. The isoelectric point was determined with Phast system and PhastGel IEF 3-9 (Amersham Biosciences) following the manufacturer's instructions. N-terminal sequencing was performed as described (23). The E 280 nm,1 cm 1% of purified MAP was determined with water solutions of freeze-dried samples. The cross-reactivity of MAP with antibodies against other RIPs was measured with an enzyme-linked immunosorbent assay as described previously (22).

Identification of MAP Isoforms in Basic Protein Fractions from M. jalapa Tissues-
The identification of MAP isoforms in basic protein fractions was performed by LC/MS as described above. Protein was applied at ϳ100 g in 200 l of phosphate-buffered saline.
In Vitro Inhibition of Protein Synthesis by Cell Cultures-Murine 3T3 (fibroblasts) and human HeLa (carcinoma), NB100 (neuroblastoma), and BeWo (choriocarcinoma) cells were maintained as monolayer cultures in RPMI 1640 medium supplemented with antibiotics and 10% fetal calf serum in a humidified atmosphere containing 5% CO 2 at 37°C. Subcultures were obtained by trypsin treatment of confluent cultures. The JM cell line (human monocyte-derived) was grown in suspension and treated with phorbol myristate to induce adhesion as described by Bolognesi et al. (22). Protein synthesis by cells was determined as described by Ferreras et al. (14). Other experimental details are described in the legend to Fig. 4.
In Vitro Translation by a Rabbit Reticulocyte Lysate-The effect of protein from M. jalapa on translation in a cell-free system (a rabbit reticulocyte lysate) was studied essentially as described by Parente et al. (23). Reaction mixtures contained, in a final volume of 62.5 l, 10 mM Tris-HCl (pH 7.4), 100 mM ammonium acetate, 2 mM magnesium acetate, 1 mM ATP, 0.2 mM GTP, 15 mM phosphocreatine, 3 g of creatine kinase, 0.05 mM amino acids (minus leucine), 89 nCi of L-[U-14 C]leucine, and 25 l of rabbit reticulocyte lysate. Incubation was at 28°C for 5 min.
Determination of Adenine Polynucleotide Glycosylase Activity-The enzymatic activity of the purified protein was determined by measuring adenine released from various substrates by HPLC (24) essentially following the procedure of McCann et al. (25) as described by Barbieri et al. (26). Reactions were run for 40 min at 30°C in a final volume of 50 l containing 100 mM KCl, 50 mM sodium acetate (pH 4.0), increasing concentrations of RIPs, and the indicated amounts of polynucleotide substrates. A standard curve of adenine was run with each experiment. The determination of bases other than adenine was performed as described (26). The determination of adenine released by crude extracts or basic protein fractions was done by LC/MS because of interference of compounds present in crude preparations with the derivatization step of the method described above. Briefly, the reaction was stopped in ice by the addition of 100 l of ice-cold 10 mM ammonium acetate, and reagents were separated by solid-phase extraction on Bond Elute-NH 2 minicolumns (Varian) equilibrated with 10 mM ammonium acetate as described (26). Adenine was measured by LC/MS on a Waters Alliance/zq apparatus. Chromatography to separate adenine was carried out on a Waters XTerra MS C 18 column (2.1 ϫ 50 mm, 2.5-m beads) equilibrated and eluted with 10 mM ammonium acetate (solvent A)/ methanol (solvent B) at 0.3 ml/min at 15°C. Equilibration was in 98:2 solvent A/solvent B; and after sample injection (120 l), the column was eluted with equilibration solvents for 2 min and then with 90:10 solvent A/solvent B for 5 min. Tightly bound material was eluted with 20:80 acetonitrile/solvent A for 0.6 min, and equilibration was attained with 90:10 solvent A/solvent B for 1.2 min, followed by 98:2 solvent A/solvent B for 9 min. Mass spectrometric analysis was carried out in positive electrospray with a single-ion recording (m/z 135 ϩ 1) on a split flow of ϳ50 l/min. Parameters were optimized manually for maximum sensitivity under the present column elution conditions. Duplicate chromatograms were combined to reduce noise using Micromass MaxLynx software. Samples containing standard adenine (from 1 to 300 pmol) were incubated and processed together with the experimental samples. A standard curve run with each experiment was fit by linear regression analysis, and adenine released from experimental samples was determined by plotting against the standard curve.

RESULTS
Tissue Distribution-The tissue distribution of RIP and APG activities was assayed both in crude extracts from most M. jalapa tissues and in basic protein fractions from root, mature shelled seed, immature unshelled seed, green leaf, flower bud, and sepal ϩ gynoecium (Table I). The highest activity was found in mature seed.
Purification from Seeds-The basic protein fraction from 100 g of seeds was loaded onto a CM-Sepharose column. Bound proteins were eluted with a linear NaCl gradient, and the inhibitory activity was resolved into four protein peaks (Fig. 1). The first eluted protein, which is identical to MAP from roots (see below), and the fourth protein (MAP-4) appeared to be Ͼ98% homogeneous upon reverse-phase HPLC (Fig. 2), and both gave a single band upon SDS-PAGE (data not shown). MAP was also pure when analyzed by gel filtration under nondenaturing conditions (data not shown).
The inhibitory activity and protein yield at the various steps of purification are summarized in Table II. It should be noticed that the total activity recovered after the acidification and SP-Sepharose steps was higher than that present in the original crude extract. This has been observed previously with other RIPs (27) and possibly is due to removal of an inhibitor or to "activation" of the M. jalapa ribosome-inactivating proteins.
Identification and Quantification of MAP and MAP-4 in Tissue Protein Extracts-MAP and MAP-4 were identified by LC/MS in all basic protein extracts examined, with the exception of root extracts, in which only MAP was detected ( Fig. 3 and Table III). MAP-2 and MAP-3, identified in preparative ion-exchange chromatographic fractions, were below detection limits in this assay.
Physicochemical Properties-LC/MS molecular masses (Ϯ2) were 27,788 (MAP), 30,412 (MAP-2), 29,771 (MAP-3) and 29,339 (MAP-4) Da. The first isoform that eluted from CM-Sepharose differs from MAP from roots described in the literature at only four amino acids, which are not involved in the putative active site (16), and has a pI Ͼ9, like that of most other RIPs; and its E 280 nm 1% is 7.06. Immunological Properties-MAP gave no reaction with antisera specific for seven other RIPs (bouganin, dianthin-32, momordin I, momorcochin-S, PAP-R, saporin-S6, and trichokirin).
Effects on Protein Synthesis-The RIP activities of MAP isoforms in a rabbit reticulocyte lysate are reported in Table IV. Inhibition of protein synthesis by various cell lines (Fig. 4) was observed at concentrations of MAP much higher than those effective on cell-free protein synthesis. The effect varied greatly from one cell line to another, with the IC 50 values ranging by more than an order of magnitude from the most resistant HeLa cells to the most sensitive JM and BeWo cells.
Adenine Polynucleotide Glycosylase Activity-MAP released adenine in a concentration-dependent manner from all substrates tested, viz. DNA from herring sperm and M. jalapa, genomic TMV RNA, poly(A) ϩ RNA from B. dioica, E. coli rRNA, and poly(A) (Fig. 5). Under the present experimental conditions, autologous DNA appeared to be the best substrate, with 24.4 mol of adenine released per mol of enzyme/min, followed TABLE I Distribution of RIP and APG activities in M. jalapa tissues Experimental details are given under "Experimental Procedures." One unit of RIP activity is defined as the amount of protein necessary to reduce translation by 50% in 1 ml of a rabbit reticulocyte lysate system under the present experimental conditions. One unit of APG activity is defined as the amount of protein necessary to release 1 mol of adenine from DNA under the present experimental conditions. Values have been normalized to activity in roots, taken as 100%. Values for root tissue were as follows: for crude extract, 9261 Ϯ 1177 (mean Ϯ S.D.) and 1.55 Ϯ 0.18 units/g of starting tissue for RIP and APG activities, respectively; and for basic protein fraction, 12,400 and 1.06 for RIP and APG activities, respectively. Values refer to the mean of the results of three different crude extracts or to a single basic protein fraction. Samples were all run in duplicate, and controls with all reagents but no incubation were run in each experiment.  by hsDNA (16.3 mol/mol/min), whereas other substrates were less sensitive (0.6 mol/mol/min for TMV RNA, 0.5 mol/mol/min for poly(A) ϩ RNA, 0.4 mol/mol/min for poly(A), and 0.35 mol/ mol/min for rRNA) at the lowest enzyme concentration assayed (Fig. 5). The APG activities of MAP isoforms on hsDNA are reported in Table IV, with MAP-4 showing the highest activity.

DISCUSSION
Methodology-Two innovations in the methodological approach were fundamental in obtaining the results described here: the use of LC/MS for (i) the determination of adenine and (ii) the identification and quantification of enzyme isoforms in partially purified extracts. These different approaches allow for the direct measurement of RIPs, as protein or as activity, in rather crude extracts, thus without the inevitable artifacts induced by variable purification yields. Furthermore, the possibility of detecting and identifying specific proteins in preparations subjected to a very limited treatment (basic protein fractions) allows for the determination of isoform distribution in many tissues, even if present in very low quantities.
Distribution-The distribution of RIP activity in the various tissues of a given plant has been described only for very few species. The antiviral properties of some tissue extracts from M. jalapa were studied, and the highest activity was found in roots (9), from which a RIP denominated MAP was purified (11). Here, we have described the distribution, in M. jalapa anatomical parts, of both main activities attributed to the group of plant enzymes provisionally called ribosome-inactivating proteins and so far classified as rRNA N-glycosylases, viz. inhibition of in vitro translation (RIP activity) and APG activity on various polynucleotides. The last activity has not been described so far in partially purified enzyme preparations due to the interference of several substances present in crude plant extracts with the highly sophisticated methodology involving derivatization of released adenine to its fluorescent derivative, ethenoadenine. Both RIP and APG activities were determined in crude extracts and partially purified samples containing basic proteins. These last preparations were chosen for two main reasons. (i) Recovery of RIP activity was often complete after batch-wise cation-exchange chromatography (27), and (ii) inconsistent results were obtained in measuring glycosylase activity in crude extracts due to interference of nucleases that degrade the substrate and other substances that inhibit enzymatic activity. Distribution of RIP activity was similar in crude extracts and basic protein fractions, whereas it was confirmed that glycosylase activity was sometimes hindered in several crude extracts, as could be inferred by the activities found in basic protein fractions (Table I). The distribution pattern in basic protein fractions is similar for both activities. At least in M. jalapa tissues, it may be said that all proteins with APG activity are also ribosome-inactivating proteins. The tissue with the highest activities was mature seed, in good agreement with what was found in the unrelated plant S. officinalis (family of Caryophyllaceae) (14). The distribution of RIP activity in M. jalapa organs is similar to that of saporins in S. officinalis, the only other plant in which RIP distribution has been most thoroughly studied (14). The distribution of activity described here differs somewhat from MAP contents estimated by Kubo et al. (9) with an enzyme-linked immunosorbent assay using anti-MAP antibodies. This difference may be due to the cultivar (red versus yellow flower), to the growing environmental conditions of the plants, or to non-homogeneous reaction with the antisera of the different MAP isoforms. No clear difference in activity was observed in senescent leaf tissue, in contrast to what was observed in other species (28).
Purification-From seeds, we purified the most abundant protein with RIP and APG activities together with three other isoforms. The most abundant RIP was analyzed by LC/MS, and the molecular mass obtained was very similar to that described for the major form found in and purified from roots (16,29): only four amino acids were different, none of them in putative regions implicated in enzymatic activity (reviewed in Ref. 1). This result could be due either to the presence in seeds of a different isoform or to a microvariability between subspecies (i.e. yellow versus red flower cultivar) of the same protein gene. To solve this problem, we purified, by reverse-phase HPLC, the main RIP from roots from the same cultivar we utilized for seeds. The retention time and molecular mass were absolutely identical to those of the protein purified from seeds. Thus, we demonstrated that the main isoform of MAP is present also in seed and that there are minor subspecies variations in the sequences of the same isoforms. The second abundant isoform from seed (MAP-4) was isolated also from leaf tissue by reverse-phase HPLC and identified by retention time and molecular mass. This highly purified protein was sequenced in the N-terminal portion. The identification of this form as a RIP was confirmed by the presence of the invariable residues Tyr, Arg, and Phe, which are totally conserved in all amino-terminal sequences of RIPs reported so far (Fig. 6).
Immunological Properties-MAP is not recognized by antibodies against several commonly used RIPs. Thus, this protein, which is easily purified in large enough quantities, may be useful to prepare immunologically distinct immunotoxins to overcome the immune response caused by the in vivo administration of these compounds.

TABLE II
Purification of RIP from M. jalapa seeds Results refer to 100 g of seeds. Activity was determined as inhibition of translation by a rabbit reticulocyte lysate as described under "Experimental Procedures." IC 50 is the concentration of protein that inhibited protein synthesis by 50% in a rabbit reticulocyte lysate system. One unit is the amount of protein causing 50% inhibition in 1 ml of reaction mixture. The CM-Sepharose elution profile is shown in Fig. 1   Analysis of Isoforms-Several tissues were then assayed for the presence of the two major isoforms, MAP and MAP-4. The substantial difference in N-terminal sequence (Fig. 6) indicates that they are products of different genes and not just minor variations due to non-significant random mutations. Both isoforms were found in all tissues examined, with the exception of root, in which only MAP could be detected. The absolute amount of each protein and their relative proportions varied from tissue to tissue; thus, it may be inferred that the expression of the relative genes is tissue-regulated. Comparison with the RIP isoforms purified from the closely related plant Mirabilis expansa shows that there is a strict similarity between pairs of isoforms from the two plants (MAP/ME2 and MAP-4/ ME1), as shown in Fig. 6, a similarity that is much greater than that of the isoforms from the same plant (very low apart from the three invariable Tyr, Arg, and Phe residues). This may allow the classification of the two different pairs as different classes of RIPs, as happens in S. officinalis (14). All isoforms have little similarity to bouganin, a RIP isolated from Bougainvillea spectabilis, another member of the Nyctaginaceae family (18). Work is in progress to verify whether MAP-4 has enzymatic and substrate specificities different from those of MAP. The existence in the same plant of several forms of RIPs, sometimes functionally different, has been frequently observed (e.g. PAP from P. americana, reviewed by Irvin (17); and saporins from S. officinalis (14)).
Enzymatic Activities-So far, MAP has been assayed only for activity in translation systems and purified animal ribosomes. Here, we have reported the determination of the direct enzymatic activity of MAP on several substrates, including autologous DNA. MAP depurinated all substrates assayed, including deoxy-and ribonucleic acids. It should be noted that autologous DNA was one of the best substrates; this is the first demonstration of activity of a RIP on the DNA of its own plant. Furthermore, no bases other than adenine were released from hsDNA (data not shown), and Ͼ1 mol of adenine/mol of enzyme was released from all substrates without the need for any cofactor, required by some other RIPs to act efficiently on ribosomes (30). MAP belongs to the restricted group active on poly(A) (6). These results indicate that MAP acts catalytically and does not require a highly restricted sequence on purified substrates. Moreover, (i) the activity on poly(A), observed so far Cytotoxicity-MAP inhibits protein synthesis more efficiently in a cell-free system than in whole cells, like all type 1 RIPs that are internalized with low efficiency. Toxicity varied depending upon the cell type tested, with HeLa cells being the least sensitive, as it is usually with other RIPs, and JM and BeWo cells being the most sensitive. As JM and BeWo cells are derived from monocytes and choriocarcinoma cells, respectively, both with a high pinocytotic activity, their high sensitivity could be the consequence of a better cellular uptake of the protein, although differences in intracellular routing cannot be excluded. All lethally intoxicated cells showed the morphological features of cell death by apoptosis (data not shown), as previously described for some other RIPs (31,32). Cytotoxicity of RIPs has long been attributed entirely to protein synthesis inhibition; however, direct damages to DNA and/or RNA other than ribosomal RNA may also have a role in the induction of apoptosis in xenobiotic cells.
Implications for the Biological Role-Some variations in RIP distribution and content in M. jalapa are to be noted. (i) The levels of RIP and APG activities are 8-fold higher in mature as compared with immature seeds, as observed in S. officinalis (14); and (ii) the ratio between MAP and MAP-4 is in favor of MAP in organs with storage tissues (root and seed), whereas MAP-4 is the prevalent isoform in leaf-related tissues. These variations in activity content point to the question of the function of RIPs in plants, which is still not clear. The notion was put forward that they may be storage proteins and/or defense systems (33); and indeed, they may confer some protection against a broad spectrum of viruses and fungal pathogens, as observed in transgenic plants expressing RIPs (reviewed in Ref. 4). Nevertheless, RIP expression rises in mature seeds and in stressed (5,28,34) and virally infected (35) plant tissues, conditions that may all require programmed cell death activation. RIPs may have a primary role in the mechanisms leading to apoptosis. This could be through the effect on ribosomes, which occurs in vivo (36). However, as discussed above, MAP and other RIPs release adenine from DNA more efficiently than from RNA. Should this occur in vivo, it could be a very efficient mechanism to kill a cell because the few alterations sufficient to induce lethal damage to DNA could be caused more rapidly than the inactivation of the relatively high number of ribosomes required to significantly impair protein synthesis.