Isolation and Characterization of Ribosome-Inactivating Proteins from Cucurbitaceae

Isolation and Characterization of Ribosome-Inactivating Proteins fromCucurbitaceae

by Daoning Zhanga) and Fathi T. Halaweish*b)

a) Center for Biomolecular Structure and Organization, Department of Chemistry & Biochemistry,University of Maryland, College Park, MD 57007, U.S.A.

b) Department of Chemistry & Biochemistry, South Dakota State University, Box 2202, Brookings,SD 57007, U.S.A. (phone: þ1-605-688-4269; fax: þ1-605-688-6364;

e-mail: [email protected])

Due to their RNA-N-glycosidase activity, ribosome-inactivating proteins (RIPs) are attractivecandidates as antitumor and antiviral agents in biomedical and agricultural research. We have isolatedand characterized two such proteins, foetidissimin II and texanin, from two Cucurbitaceae species.Foetidissimin II, obtained from the roots of Cucurbita foetidissima, was identified as a type-2 RIP, with amolecular weight of 61 kDa, as estimated by gel electrophoresis. It is composed of two chains, a 29-kDachain A, and a 32-kDa chain B. Texanin, isolated from the fruits of Cucurbita texana, is a type-I RIP, witha single chain of molecular weight 29.7 kDa, as estimated by MALDI-TOF-MS. Both proteins exhibitRNA-N-glycosidase activity, with aniline playing a critical role in rRNA cleavage. The IC50 value offoetidissimin II, determined by cell-free protein-synthesis inhibition, was 0.251 mm. In an in vitrocytotoxicity assay, foetidissimin II exhibited IC50 values of ca. 70 nm to both adenocarcinoma anderythroleukemia cells. Texanin exhibited a weaker anticancer activity against erythroleukemia cells, withan IC50 value of 95 mm, but no activity against adenocarcinoma cells. The N-terminal sequences of bothproteins were compared with those of reported RIPs.

1. Introduction. – Ribosome-inactivating proteins (RIPs) are a group of proteinsbeing able to inactivate eukaryotic ribosomes through site-specific de-adenylation ofthe large ribosomal RNA [1]. They inactivate ribosomes by modifying both or either oftwo nucleoside residues, G4323 and/or A4326, in 28S rRNA (Fig. 1), consequentlycause the inactivation of ribosome and, eventually, the death of cells being invaded.This specific enzyme activity, termed RNA-N-glycosidase activity, distinguishes RIPsfrom other toxins in the way of damaging cells.

RIPs are widely distributed in the plant kingdom; they also exist in fungi, algae, andbacteria [2–4]. They are compartmentalized in vacuoles and cell walls, whichapparently allow the ribosome-inactivating activity to be sequestered from their ownribosomes [5] [6]. RIPs may be released or induced in response to pathogen infection orinjury, and several other functions have been proposed for their roles in plants, but aclear function of RIPs in plants has yet to be elucidated [7]. Three types of RIPs havebeen identified [2] [8] [9]. Type-1 RIPs contain a single polypeptide chain with amolecular weight between 26 and 32 kDa, have alkaline isoelectric points (pI> 9), andare often glycosylated. Type-2 RIPs are composed of two chains (chains A and B),linked with each other by a single disulfide (S�S) bond [2]. Chain A of type-2 RIPs

CHEMISTRY & BIODIVERSITY – Vol. 4 (2007) 431

G 2007 Verlag Helvetica Chimica Acta AG, ZIrich

possesses N-glycosidase activity, and chain B probably comprises a lectin-type chainassisting chain A to enter target cells [1]. Finally, type-3 RIPs are composed of twodomains that are directly linked with each other. One domain of type-3 RIPs possessesthe N-glycosidase activity; the function of the other domain is not known yet. All theRIPs are able to cleave a specific glycosidic bond on 28S rRNA. The action results inthe inactivation of the ribosome, further causing cell death.

Recent experiments established that RIPs can remove adenine from various nucleicacids other than rRNA [10]. This site-specific N-glycosidase activity has drawnattention from scientists because of the potential application of RIPs in medicine andplant protection [1] [11–13]. Much research has been undertaken to use RIPs directlyas antiviral and antitumor agents, or as the toxin part of immunotoxins in drugdevelopment [14–18]. RIPs are also an important topic in terms of bio-terrorism.Ricin, one of the first studied RIPs for therapeutic applications, has been listed by theUnited State Government as one of a few chemical agents that could be used byterrorists to pose threat to public because no vaccine or antitoxin against ricin has beenfound or developed.

Herein, we report the isolation, identification, and biological properties of twoRIPs, the type-2 compound foetidissimin II, and the type-1 compound texanin. In ourprevious work [19], foetidissimin, another type-2 RIP, had been isolated fromCucurbita foetidissima. Further investigation of this plant now led to the discovery offoetidissimin II. Texanin, in turn, was isolated from Cucurbita texana. Both RIPs wereidentified with the aid of an RNA-N-glycosidase activity assay. A cell-free protein-synthesis-inhibition assay was performed as well as an in vitro cytotoxicity assay toexplore their toxicities in different environments. Finally, N-terminal protein sequenc-ing was conducted to compare the amino acid sequences of our samples to those ofreported RIPs.

2. Results and Discussion. – 2.1. Isolation and Purification of Foetidissimin II.Foetidissimin II was extracted from the dried roots of C. foetidissima, and purified bymeans of a three-step chromatographic procedure (Fig. 2). Crude proteins were firstseparated by size-exclusion chromatography (Sephadex G25) into four groupsdesignated as CF1–CF4 (Fig. 2,a). Group CF2, which contained proteins rangingfrom 20–70 kDa in molecular weight (as determined by gel electrophoresis), wasfurther investigated since most RIPs have molecular weights of ca. 30–60 kDa. In a

CHEMISTRY & BIODIVERSITY – Vol. 4 (2007)432

Fig. 1. RNA-N-Glycosidase activity of RIPs

second step, anion-exchange chromatography was applied, because most RIPs areacidic proteins. First, a micro-scale separation was conducted with the help of theHiTrap IEX test kit to select a proper column medium (data not shown), and Q-Sepharose was finally chosen. Then, fraction CF2 was separated to afford threesubfractions, CF21, CF23, and CF25 (Fig. 2,b). Further purification of fraction CF23by means of gel filtration (Superdex-75 column) resulted in three further subfractions,CF231, CF232, and CF233 (Fig. 2,c), respectively.

Electrophoresis on sodium dodecylsulfate (SDS) gel showed that the amounts ofproteins in fractions CF231 and CF233 were below the detection limit, while CF232contained two protein bands, with molecular weights of 33 and 29 kDa, respectively(Fig. 3). Further analysis by analytical gel filtration indicated that CF232 contained aprotein with a molecular weight of ca. 61 kDa, designated as foetidissimin II, and

Fig. 2. Three-step chromatographic elution profiles for the purification of foetidissimin II from Cucurbitafoetidissima (CF). a) Size-exclusion chromatography (Sephadex G25) for group separation; b) anion-exchange chromatography (Q-Sepharose); c) gel-filtration chromatography (Superdex 75). For details,

see Exper. Part.


identified as a type-2 RIP using an RNA-N-glycosidase activity assay (see below). Thetotal yield of foetidissimin II was 10 mg/kg dry root powder.

The purification and characterization of foetidissimin II was conducted in analogyto our previous work, which had led to the isolation and characterization offoetidissimin, another type-2 RIP [19]. Both foetidissimin and foetidissimin II wereobtained from the dry roots of C. foetidissima, and purified by a three-stepchromatographic procedure. However, the isolation of foetidissimin II demandedmore effort due to its lower yield upon purification. Both foetidissimin andfoetidissimin II were separated by anion-exchange chromatography, the latter (inCF23) having a shorter retention time than the former (in CF25). This indicated thatfoetidissimin II was probably a less-acidic protein.

2.2. Isolation and Purification of Texanin. Texanin was extracted from the freshfruits of C. texana (Cucurbitaceae) growing in the south-central United States. Thecrude protein extract was first desalted on a SR 25/45 column packed with SephadexG25 by size exclusion. The separation profile is shown in Fig. 4. Two major proteingroups were collected, and designated asCT–G1 andCT–G2.Bradford tests indicatedthat the protein concentration in CT–G1 was much higher than in CT–G2. Andbarium chloride (BaCl2) test mapped out where ammonium sulfate was during thisseparation (tR 14–23 min, marked area in Fig. 4,a). This test displayed that thedesalting process was complete. Next, six fractions were collected upon anion-exchangechromatography of CT–G1, designated as CT1–CT6, as shown in Fig. 4,b. FractionCT1 showed N-glycosidase activity, and was further fractionated by gel filtration(Superdex 75 HR 10/30 column) to afford four subfractions, CT11–CT14. Theseparation was monitored by gel electrophoresis (Fig. 5,a), and CT13 (Fig. 4,c) wasfound to contain texanin, which was identified as a type-1 RIP, based on its N-glycosidase activity. This protein had a molecular weight of 29.7 kDa, as found byMALDI-MS, which is a little higher than it appeared to be on the gel. Fraction CT12was a mixture of texanin and another protein with a molecular weight of 33.8 kD. Theycould be successfully separated by means of analytical reverse-phase HPLC (Fig. 5,b).The yield of texanin was 5 mg/kg of plant material.

Fig. 3. SDS-PAGE Analysis of foetidissimin-II chromatographic fractions (see Fig. 2). Foetidissimin IIwas in fraction CF232.


2.3. Cell-Free Protein-Synthesis-Inhibition Assay. Foetidissimin II was tested with acell-free rabbit-reticulocyte-lysate-translation system to investigate whether its pres-ence would affect the luciferase-synthesis machinery in a cell-free environment.According to Fig. 6, foetidissimin II effectively reduced the yield of luciferase, with anIC50 value1) of 251.6 nm, as derived by graphic extrapolation. In this experiment theprotein appeared to be less toxic than the previously studied foetidissimin, which gaverise to an IC50 value of 26 nm (Table 1).

2.4. RNA N-Glycosidase Activity Assay. Cell-free protein-synthesis-inhibitionassays can suggest whether a sample inhibits the protein-synthesis machinery, but itcannot tell the mechanism of inhibition. We, thus, further conducted an RNA-N-

Fig. 4. Three-step chromatographic elution profiles for the purification of texanin from Cucurbita texana(CT). a) Size-exclusion chromatography (Sephadex G25) for protein desalting; b) anion-exchangechromatography (Q-Sepharose); c) gel-filtration chromatography (Superdex 75). For details, see Exper.

Part.


1) IC50 corresponds to the concentration causing 50% inhibition.


Fig. 5. a) SDS-PAGE Analysis of chromatographic fractions (see Fig. 4) from Cucurbita texana (CT)after b) reverse-phase HPLC separation of fraction CT1 containing texanin

Fig. 6. Cell-free protein-inhibition assay of foetidissimin II (fraction CF232). The inhibition activity wascalculated based on the yields of luciferase produced in the presence of various amounts of RIP.

glycosidase activity assay to verify ribosome-inactivating proteins. The activity wasmeasured by analyzing rRNA treated with RIP samples [1] [20] [21]. Both foetidissiminII and texanin exhibited N-glycosidase activities, with a 450-bp fragment appearingupon gel electrophoresis of the rRNA sample (Fig. 7). The experiment with texanindemonstrated better the role of aniline in the enzymatic cleavage reaction. The removalof adenine took place when 28S rRNAwas treated withCT13, consequently weakeningthe phosphodiester bond at the nearby 3’-end of the rRNA, but without cleaving thisbond. No RNA fragment was released at this stage, as evident from lane CT13�(Fig. 7,b). Only after being treated with aniline, the depurinated rRNA broke down,generating a 450-bp fragment (lane CT13þ ). Note that CT12 also contained texanin,but the result from the N-glycosidase activity assay of this fraction was ambiguous,probably because CT12 is a protein mixture with a relatively lower concentration oftexanin.

2.5. In vitro Cytotoxicity Assay. It is thought in general that type-2 RIPs are moretoxic to cells than type-1 RIPs because the former are assumed to have a lectin-typechain to assist the protein-entering cells [22] [23]. We tested foetidissimin II and texaninin two different tumor cell lines to compare their cytotoxicities. Foetidissimin IIexhibited the same level of in vitro cytotoxicity to both adenocarcinoma cells anderythroleukemia cells, with estimated IC50 values of 70�77 and 70�73 nm, respec-tively (Table 1). The IC50 value for texanin against erythroleukemia cells was 95�101 mm. Against adenocarcinoma cells, texanin was inactive in our experiment.Although the structural details of foetidissimin II and texanin are unknown yet, theevidently higher cytotoxicity of foetidissimin II towards these cell lines is probably dueto the presence of a lectin-like chain B.

Several purified RIPs are known to be cytotoxic, the published IC50 values varyingfrom the nanomolar to the millimolar range. Cinnamomin (with two chains) andcamphorin (with a single chain), obtained from the seeds of Cinnamomum camphora,produced inhibitory effects in several kinds of carcinoma cells, with IC50 values at thenanomolar level [24]. A type-1 RIP, sechiumin, obtained from the seeds of ediblegourd, showed an inhibitory effect in intact HeLa cells, with an IC50 value of 5 mm [25].The cytotoxicities of nigrin b and of ricin, two type-2 RIPs, were determined in a varietyof cancer cells [26]. Thereby, nigrin b was less cytotoxic by a factor of ca. 104–105

compared to ricin, with IC50 values ranging from 10�8 to 10�6m. Two aspects must be


Table 1. Characterization of Foetidissimins and Texanin

Foetidissimina) Foetidissimin II Texanin

Origin C. foetidissima C. foetidissima C. texanaSource root root fruitYield [mg/kg] 35 10 5Mr [kDa] 63 62 29.7RIP Type 2 2 1CF-PSIA [nm] 25.9 251 n.d.c)IC50 [nm] (adenocarcinoma) n.d. 70�77 not activeIC50 [nm] (erythroleukemia) n.d. 70�73 95�101

a) Data taken from [19]. b) Cell-free protein-synthesis-inhibition activity. c) Not determined.

considered to evaluate the cytotoxicity of type-2 RIPs. The first question to be posed iswhether the difference in chain B makes some RIPs entering cells easier, and thesecond is whether the structural difference in chain A gives rise to a higher enzymaticactivity in site-specific depurination of nuclear acids.

2.6.Homology Analysis of N-Terminal Protein Sequence. The N-terminal sequencesof foetidissimin II and texanin were analyzed and compared with those of reportedRIPs, using the database from the European Bioinformatics Institute. The first aminoacids, starting from the N-terminus of foetidissimin II, were determined asDVTFDLXGAT in chain A, and as NIQFDLSQA in chain B. The N-terminalsequence of texanin was determined to be NVRFDLSGATSSSYKT. Sequencesimilarities to other RIPs are shown in Table 2.

Foetidissimin II and foetidissimin, isolated from the same plant species, were foundto share 80% similarity in their N-terminal sequence. Thus, these two proteins are mostprobably isoforms. It is interesting to see that the B-chain of foetidissimin II also sharesa certain degree of similarity in N-terminal sequence to other RIPs. The N-glycosidaseactivity of chain B can be further examined when separated from chain A. In addition,


Fig. 7. N-Glycosidase activities of foetidissimin-II and texanin fractions analyzed by RNA gel electro-phoresis. a) rRNA treated with C. foetidissima (CF) fractions; lane CF232þ corresponds to RNAtreated with foetidissimin II. b) rRNA treated with C. texana (CT) fractions; lane CT13 representstexanin-treated RNA. Note: a plus (þ) indicates analysis after addition of aniline, and a minus (�) refers

to the experiment before addition of aniline (see Exper. Part).

the galactose-binding activity of chain B is worth investigating to determine whetherchain B is a lectin-like protein [1] [2].

As can be seen from Table 2, the first 16 amino acid residues of texanin (from C.texana) and of pepocin (from the sarcocarp of C. pepo [27]) are identical, indicatingthat these two proteins share a high degree of similarity. Belonging to theCucurbitaceae family, Cucurbita pepo is regarded as a domesticated squash andpumpkin species, and its domestication can be dated back to ca. 8,000 BC fromarcheological excavations in Mexico [28]. Pepocin is an RIP with a molecular weight of26 kDa, slightly smaller than texanin. Its elution profile upon anion-exchangechromatography indicates that pepocin has a higher isoelectric point than texanin.The genetic connection may account for the similarity of these two RIPs, but furtherstudies need to be performed to examine their structural resemblance.

3. Conclusions. – Two RIPs, foetidissimin II and texanin, were for the first timeidentified and isolated from C. foetidissima and C. texana. They were purified byconsecutive chromatographic methods, using a combination of size-exclusion and ion-exchange columns. Our results demonstrate that RIP treatment to ribosomal RNA onlyremoves adenine from 28rRNA, whereas aniline accounts for the cleavage of the RNAchain. Foetidissimin II was identified as a type-2 RIP, and exhibits cytotoxicity towardsboth adenocarcinoma and erythroleukemia cancer cells. Texanin was identified as atype-1 RIP, being less cytotoxic towards adenocarcinoma and erythroleukemia cells,probably due to lacking of lectin-like chain B facilitating RIP-entering cells. Homologycomparison of the N-terminal amino acid sequence suggested a high similarity of bothproteins to known RIPs.

The potential usage of RIPs has been mainly relied on their site-specific N-glycosidase activity. Direct tests of RIPs on virus or tumor cells have demonstratedtheir antiviral or antitumor activities, but great efforts are still required to make RIPsapplicable as antivirus and antitumor agents. On one hand, protein engineering mayprovide great opportunities to convert RIPs to highly efficient and selective


Table 2. N-Terminal Sequences of Selected RIPs. Partial-sequence identities are given relative tofoetidissimin II (A- and B-chains) or texanin, resp.

Protein Chain Sequence Identity [%]

Foetidissimin II A DVTFDLXGAT 100Foetidissimin A DVSFDLAGAT 80RIP from C. pepo NVRFDLSGAT 70Trichoanguina precursor DVSFDLSTAT 70Momorcochin S DVTFSLLGA 77.8Foetidissimin II B NIQFDLSQA 100RIP from C. pepo NVRFDLSGA 66.7Momordin-II precursor DVNFDLSTA 55.6Texanin NVRFDLSGATSSSYKT 100Pepocin NVRFDLSGATSSSYKT 100Trichobakin DVSFRLSGATSSSYGV 78.6Trichosanthin precursor DVSFRLSGATSSSYGV 78.6

immunotoxins, on the other hand, Nature is always an abundant source to search foranswers.

This study was supported by start-up funds for F. H. from the Department of Chemistry &Biochemistry at South Dakota State University (SDSU), and by a graduate assistantship from the SDSUNSF-EPSCoR fund. We thank Dr. Susan Rupp and Dr. Chris Chase from the Veterinary ScienceDepartment for assistance in conducting cell assays, Dr.Manisha Sonee and Dr. Suman Mukherjee fromthe College of Pharmacy, and Dr. Yen Yang from the Department of Biology for kindly providingfacilities and offering experimental assistance.

Experimental Part

General. Dried root powder ofCucurbita foetidissima (buffalo gourd) was purchased fromMicro FloCo. (Sparks, GA). The fresh fruits of Cucurbita texana were generously provided by Dr. D. Tallamy,University of Delaware. Prep. and semi-prep. columns and packing media were purchased fromAmersham Bioscience, including C16/70, C16/20, and HR10/30 columns, and Sephadex-G25, Q-Sepharose, and Superdex-75 column media. The cell-free, non-radioactive, in vitro rabbit-reticulocyte-lysate-translation system was purchased from Promega. Two different cell lines, adenocarcinoma anderythroleukemia, and cell-culture media as well as other related reagents were purchased fromAmericanType Culture Collection (ATCC). Protein-assay reagent and protein standards for chromatography werepurchased from Bio-Rad Laboratories. Electrophoresis markers and all other chemicals required werepurchased from Sigma-Aldrich. The following aq. buffers were used for extraction and purificationprocedures (unless noted otherwise): buffer A : 50 mm sodium phosphate (pH 7.5) containing 50 mm

NaCl; buffer B : 5 mm aq. NaHCO3 (pH 9).Protein Extraction. All experiments were performed at a temp. of 48. Foetidissimin II was isolated

from the dried roots ofC. foetidissima. Thus, dried, powdered root material (100 g) was homogenized in ablender with 1 l of buffer A. Insoluble substances were removed by centrifugation (Fisher ScientificMarathon 21000 multi-purpose refrigerated centrifuge) at 3,760g for 30 min, followed by filtration usingan ADVANTEC pressure filter. Proteins in soln. were further precipitated by adding ammonium sulfateto 80% saturation, and redissolved in buffer A (pH 7.5). Texanin was isolated from fresh fruits of C.texana in analogy to the method described above, but starting from 400 g of fruits.

Chromatographic Separation, and Concentration and Molecular-Weight Determination of RIPs.Separation was conducted in three steps. The first step consisted of a group separation using a C16/70column packed with Sephadex G25. The sample vol. was 10 ml, and the flow rate was 2 ml/min of bufferA. AC16/20 column packed with Q-Sepharose fast-flow medium was used in the second step under thefollowing conditions: flow rate, 2 ml/min; gradient, 0–10 min buffer B (5 mm NaHCO3 (pH 9)), 10–80 min linear gradient of 0–1m NaCl in buffer B. The final separation was conducted on an HR10/30 column packed with Superdex 75, eluting with buffer A. The loading vol. was 1 ml, and the flow ratewas 0.3 ml/min.

In the purification of texanin, the first step was modified by switching to an SR25/45 column packedwith Sephadex G25. A simple BaCl2 test was carried out to estimate the efficiency of the Sephadex-G25-packed SR25/45 column in desalting the crude protein extract. Thus, an aliquot (100 ml) of the eluate fromeach tube was mixed with 1m aq. BaCl soln. (100 ml) to detect sulfate by means of ammonium sulfateprecipitation. Reverse-phase chromatography was also applied in the final purification. A sample in 0.1%aq. TFA (CF3COOH) was loaded on a Prosphere C18 column (300 N, 5 mm; 250�4.6 mm; sampleloading: 1 mm), and eluted with a linear gradient of TFA/H2O/MeCN 0.1 :29.9 :70 from 0% at 2 min to100% at 45 min, at a flow rate of 0.5 ml/min. The detection wavelength was 280 nm.

Protein concentrations were determined bymeans of theBradford assay using bovine serum albumin(BSA) as standard [29]. Sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS-PAGE) wasperformed according to the Laemmli protocol [30]. Non-reducing PAGE was also used to preserve theS�S bond between two RIP subunits. A Might-Small-II system (Hoefer, San Francisco) was used, and10% acrylamide was prepared for the analysis. The molecular standards used were bovine lung aprotinin(6.5 kDa), bovine milk a-lactalbumin (14.2 kDa), soybean trypsin inhibitor (20.0 kDa), bovine pancreas


trypsinogen (24.0 kDa), bovine erythrocyte carbonic anhydrase (29.0 kDa), glyceraldehyde-3-phosphatedehydrogenase (36.0 kDa), chicken egg ovalbumin (45.0 kDa), and BSA (66.0 kDa). CoomassieBrilliant Blue was used to stain the proteins. Molecular weights were also determined by MALDI-TOF-MS on a Bruker Biflex IV apparatus. All the protein samples were desalted by dialysis or size exclusionchromatography (Sephadex G25), all experiments being conducted according to standard proceduresprovided by the manufacturer.

Cell-Free Protein-Synthesis-Inhibition Assay. A cell-free, non-radioactive, in vitro rabbit-reticulo-cyte-lysate protein-synthesis system was adapted to evaluate the inhibition activities of the isolatedprotein fractions [31] [32]. Purified RIP was added to the protein-synthesis system at variousconcentrations, and the inhibition activity of the RIP was deduced from the yield of protein synthesized.Thus, protein sample (5 ml) at 10, 2, 1, and 0.5 mg/ml in PBS buffer was added to the protein-synthesisreaction mixture, which was composed of rabbit reticulocyte lysate (17.5 ml ; 70% of reaction volume), a1 mm Oall amino acidP mixture (0.25 ml ; except Leu), a 1 mm Oall amino acidP mixture (0.25 ml ; exceptMet), 2.5m KCl (0.7 ml), RNasin ribonuclease inhibitor (0.5 ml), and luciferase control RNA (0.5 ml ;1 mg/ml). The mixture was incubated at 308 for 90 min. The reaction was terminated by placing thereaction tube on ice. The reaction product, luciferase, was diluted with BSA (1 mg/ml) by a factor of5,000. An aliquot of the resulting soln. (2 ml) was mixed with luciferase assay reagent (25 ml), and theluminescence intensity of the mixture was measured with a Beckman LS6500 scintillation counter. Basedon the luminescence intensity, the reaction yield of luciferase was calculated, and then the inhibitionactivity of the isolated protein was evaluated.

RNA-N-Glycosidase Activity Assay. This assay was conducted to directly identify the two RIPs.Rabbit reticulocyte lysate (70 ml) was mixed with ribosome protein soln. (12.5 ml, 1 mg/ml), together withthe other reagents, to obtain a final reaction mixture containing 62.5 mm KCl, 2.5 mmMgCl2, and 0.05%2-mercaptoethanol. The mixture was incubated at 378 for 15 min, the reaction was stopped by addition ofice-cold 10% SDS (5 ml), and the RNA in the reaction mixture was extracted with a Promega SV total-RNA-isolation system. To selectively cleave the 28S rRNA, the extracted RNA (50 ml) was incubated atpH 4.5 on ice for 10 min with 0.8m acidic aniline soln. (13.2 ml), prepared from AcOH (24 ml), aniline(15 ml), and deionized H2O (111 ml). The cleavage reaction was stopped by adding H2O (200 ml), and theaniline was removed by evaporating the sample to dryness at ambient temp. in vacuo. The RNA wasfinally analyzed by gel electrophoresis (1.2% agarose) at 80 V for 60 min [20] [21].

In vitro Cytotoxicity Assay. The activities of the purified protein fractions were assayed onadenocarcinoma (ATCC No. CCL-2) and erythroleukemia (ATTC No. CRL-2451) cell lines.Adenocarcinoma cells were cultured in MEM medium, and erythroleukemia cells were cultured inRPMI-1640 medium. The cells were harvested, and adjusted to a concentration of 5�105 cells/ml. Theassay was conducted in a 96-well flat-bottom tissue plate. All the wells were first filled with cellsuspension (180 ml). Then, RIP soln. (20 ml) was added to the first row of wells, the starting RIPconcentration being high, as long as the protein could be dissolved completely. Then, a serial dilution wasmade to provide a total of ten sample concentrations, the last two rows being controls with only cellsuspensions. The cells were then incubated for 72 h at 378 in a humified atmosphere (5% CO2, 100% rel.humidity). Then, cell population was examined under a microscope, and IC50 values were calculated toevaluate the RIPPs cytotoxicity against the tumor cells.

N-Terminal Amino Acid Sequencing. The appropriate protein sample was subjected to SDS-PAGEin the presence of 2-mercaptoethanol to dissociate the A- and B-chains, and electro-blotted onto PVDFmembranes. Detection was achieved by staining with 0.2% (w/v) Coomassie Blue R-250 in MeOH/H2O/AcOH 50 :40 :10 for 0.5–2 min. The membranes were washed with MeOH/H2O/AcOH 48 :47 :5, theprotein bands were cut out of the PVDF, and subjected to sequencing (Applied Biosystems PreciseSequencer) according to the manufacturerPs instructions. The rel. degree of sequence similarity amongvarious RIPs was compared by searching the EMBL database of the European Bioinformatics Institute.


REFERENCES

[1] Y. Endo, K. Mitsui, M. Motizuki, K. Tsurugi, J. Biol. Chem. 1987, 262, 5908.[2] F. Stirpe, L. Barbieri, M. G. Battelli, M. S. Soria, D. A. Lappi, Bio/Technology 1992, 10, 405.[3] T. Girbes, J. M. Ferreras, F. J. Arias, F. Stirpe, Mini Rev. Med. Chem. 2004, 4, 461.[4] T. B. Ng, W. Y. Chan, H. W. Yeung, General Pharmacol. 1992, 23, 575.[5] J. Kataoka, N. Habuka, M. Furuno, M. Miyano, Y. Takanami, A. Koiwai, J. Biol. Chem. 1991, 266,

8426.[6] J. M. Vivanco, B. J. Savary, H. E. Flores, Plant Physiol. 1999, 119, 1447.[7] W. J. Peumans, Q. Hao, E. J. M. V. Damme, FASEB J. 2001, 15, 1493.[8] H. W. Bass, C. Webster, G. R. OPBrian, J. K. M. Roberts, R. S. Boston, Plant Cell 1992, 4, 225.[9] B. Chaudhry, F. Muller-Uri, V. Cameron-Mills, S. Gough, D. Simpson, K. Skriver, J. Mundy, Plant J.

1994, 6, 815.[10] L. Barbieri, P. Valbonesi, M. Govoni, A. Pession, F. Stirpe, Biochim. Biophys. Acta 2000, 1480, 258.[11] J. W. Uhr, Cell Biol. 1991, 2, 1.[12] B. A. Parikh, N. E. Tumer, Mini. Rev. Med. Chem. 2004, 4, 523.[13] L. Barbieri, A. Bolognesi, F. Stirpe, Methods Mol. Biol. 2001, 166, 71.[14] C. Y. Zhang, Y. X. Gong, H. Ma, C. C. An, D. Y. Chen, Analyst 2000, 125, 1539.[15] L. Citores, J. M. Ferreras, R. Munoz, J. Benitez, P. Jimenez, T. Girbes, Cancer Lett. 2002, 184, 29.[16] J. L. Lambert, V. S. Goldmacher, A. R. Collinson, L. M. Nadler, W. A. Blattler, Cancer Res. 1991, 51,

6236.[17] M. S. McGrath, K. M. Hwang, S. E. Caldwell, I. Gaston, K.-C. Luk, P. Wu, V. L. Ng, S. Crowe, J.

Daniels, J. Marsh, T. Deinhart, P. V. Lekas, J. C. Vennari, H.-W. Yeung, J. D. Lifson,Proc. Natl. Acad.Sci. U.S.A. 1989, 86, 2844.

[18] A. Bolognesi, L. Polito, P. L. Tazzari, R. M. Lemoli, C. Lubelli, M. Fogli, L. Boon, M. D. Boer, F.Stirpe, Br. J. Haematol. 2000, 110, 351.

[19] D. Zhang, F. T. Halaweish, Plant Sci. 2003, 164, 387.[20] T. Girbes, L. Citores, J. M. Ferreras, M. A. Rojo, R. Iglesias, R. Munoz, F. J. Arias, M. Calonge, J. R.

Garcia, E. Mendez, Plant Mol. Biol. 1993, 22, 1181.[21] K. Y. Tsang, T. B. Ng, Life Sci. 2001, 68, 773.[22] G. R. Thrush, L. R. Lark, B. C. Clinchy, E. S. Vitetta, Annu. Rev. Immunol. 1996, 14, 49.[23] I. I. Pastan, R. J. Kreitman, Adv. Drug Deliv. Rev. 1998, 31, 53.[24] J. Ling, W. Y. Liu, Cell Biochem. Funct. 1996, 14, 157.[25] T. H. Wu, L. Chow, J. Y. Lin, Eur. J. Biochem. 1998, 255, 400.[26] R. Munoz, Y. Arias, J. M. Ferreras, P. Jimenez, M. A. Rojo, T. Girbes, Cancer Lett. 2001, 167, 163.[27] S. Yoshinari, S. Yokota, H. Sawamoto, S. Koresawa, M. Tamura, Y. Endo, Eur. J. Biochem. 1996, 242,

585.[28] H. Wilson, OTexas Endemics: Distribution of Cucurbita texanP, 1998 (see http://www.csdl.tamu.edu/

FLORA/flcp/flcp3.htm).[29] M. Rosenberg, OProtein Analysis and Purification: Benchtop TechniquesP, BirkhRuser, Boston, 1996.[30] U. K. Laemmli, Nature 1970, 227, 680.[31] P. E. Thorpe, A. N. F. Brown, W. C. J. Ross, A. J. Cumber, S. I. Detre, D. C. Edwards, A. J. S. Davies,

F. Stirpe, Eur. J. Biochem. 1981, 116, 447.[32] Promega, Flexi Rabbit Reticulocyte Lysate System, Technical Bulletin No. 127, 2006.

Received December 8, 2006


Documents

Isolation and Characterization of Ribosome-Inactivating Proteins from Cucurbitaceae