STUDIES ON RIBOSOME-INACTIVATING PROTEINS FROM MOMORDICA CHARANTIA
By TSE MAN FAI B. SC. (Hon), C. U. H. K.
.� \ ^ *
A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENT FOR THE DEGREE OF MASTER OF PHILOSOPHY IN BIOCHEMISTRY
JULY 1997 DEPARTMENT OF BIOCHEMISTRY THE CHINESE UNIVERSITY OF HONG KONG � ^^ / Y_ \ V /i,�:大 x^ A/统系馆書圖\A pf 2 2 m 1998 ]|j ^Xr UN!VERSiTY jM X^VvLIBRARY %r^m/^/J ^^^/ ACKNOWLEDGMENTS First of all I wouM like to express my most sincere gratitude to my supervisor Dr. Hin-Wing Yeung for his support throughout the project. I would like to give special thanks to Dr. Wing-Ping Fong, Dr. Tzi-Bua Ng and Dr. Chi-Cheong Wan for their advice and support, and their continuous guidance throughout my study. I would also like to thank my external examiner, Professor Eric J. Lien. I also wish to thank all staff and students in the Department of Biochemistry for their advice and help during my postgraduate study. Last, but not the least, I would also like to thank my family and friends for their encouragement during my study.
I ABSTRACT Two new ribosome-inactivating proteins (Rffs), designated 5- and s- momorcharin (6- and s-MMC), were isolated from the seeds and fruits of Momordica charantia, respectively. The method for the isolation of a- and P_ momorcharin (a- and P-MMC) was adopted for the isolation of these two new REPs. Li this method, an affinity column of Affi-Gel Blue gel was used to remove most of the non-RP like proteins from the crude extract. A fast protein liquid chromatography (FPLC) column, Mono S, was used to resolve proteins adsorbed on the affinity column. 6- and s-MMC were shown to possess molecular weights of 30 kDa and 24 kDa, respectively. They were also shown to have A^-glycosidase activity, which is characteristic of other Rffs. Like other Rffs, 5-MMC was highly potent in inhibiting cell-free translation in rabbit reticulocyte lysate with a 50 % inhibitory concentration (IC50) at 0.15 nM. On the other hand, s-MMC was much less potent. Its IC50 was shown to be 0.17 |iM. The yields of 5- and s-MMC were low. From 5 g of seeds, 34 ^ig of 5-MMC were isolated. From 500 g of fruits, 62 p,g of s-MMC were isolated. The reason for the low content of these two Rffs in their tissues is unknown. Further research is required.
n LIST OF ABBREVLVTIONS APA : Abrus precatorius agglutinin a-MMC : alpha-momorcharin P-MMC : beta-momorcharin ConA : Concanavalin A DEPC : Diethyl pryocarbonate DNase : Deoxyribonuclease DTH : Delayed-type hypersensitivity EP : Electrophoresis buffer FPLC : Fast protein liquid chromatography y-MMC : Gamma-momorcharin HTV : Human immunodeficiency virus kb : Kilobase kDa : Kilodalton LPS : Lipopolysaccharide IC50 : 50 % inhibitory concentration IT : Lnmunotoxin MAP30 : Momordica anti-HTV protein 30 MWCO : Molecular weight cut-off NAD : Nicotinamide adenine dinucleotide NK : Natural killer PHA : Phytohaemagglutinin PVDF : Polyvinylidene difluoride Rff : Ribosome-inactivating protein RCA : Ricinus communis agglutinin RNase : Ribonuclease SDS-PAGE : Sodium dodecyl sulphate-polyacrylamide gel electrophoresis sRn> : Small RH> SV : Simian virus TCS : Trichosanthin TEMED : N’ N, N, iV-tetramethyl ethylene diamine VAA : Viscum album agglutinin
m Table of Contents
ACKNOWLEDGMENTS I ABSTRACT n LEST OF ABBREVLiTIONS DDE TABLE OF CONTENTS 1 CHAPTER 1 n^TRODUCTION 2 1.1 RffiOSOME-mACnVATTNG PROTEDSfS (REPS) 2 1.1.1 Classification ofRIPs 2 1.1.2 Distribution ofRIPs 6 1.1.3 Molecular biology of RIPs 7 1.1.4 Physical and chemical properties ofRIPs 9 1.1.5 Enzymatic and translation-inhibitory activities 12 1 • 1.6 RIP-based Immunotoxins 16 1.2 MOMORDICA CHARANTLi AND FTS RmOSOME-^CnVATD^G PROTEDsfS (REP) 17 1.2.1 Momordica charantia 17 1.2.2 Ribosome-inactivating proteins (RIPs) in Momordica charantia 18 1.3 OBJECTTVE OF TfflS STUDY 28 CHAPTER 2 STUDY ON A NEW mBOSOME-ESACTrVAXmG PROTEBV (Rff) FROM MOMORDICA CHARANTLi SEEDS 30 2.1 LSTTR0DUCTI0N 3 0 2.2 MATERL\LS AND METHODS 33 2.2.1 Materials 33 2.2.2 RIP isolation 34 2.2.3 Characterization 35 2.3 RESULTS 42 2.4 DISCUSSION 48 CHAPTER 3 STUDY ON A NEW RffiOSOME-EVACTrVATEVG PROTEEV (RBP) FROM MOMORDICA CHARANTM FRUITS 51 3.1 blTRODUCTION 51 3.2 MATERLVLS AND METHODS 53 3.2.1 Materials 53 3.2.2 RIP isolation 54 3.2.3 Characterization 56 3.3 RESULTS 56 3.4 DISCUSSION 62 CHAPTER 4 GENERAL DISCUSSION AND CONCLUSION 64 4.1 GENERAL DISCUSSION 64 4.2 CONCLUSION 72 REFERENCES 74
1 Chapter 1 Introduction
1.1 Ribosome-inactivating proteins ^RIPs) 1.1.1 Classification ofRIPs Ribosome-inactivating proteins (RBPs) are a group of toxic proteins widely distributed in plants and they inhibit protein translation inside cells. RJPs are classified into four types, namely, type I，type H, type TV and a novel type called small RJPs. Type I RJPs are single-chain proteins with a molecular weight of around 30 kDa. They are the most abundant type among the four types (Table 1.1 and 1.2). Type H RPs are double-chain proteins which contain an active A chain and a cell-binding B chain linked together through disulphide linkage or hydrophobic interactions (Fig. 1.1). The A chain is similar to the type I PJPs and possesses the activities of RJPs. Its molecular weight is between 26 - 32 kDa. The B chain is a glycosylated protein with a molecular weight of around 34 kDa. It is a lectin which enables it to bind to galactosyl- terminated receptors on the cell surface. As a result, the active A chain can enter cells much more easily than type I RCPs and so type H RJPs are more toxic than type I RJOPs. There are some relatively non-toxic type H RJDPs. Ebulin 1 (Girbes et al., 1993b) and nigrin b (Girbes et al., 1993a) have toxicity at nanomolar levels, as
2 high as that of other type H RJDPs on cell-free translation system. But, they are relatively non-toxic towards intact cells in comparison with other type E RJDPs. This led to further classification into toxic and non-toxic type E RPs (Citores et aL, 1993) (Table 1.3).
Table 1.1 Families and species of plants with type I RIPs and some examples ofthe purified type I RJPs QBarbieri et aL^ 1993). Family Genus and species Name ofRJP Asparagaceae Asparagus officinalis asparin 1 asparin2 Caryophyllaceae Agrostemma githago agrostin2 agrostin 5 agrostin 6 Dianthus barbatus dianthin 29 Dianthus caryophyllus dianthin 30 dianthin 32 Lychnis chalcedonica lychnin Petrocoptis glaucifolia petroglaucin Saponaria officinalis saporin-Ll saporin-L2 saporin-Rl saporin-R2 saporin-R3 saporin-S5 saporin-S6 saporin-S8 saporin-S9 Cucurbitaceae Bryonia dioica bryodin-L bryodin-R Citrullus colocynthis colocin 1 colocin 2 Cucumis melo melonin Luffa acutangula luffaculin Luffa cylindrica luffin a luffmb (to be continued on p.4)
3 Table 1.1 (continued) Family genus and species Name ofREP Cucurbitaceae Momordica charantia a-momorchain P-momorchain Momordica cochinchinensis momorcochin-S Trichosanthes kirilowii trichosanthin TAP29 trichokirin a-kirilowin P-kirilowin Trichosanthes cucumeroides p-trichosanthin Euphorbiaceae Croton tiglium crotin Gelonium multiflorum gelonin Hura crepitans Hura crepitans RBP Jatropha curcas curcin 2 Manihot palmata mapalmin Manihot utilissima manutin 1 manutin 2 Nyctaginaceae Mirabilis jalapa MAP Phytolaccaceae Phytolacca americana PAP Phytolacca dioica PD-S1 PD-S2 PD-S3 Phytolacca dodecandra dodecandrin dodecandrin-C Poaceae Hordeum vulgare barley Rff Secale cereale Secale cereale RP Triticum aestivum tritin \Zea mays |maize RP
Table 1.2 Families and species of plant with type H REPs and some examples of purified type n RJPs CBarbieri et al” 1993). Family Genus and species Name ofRJP Euphorbiaceae Ricinus communis ricin Ricinus agglutinin Fabaceae Abrus precatorius abrin PassifIoraceae Adenia digitata modeccin Adenia volkensii voUcensin Sambucaceae Sambucus ebulus ebulin 1 Viscaceae Phoradendron californicum P. californ. lectin Viscum album viscumin
4 TypeimP TypeHRff
AH chain B chai n Rg/. 1.1 Structures oftype I and type H REPs. Type rV REPs are tetrameric. They resemble type E RBPs but each type IV REP has two A chains and two B chains. The molecular weights ofA chain and B chain are around 32 kDa and 34 kDa, respectively. Type IV RPs are agglutinins. They can also be classified into toxic and non-toxic ones (Citores et al., 1993) (Table 1.3). Ricinus communis agglutinins (RCA), Abrus precatorius agglutinins (APA) and Viscum album agglutinins fVAA) are examples oftype W RJCPs. Table 1.3 Classification ofRgs. Type Sub-classes Examples References I normal Asparin 1，Agrostin 2，a- (Barbieri et al., 1993) momorcharin, Crotin H toxic Ricin, abrin (Barbieri et al., 1993) non-toxic Ebulin 1，nigrin b (Girbes et al., 1993a; Girbes et al., 1993b) IV toxic Ricinus communis agglutinin (Citores et al., 1993) (RCA), Abrus precatorius agglutinin (APA) non-toxic Viscum album agglutinin (VAA) (Citores et al., 1993) Small RIP toxic LufBn S, g-MMC (Gao et al., 1994； Pu et al., 1996)， ，
Recently, there are some single-chain Rffs found to have a molecular weight of around 10 kDa. They are classified as small RJCPs (sRff). Luffm S from
5 Luffa cylindrica (Gao et al., 1994) and g-momorcharin from Momordica charantia (Zheng et al., 1996) are examples of sREP.
1.1.2 Distribution ofRIPs iUPs are widely distributed in the plant kingdom. It is found in both dicotyledonae and monocotyledonae of Angiospermae. It has not yet been found in Gymnosperaiae. Type I RJDPs are much more abundant than type H REPs (Table 1.1). Type n Rffs can only be found in five families (Table 1.2). Li a single plant speices, it is possible that more than one form of Rff exist. Examples are the RPs found in Phytolacca americana (pokeweed) OBarbieri et al., 1982; Lrvin，1975; L:vin et al., 1980)，Bryonia dioica (bryony) 03arbieri et al., 1989; Gasperi-Campani et al., 1980; Stirpe et al., 1986) and Saponaria officinalis (Ferreras et al., 1993)，which produce many different isoforms of type I Rffs. Many RJDPs found so far are located in the seeds fBarbieri et al., 1993). They can also be found at various parts of a plant, even at the latex (Stirpe et al., 1983). The different forms ofRJPs from Phytolacca americana, Bryonia dioica and Saponaria officinalis mentioned above are found in the seeds, roots and leaves of the plants OBarbieri et al., 1982; Barbieri et al., 1989; Ferreras et al., 1993; Gasperi-Campani et aL, 1980; L:vin, 1975; Lrvin et aL, 1980; Stirpe et al.,
6 1986). The type H Rffs modeccin and voUcensin can be found in roots and seeds (Gasperi-Campani etal” 1978; Stirpe etal., 1985). RPs are not only found in higher plants. They are also found in bacteria. Shiga toxin was purified from Shigella dysenteriae type I. Shiga-like toxin I (SLT-I) and Shiga-like toxin H (SLT-E) were purified from enterohaemonphagic Escherichia coli. These baterial toxins have the same enzymatic properties of type I RDPs (Strockbine et aL, 1986). ]n fungi, Aspergillus giganteus produces the toxin a-sarcin with specific RNase activity on the 28S rRNA ^)ndo and Wool, 1982). Aspergillus restrictus produces restrictocin and mitogillin (Fando et al., 1985). a-Sarcin, restrictocin and mitogillin cleave the phosphodiester bond between G4325 and A4326. It is very near the site of action of conventional RJDPs on the eukaryotic ribosome. Their effect on cells also involves protein synthesis inhibition.
1.1.3 Molecular biology ofRIPs The majority of the mRNA sequences ofRffs code for iV-terminal and C- terminal peptides which are post-translationally cleaved to form the mature RJP CFig 1.2). Most of the REP genes contain no introns in their genomic sequences. One exception is the gene ofMAP from M jalapa (Kataoka et al., 1993). Maize b-32 RP has an internal sequence in the nascent polypeptide This internal sequence is cleaved to form the mature protein (Fig 1.2). The TV-terminal signal
7 sequences are hydrophobic which resemble the signal sequences of other eukaryotic and prokaryotic proteins (Kataoka et aL, 1991).
Leader C-terminal ^ Mature peptide ~^^^ Dianthin30 23 252 18 Saporin-S6 24 253 22 MAP 28 250 --
Leader Unker C-terminal |jj^ ^peptide ^jj p-peptide J^^ Maize RIP 16 141 25 71 44
Leader Unker mmi A chain jJj|| B chain Ricin D 24 267 12 262 Ricinu3 agglutinin 24 266 12 262 Abrin 34 241 14 276
Fig. 1.2 Structures of prepro-forms of some type I and H RBPs. MAP=Mirabilis antiviral proteins. Maize REP is a type I RCP; but unlike other type I MPs, it is synthesied with the a- and P-peptides linked by a Unker sequence. Leader sequence, linker and C-terminal are absent in mature REPs. The number of amino acids in each peptide chain is indicated.
The feature of the iV-teraiinal signal sequence enables the biosynthesis of RJPs to be compatible with their enzymatic activity. It is because the prepro-forai of RJDPs may be inactive so that it would not damage its own ribosomes. This N- terminal sequence can direct the prepro-form to the endoplasmic reticulum OER).
8 The terminal sequences can be cleaved to form the active protein only at the ER. Other features, such as insensitivity to homologous ribosome may also contribute to the compatibility of toxicity to biosynthesis. The biosynthesis of ricin and Ricinus communis agglutinin (RCA) is well elucidated OHarley and Lord, 1985; Lord, 1985a; Lord, 1985b; Roberts and Lord, 1981). The A^-terminal signal sequences of the prepro-forms direct the proteins to the endoplasmic reticulum. Then, the signal sequences are removed to form the pro-form. Glycosylation and di-sulphide bond formation occur at the ER as well. The proteins are then transported by Golgi apparatus and Golgi vesicles to the protein bodies. Then, an acid proteinase processes the pro-form to form the mature RJPs.
1.1.4 Physical and chemical properties ofRIPs The molecular weights of type I REPs are around 30 kDa (Table 1.4). The A chains of type H Rffs also have similar molecular weights as type I Rffs and the B chains are about 34 kDa in molecular weight (Table 1.5). The majority of the REPs found so far are glycosylated. But there are some notable exceptions. For example, trichosanthin and abrin A chain are non- glycosylated. The percentage of glycosylation is different for different Rffs (Tables 1.4 & 1.5). The sugars in Rffs are glucose, galactose, mannose, fucose, xylose and A^-acetyl-D-glucosamine (Barbieri et al., 1993).
9 Table 1.4 Molecular weights of some type 1 RIPs ^Barbieri et al； 1993) RIPs Molecular Weight (kDa) Glycosylarion (%) Agrostin 5 29.5 6.87 Asparin 1 30.5 * Barley RIP 30.0 ** Bryodin (from root) 30.0 6.3 Colocin 2 26.3 1.59 Crotin 2 30.2 0.41 Curcin 2 28.1 糾 Dianthin 30 29.5 1.56 Dodecandrin 29.0 0 Gelonin 30.0 ** Hura crepitans RIP 27.5 0.94 LuffacuIin 28.0 * Luffm a 28.2 0 Maize RIP 25.0 0 Mapalmin 32.3 5.99 MAP 27.8 0 Momorcochin-S 30.7 2.82 a-Momorcharin 29.1 1.74 PAP 29.0 0 PD-S2 32.0 2.22 Saporin-S6 29.5 0 Trichokirin 27.0 1.27 Trichosanthin 26.0 0 Tritin ^ ^ * Glycosylation exists but amount is unknown ** Glycosylation is unknown
10 Table 1.5 Molecular weight & glycosylation of some type H RBPs flBarbierigM/.，1993) Type II RIPs Molecular weight (kDa) Glycosylation (%) Abrin c 63.8 7.4 A chain 30.0 0 B chain 36.0 7.4 Ebulin 1 56.0 ** A chain 26.0 ** B chain 30.0 ** Modeccin 60.0 2.7 A chain 27.8 * B chain 31.0 * Ricin D 63.0 * A chain 30.0 4.5 B chain 33.0 * Viscumin 60.0 11.8 A chain 29.0 * B chain 32.0 * Volkensin 62.0 5.7 A chain 29.0 * B chain 36^ * * Glycosylation exists but amount is unknown ** Glycosylation is unknown
Glycosylation of the type I RJPs and A chains of type H Rffs may be non- essential to the activity. It is because the type and percentage of glycosylation vary from one RP to another. Rffs such as trichosanthin and abrin A are non- glycosylated. When gelonin and ricin A chain are partially deglycosylated, their activities are not affected (Barbieri et aL, 1993). More striking is that when recombinant ricin A chain is produced in E. coli and is not glycosylated, it is still fullyfimctional (Barbier i et al” 1993).
11 Type I REPs are invariably basic proteins. Their pI values are higher than 7. Usually, their pI values are equal to or higher than 9 (Gelfi et al., 1987). The pI values of the A chains of type H and type W RBPs range from 4.6 to 8.6 (Barbieri etal； 1993).
1.1.5 Enzymatic and translation-inhibitory activities The characteristic property of RBPs is their A^-glycosidase activity. It was elucidated by Endo and colleagues in 1986 OEndo et al., 1990). RTPs attack the 28S rRNA. As the rRNA is located at the junction between the 60S and 40S subunit which interacts with elongation factors EF2 and the mRNA for successful translation, destruction of rRNA by REPs would result in failure of translation.
Rff target a-sarcin target
C G ^V I G / A G
� / ...ribose - p - ribose - p - rioose - p - ribose - p ^bose - p - ribose ... •.....•..••...•. < ..•...•..•....•••••…• A G..---" U G G A A A C C U C C—G G_C G_C C_G U C—G A_U 4308 A_U 4339 Rat1 282 S rRNA Fig. 1.3 The site of cleavage by the A^-glycosidase activity ofRIPs (Stirpe etal” 1992). Li the 28S rRNA, there is a highly conserved stem and loop structure. REPs specifically depurinate the A4324 on the rat liver ribosome (Fig. 1.3). The cleavage leads to the failure of GTP-dependent binding ofEF-2, EF-2-dependent hydrolysis ofGTP and binding ofEF-l. As a result, protein synthesis is inhibited. The activity is very potent. It inhibits the ribosome at nanomolar concentrations (Table 1.6). The mechanism of RJPs' iV-glycosidase activity was proposed by Monzingo and Robertus (1992) (Fig. 1.4). This mechanism was applied to ricin A chain. But this may also be applied to other iUPs ^VIonzingo and Robertus, 1992). Ln this mechanism, there is oxycarbonium ion formation. There are two variations. Li the first one ^Fig. 1.4), the target adenosine is stacked between two aromatic side chains tyrosine 80 and 123 (not shown). Arginine 180 protonates the adenine to form an oxycarbonium ion which is stabilized by the negatively charged glutamate 177 and the hydrogen bonding from glycine 121 and valine 81. After protonation, the arginine 180 attacks a nearby water molecule. The activated water molecule attacks the oxycarbonium ion and releases the adenine leaving group. Another proposed mechanism is similar. But, the water molecule is attacked by the hydrogen-bonded oxycarbonium structure instead of the arginine 180 in the first mechanism. The final structure is that the N9 becomes covalently bound to a hydrogen instead of the N3.
13 A. B. G121 C. G,21
3 X ...A �V^ �V^ >-。《邊，.、4 )�01, ^ )�ct>._..
*>:。德:《 T^ >>'" A. B. C. 0121 ^ , ^ J^j ElT7 J • ...0^" c.77 0-..., ..--°^^ Ein °---w""°Jl^
\。终....二 >^#^i^Z
^->-树>"、>-
Fig. 1.4 The proposed mechanism of N-gIycosidase activity of RBPs OMonzingo and Robertus, 1992). The top mechanism has N3 being protonated. The low mechanism has N9 being protonated. Please refer to the text for details.
As a consequence ofthe 7V-glycosidase activity, RJDPs inhibit the translation in cell-free and whole cell systems (Barbieri et al., 1993). The values of 50 % inhibitory concentration fECso) of some REPs on the translation of rabbit reticulocyte lysate and HeLa cells are shown in Table 1.6. Their IC50 values are as low as at 0.06 nM in the cell-free system. Apart from eukaryotic ribosomes, REPs can also inhibit prokaryotic ribosomes. Bacterial ribosomes, like those from E. coli, are very sensitive to type
14 I REPs (Habuka et al., 1990). As a result, expression of type I RJCPs in E. coli is often limited. But, on the contrary, ricin, a type H RJP, does not depurinate the E. coli ribosome (Cammarano et al., 1985) and so, it can be expressed in E. coli. Table 1.6 The inhibitory effect of some RBPs on protein synthesis in a ceU-free system (rabbit reticulocyte lysate) and a whole cell system CHeLa ceUs) and their toxicity to mice QBarbieri et al； 1993), RBPs Cell-Free IC^Q Whole cells Mouse LD50 ^ IC50 (nM) (mgAcg)* TypeI Agrostin 5 0.47 9200 1.0 Asparin 1 0.27 >3300 20 Bryodin-L 0.09 >3300 >10 Colocin 1 0.04 >3300”* 10.7 Crotin 2 0.48 >3300 ND Curcin 2 0.19 >3300 ND GeIonin 0.4 >3300 40 Hura crepitans REP 0.1 2000 ND Lychnin 0.17 >3300 9.3 Manutin 1 0.03 210 ND Mapalmin 0.05 >3300 >8.0 a-momorcharin 0.06 >3300 7.4 PAP 0.07 3400 0.95 PD 0.06 1620 1.12 Saporin S6 0.037 610-2340 ^ Type n Abrin 88 0.0039 0.00056 A chain 0.5 >0.4 Ricin 84 0.0011 0.0026 A chain 0.1 >0.4 Modeccin 45 0.0003 0.0023 A chain 2.3 Viscumin 43.3 0.008 0.0024 A chain 3.5 Volkensin 84 0.0123 0.00138 A chain 0^ *LD50 = 50 0/0 lethal dose to mice; ND = not determined.
The frnigal ribosomes are generally sensitive to Rffs fBradley et al., 1987). Ribosomes of fungi like S. cerevisiae and N. crassa can be inhibited by RPs with high efficiency (Bradley et aL, 1987).
15 Although many Rffs are inactive towards their own (autologous) ribosomes, it was shown that they can be active on autologous ribosomes at high concentrations O^restle etal” 1992). ti short, the activity of REPs on different types of ribosomes are varied. More research will be needed for the complete elucidation of the basis for the differences. Some RJDPs were found to possess DNase (Go et al., 1992) and RNase (Mock et al., 1996) activities. Those activities will be discussed in detail in section 1.2.2.
1.1.6 RIP-based Immunotoxins As RJDPs are effective in killing cells, they can be used to specifically target to undesired or harmful cells. As a result, a large number of immunotoxins (JTs) in which RJCPs conjugated with antibodies were constructed and examined for different kinetics and pharmacological studies (Ramakrishnan et al., 1992). Rn^-containing ITs were used in many in vitro studies, mainly for anti- tumour study (Ramakrishnan et al., 1992). Li an in vivo local treatment in mice, it was found that the ricin-containing ITs can effectively control the thymoma grafts and human tumour xenografts (Kanellos et al., 1989). In an in vivo human treatment, an IT of saporin-S6-containing anti-CD30 monoclonal antibody was used to treat Hodgkin's lymphoma (Falini et al., 1992).
16 The studies of TTs containing Rffs have been mostly in vitro or animal model studies (Barbieri et al., 1993). However, as more data will be obtained there will be more clinical trials for the different therapies using Rff-containing ITs.
1.2 Momordica charantia and its ribosome-inactivating proteins ORIP)
1.2.1 Momordica charantia Momordica charantia Linn is a climbing herbaceous annual. The fruits of M charantia are called bitter gourds. They are a favorite dish of Southem Chinese. The various parts of M charantia are used as traditional medicine and regarded as having the effect of eye-brightening, heat-dissipation and detoxification. They have been used to treat sunstroke, dysentery and eye pains. In fact, a wide variety of biological activities were identified from its fruits. These included guanylate cyclase inhibition (ClafIin et al., 1978), human lymphocytes cytotoxic action (Takemoto et al； 1982), cataractogenesis inhibition (Srivastava et aL, 1988)，antilipolytic & lipogenic action Q^Jg et al., 1987), insulinomimetic action (Ng et al., 1987)，hypoglycemic action (Ali et al, 1993)， tumour growth inhibition (Fletcher et al； 1980) and lectin-like action (Barbieri et al., 1987).
17 1.2.2 Ribosome-inactivating proteins ORIPs) in Momordica charantia Like many other plants, M. charantia has RJDPs. There are four RJPs existing in M. charantia, namely, a-momorcharin (a-MMC), p-momorcharin (P_ MMC) C^eung et al.，1988)，y-momorcharin (y-MMC) (Pu et al； 1996) and Momordica anti-HTV proteins 30 (MAP30) OLee-Huang et al” 1990). Their physical, chemical and biological activities were well characterized.
Classification and distribution There are two classes ofRIPs in M. charantia. They are respectively type I and small REPs. a-MMC, p-MMC and MAP30 are type I RJDPs. The thirty in MAP30 means 30 kDa. yMMC belongs to the novel class of small RPs. Up to now, the existence of type E and type YV RCPs in M. charantia has not been reported. a-MMC, p-MMC and MAP30 are type I RPs. y-MMC belongs to the novel class of small REPs. There is no report of the existence of type E or type W RBPs in M charantia. All these four REPs have been isolated from the seeds of M. charantia. But MAP30 can also be isolated from the fruits. At present, there is no report of isolation ofRBPs from other parts of the plant.
18 Physical and chemical properties The physical and chemical properties of FJDPs from M charantia were shown in Table 1.7. Like other type I RPs, the molecular weights of a-MMC, P_ MMC and MAP30 are around 30 kDa. y-MMC, being a small RH^, has a molecular weight of 11.5 kDa. a-, P_ and y-MMC are glycosylated while the extent of glycosylation ofMAP30 is unknown. a- and p-MMC are basic proteins. Their pI values are above 7. This is consistent with other type I RIPs which are also basic proteins. y-MMC is also basic. The pI value ofMAP30 is unknown.
Table 1.7. Comparison ofRBPs from Momordica charantia ^^ee-Huang et al” 1990; Pu et al., 1996; Yeung et aL，1987). M. charantia REP j oc-MMC*�丨 g-MMC*�丨[MMC*�丨 MAP30 .iiSi)i^S^;i^^3 j 29 [ 28 1•••.ii"5..………j 30 Number ofamino acids 263 248�丨 Unknown�丨 263 Neutral sugar content (%) | 1.6 | 1.3 | 0 j Unknown pI value j 8.5-9.0 j 8.5-9.0 j 9.5 j Unknown Homology with a-MMC (%) | 100 | 54 | Unknown | 48 *: a-, p- & y-MMC = a-，p- & Y_momorcharin, respectively; MAP30 = Momordica anti-HTV protein 30.
ln. considering the homology of these Rffs, they are fairly similar to each other. P-MMC and MAP30 have about 50 % homology with a-MMC. There are no studies on the homology of y-MMC with other available M. charantia Rffs.
19 Biological activities Like RJPs from other plants, RJDPs from M. charantia has the characteristic A^-glycosidase activity (Fong et al., 1996; Lee-Huang et al., 1995a; Pu et aL, 1996). They depurinate specifically the 28 S rRNA in the large subunit of the eukaryotic ribosome. They also inhibit translation in cell-free systems QLee- Huang et al., 1995a; Pu et al., 1996; Yeung et al, 1988). Apart from these activities, DNase/RNase activity, abortifacient activity, immunomodulatory activity, anti-tumour activity and anti-viral activity can also be observed in some of the M charantia RJPs.
(i) DNase/RNase activities a-MMC, p-MMC and MAP30 possess deoxyribonuclease ^)Nase) activity, a- and p-MMCs also have RNase activity. When a- and P-MMC were incubated with supercoiled, double-stranded SV-40 DNA from monkey, nicked circular and linear DNAs were produced (Go et al., 1992). On the other hand, when MAP30 was incubated with the DNA of human immunodeficiency virus (HTV), the predominant supercoiled DNA was converted to the relaxed form, even in the presence of gyrase which would have converted the relaxed form back to supercoiled form (Lee-Huang et al., 1995a). As a- and P-MMC were active towards another supercoiled DNA ofM13 mpl8 and they were inactive towards linear DNAs such as lambda, Ad-2 and T7, it was suggested they recognized the conformation of supercoiled DNA substrate
20 (Go et al., 1992). The DNase property may be the mechanism for the anti-viral activity of these RPs as will be discussed below. On the other hand, a- and P-MMC have RNase activity (Mock et al； 1996). They were also effective towards tRNA and polyU (Mock et al., 1996). The significance of this activity is not known.
(ii) abortifacient activity Another activity that can be found among the Rffs from M. charantia is abortifacient activity. This activity exists in a- and P>MMC. There are no reports on the abortifacient effect ofMAP30 or y-MMC. It was the abortion-inducing effect of M charantia fruit extract that led Yeung et al. (1986) to study the seeds of this plant and isolate a- and P-MMC. When the two MMCs were administered into mice in mid-term pregnancy, they induced 100 % abortion at the concentration of0.05 mg/25 g body weight (Yeung et al., 1986). This effect was also observed in early pregnancy but at a higher dose O^g etal., 1992) (Table 1.8). The mechanisms of the abortifacient activities ofthe MMCs in early and mid-term pregnancy are different. Li early abortion, the development ofembryo is affected. The MMCs inhibit the synthesis of macromolecules (RNA, DNA and protein) in embryo, leading to degeneration of embryos (Chan et al., 1984). They also decrease the number of implantation sites of embryos to the uterus (Chan et
21 a/” 1985). ii mid-term abortion, the trophoblasts in placenta are damaged and expelled from the matemal body ^,au, 1983). Table 1.8. Early and mid-term abortifadent effect ofa- and P-momorcharin in mice flVg et al； 1992; Yeung et aL, 1986). RIPs i Dose mg/25 g body j Days pregnant j No. ofmice 1 No. ofaborted mice wt. ofmlce when dnig was • j no. oftreated mice 1 i administrated ‘ j j (%) .b a-MMC I 0.05 | 12 | 7 j 7/7(100) I 0.20 I 3 I 8 I 5/8 (62.5) P-MMC 0.05 12 9 ！ 9/9(100) 0.20 [ 3 [ 10 9/10 (90) a： day 3 was considerei d as early pregnancy; day 12 is mid-term pregnancy . I b: a mouse was considered aborted when dead fetuses were found at over 50 % ofits implantation sites; the abortion rate was the percentage of no. ofaborted mice to no. oftotal mice.
(iii) immunomodulatory activity On further study of a- and p-MMC by Leung et al (1987)，it was found that these two MMCs possessed a wide variety of immunomodulatory effects. a- and p-MMC suppressed the mitogenic response ofsplenocytes towards mitogens including concanavalin A (Con A), phytohaemagglutinin OPHA) and lipopolysaccharide (LPS) ^.eung et al” 1987). The same research group also showed that a- and P-MMC suppressed the delayed-type hypersensitivity ODTH) of mice towards sheep red blood cells and that they inhibited the production of interleukin-2. Since Con A & PHA are T cell mitogens, DTH is a T cell response and interleukin-2 is a T cell-derived lymphokine, it was suggested that a- and P_ MMC act on T-cells in the immune system ^.eung et al., 1987).
22 On the other hand, a- and P-MMC affect B cells as well. This is supported by the inhibition of the mitogenic effect of LPS by the two MMCs (Leung et al., 1987). Meanwhile, a- and p-MMC also suppressed the humoral immune response of mice towards sheep red blood cells OLeung et al.’ 1987). a- and P-MMC also inhibit the migration as well as the phagocytic and cytostatic effects of macrophages (Leung et al., 1987). However, they did not affect the cytotoxic effect of natural killer QSfK) cells (Leung et al； 1987). The selective suppression of immune cells was not a result of lymphocytotoxicity as the viability of the treated cells was not affected ^Leung et al” 1986). It was suggested that macromolecule synthesis in the immune cells was inhibited as in the embryo 0^¾ et al； 1992). But further studies are required before the exact mechanism is elucidated.
(iv) anti-tumour activtiy a- and P-MMC also have anti-tumour effect in vitro. They inhibited protein synthesis in a variety of tumour cell lines at concentrations lower than micromolar level (Table 1.9). These cell lines included rat hepatoma (H35), mouse melanoma (B16), human squamous carcinoma 0-ICRyTO^4 & LICR>WJ11), human nasopharyngeal carcinoma QSfPC/HKl) and human choriocarcinoma (JAR). a-MMC can also inhibit a sarcoma cell line (S180) at a similar concentration QSfg et aL, 1994). Both a- and p-MMC were not inhibitory
23 to mouse liver cells at concentrations higher than micromolar level (Chan et al., 1992). On the other hand, MAP30 was also inhibitory to a variety of human tumour cell lines. The cell lines studied included the brain glioblastoma (87GM), breast carcinoma (BT20), epidemoid carcinoma (A431), liver hepatoma OHep 3B), melanoma 0VMme-3M), myeloma 0^266), neuroblastoma (SK-N-SH) and prostate carcinoma 0DU145) (Lee-Huang et aL, 1995b). The inhibitory activtiy of MAP30 on different cell lines was shown to be at nanomolar level. There is no report of anti-tumour activity for y-MMC.
Table 1.9. Inhibitory activities ofa-MMC, p-MMC and MAP30 on different tumour cell Unes O^ee-Huang et aL, 1995b; Ng et aL, 1992). Inhibitory effects on cellular protein synthesis were measured. Tumourcelllines~ |IC5oofDrugs(^M)* [ct-MMC |3-MMC MAP30 ..赵]]^9扭?1运函 ====::=]互:「r.66………:………亜•••• H]m^sgi^ .."5793 0.86 ND 】胜躲!今乡9£|^鹏经资丄亟@:亟:[[::: o"il ..….••...ND*……… �?^??]1视..?.^^-9.4?.£汪—扭色说@1^^^^::: 0"28 0.26 .pii' �yEEL?k9^i9?J?—^?.!^^^ i 0"02 0.03 S5 ..M9]^.?.B®iE^9E.^.(?.IS 0"02 0.03 Mi —吞赃视^^^^鲍£呢^^视3:面!^) 亜 i^5 0.00034… --lyEE[]iYE.ji^patOEa .....ND"...."..…0'm0S6.… ••MyE??L^?!?^?-9.E!(M?ji^?:ii^ 亜 而 0 00021… Human prostate carcinoma DU145) 而 而 0.00188… * ICso = 50 % inhibitiory concentration; ND = Not determined.
24 It is not known why a-MMC, P-MMC and MAP30 are selectively inhibitory to tumour cell lines. It was suggested that the mechanism may be common in embryonic tissue and tumor cells as both of them have common properties, e.g. rapid cell divisions and protein synthesis (Chan et al., 1992).
(v) anti-viral activity The most striking activity of the Rffs from M. charantia is their anti- human immunodeficiency virus (anti-KDV) activities. a-MMC, P-MMC and MAP30 have anti-fflV activity ¢^ et al” 1992; Lee-Huang et al, 1990). Take MAP30 as an example. MAP30 inhibited HTV in vitro. It showed an IC50 of inhibition on HTV infection-induced cell fusion (syncytium formation) in HIV-infected H9 cells at concentration lower than nanomolar level (Table 1.10). It also inhibited expression of the HTV antigen p24 and the activity offflV reverse transcriptase in the infected cells with similar values ofIC50 (Table 1.10). While HTV in H9 cells was inhibited, MAP30 was non-toxic to the H9 cells. It showed an IC50 ofinhibition on protein synthesis in H9 cells at concentrations greater than 3 nM.
Table 1.10. Inhibitory effect ofMAP30 on fflV O^ee-Huang et al” 1995b). Assays 丨 Syncytium | p24 丨 RT | Cellular protein j formation | j | synthesis IC50 (nM) of the | 0.22 0.22 1 0.33 j >3000 MAP30 in different 丨 assays | | i j IC50 = 50 % inhibitory concentration.
25 It was suggested that the mechanism of the anti-HTV activity of RPs like MAP30 was not similar to that of the abortifacient, immunomodulatory or anti- tumour activities fBarbieri et al” 1993)，i.e. the anti-HTV RJDPs do not exert their effect through inhibition of macromolecules. One possible mechanism was the topoisomerase activity of MAP30. As mentioned above, a-MMC, P-MMC and MAP30 were shown to be able to cleave circular viral DNA into a linear form. MAP30 was also confirmed to have such effect on fflV DNA 0-ee-Huang et aL, 1995a). Meanwhile, MAP30 was able to interrupt UTV integrase, preventing the integration of HTV DNA into the DNA from normal cells (Lee-Huang et al., 1995b). The topological inactivation and integration disruption were suggested to be the mechanisms for the anti-HTV activity ofMAP30 (Lee-Huang et al., 1995a). Nevertheless, the exact mechanism is still unknown and it requires further studies. Apart from inhibiting fflV, MAP30 is also inhibitory to herpes simplex virus OHSV). It inhibits both the HSV-1 and HSV-2 with values of IC50 lower than 1 nM in vitro in a lung fibroblast cell line (WI-38) ^Bourinbaiar and Lee- Huang, 1996). The inhibition ofHSV may help cure AEDS patients with resistant strain ofHSV (Bourinbaiar and Lee-Huang, 1996). However, as the case of anti- HTV activity, the mechanism for the anti-HSV activity remains to be defined.
26 (vi) immunotoxins Momordin is another name for a-MMC (Barbieri et aL, 1993). a-MMC was used to construct many immunotoxins (ITs) under the name of momordin OBattelli et al, 1996; Terenzi et al” 1996; Wawrzynczak et al.，1990; Dinota et al., 1989; Stirpe et al., 1988). Momordin has been used to construct JJ for use as anti- tumour agents. Many different types of momordin-containing ITs were made. Some of the tumours studied were bladder carcinoma (Battelli et al., 1996)， Hodgkin's disease (Terenzi et al., 1996)，lymphoma (Stirpe et al., 1988) and myeloma (Dinota et al” 1989). All these studies were in vitro. The production of ITs enabled the targeting of momordin to tumour cells with minimal non-specific cytotoxic effects OBarbieri et aL, 1993). Conjugating momordin to ITs also enhanced the cellular entry. The values of IC50 for the effect of momordin-based ITs on protein synthesis in tumour cell lines are at nanomolar concentrations or below ^)inota et al., 1989; Stirpe et al., 1988; Terenzi et al., 1996; Wawrzynczak et al.’ 1990) while the values of IC50 for the effect of momordin alone are at micromolar concentrations (Table 1.9). However, in spite of the effectiveness of the momordin-based ITs in vitro, many problems still exist in vivo. For example, there is the problem of hepatic clearance (Wawrzynczak et al., 1990). Glycosylation of the RJP may lead to the binding of the ITs to hepatic cells with consequent removal from blood stream. Meanwhile, repeated administration of IT into patients will result in immune response against the toxin (Battelli et al., 1996).
27 Switching to different ITs containing different Rffs can prevent the induction of immune response (Battelli et al., 1996). P-MMC and MAP30 are possible constituents for a-MMC. Meanwhile, new REPs can also be alternatives for making new ITs. Li fact, the present study was also prompted by the objective to search for new RDPs to prepare new ITs for different therapeutic strategies.
1.3 Objective ofthis study The objective of this study was to isolate new RJPs from M. charanta. Li the therapy using ITs, repeated and prolonged injections would result in immune response against the IT OBattelli et al., 1996). It is because the protein nature of REPs and antibodies in the ITs induced the immune system. To circumvent this problem, switching to new TTs with different REPs after prolonged treatment is required ^attelli et al； 1996). Although four Rffs (a-, P-, y-MMC and MAP30) have akeady been identified and isolated from M charantia, more RIPs may exist in the plant. Li fact, during the cloning ofMAP30 gene, it was found that there were at least eight homologous genes related to MAP30 (Lee-Huang et aL, 1995a). Three of them were identified to be MAP30, a-MMC and p-MMC (Lee-Huang et cd., 1995a). The identity of the remaining isoforms were unknown. Thus, it is likely that more
28 RDPs can be isolated. These potential new Rffs may provide a new source for the production ofITs. Seeds and fruits were the target tissues for isolation of new RDPs in this study. They will be discussed separately in the following two chapters.
29 Chapter 2 Study on a new ribosome-inactivatingprotei n Of^P) from Momordica charantia seeds 2.1 Introduction Seeds usually have high concentrations of RCPs OBarbieri et al., 1993). Meanwhile, there may be more than one form ofRP in the seeds or other tissues of a plant. For example, in Saponaria officinalis L., there are four isoforais of RPs in its seeds, three in the roots and another three in the leaves fFerreras et al., 1993). The existence ofmore than one form ofRBP in a tissue was also observed in the Cucurbitaceae plants. For example, in Trichosanthes kirilowii, there are trichosanthin and trichosanthes anti-fflV protein 29 (TAP 29) in the roots (Kubota et al” 1986; Lee-Huang et al., 1991; Toyokawa et aL, 1991). ti Luffa cylindrica, luffin-a and luffin-b exist in the seeds C^exmg et aL, 1991). ]n M. charantia, there are four REPs, namely, a-MMC, p-MMC, MAP30, and y-MMC, ab:eady isolated from its seeds. But it is possible that some other new Rffs still exist. Li the cloning of MAP30 gene, homologous but different nucleotide sequences were identified (Lee-Huang et al., 1995a). It was suggested that there were at least eight related but distinct isoforms ofRBPs in M charantia fLee-Huang et aL, 1995a). Three ofthem were identified as a-MMC, p-MMC and MAP30 (Lee-Huang et al, 1995a). But the identity of the remaining sequences were unknown. This supports that new REPs may exist in the seeds of M. charantia.
30 In this chapter the isolation of a new RJDP fromth e seeds of M charantia is reported. To isolate the new RJDP from M. charantia seeds, the method for the isolation of a- and P-MMC was adopted (Go et aL, 1992). Li this method, affinity chromatography on Affi-Gel® blue gel was used fFig. 2.1). This affinity gel has triazine dye functional groups and it can adsorb nucleotide-dependent enzymes including RH>s ^Vlunoz et al., 1990). The salt-eluted proteins from the affinity column were loaded on a fast protein liquid chromatography OFPLC) column of Mono S after dialysis. Mono S is a cation-exchanger with fine matrix particles which allow high pressure and high speed chromatography with high efficiency. With the high resolution of the Mono S column, the salt-eluted proteins from the affinity column were resolved into individual peaks. To characterize the biological activities of the isolated proteins, assays for A^-glycosidase activity and cell-free translation inhibition ^Rff) activities were carried out.
31 Decorticated seeds
(i) Grind seeds in 1 OmM NaH2PO4, pH 7.0; (ii) Centrifuge to remove debris; (iii) Dialyze against the buffer. >J/ Crude extract
sp Affinity chromatography on Affi-Gel blue gel
^/^ Breakthrough Salt-eluted proteins
(i) Concentrate by ultrafiltration; (ii) Dialyze against 2mM phosphate V buffer,pH 7.5. FPLC ion-exchange chromatography on Mono S
N^ NewREP Fig. 2.1 Procedure for the isolation of a new ribosome-inactivating protein (RIP) from Momordica charantia seeds.
32 2.2 Materials and Methods 2.2.1 Materials Ripe dried seeds of Momordica charantia (bitter gourd) were purchased from the local market. The seeds were stored in a cool dry place before use. Before extraction, the seeds were decorticated. Affi-Gel® blue gel (100-200 mesh size), Mini-PROTEAN® E electrophoresis cell for the SDS-PAGE and agarose (high strength analytical grade) were products ofBio-Rad OJ.S.A.). The Mono S HR 5/5 FPLC column was purchased from Pharmacia LKB (Sweden). SPECTRATPOR® dialysis membranes were purchased from Spectrum O^.S.A.). Diethyl pyrocarbonate ODEPC), bromophenol blue, ethidium bromide, hemin, creatine kinase and creatine phosphate were purchased from Sigma (XJ.S.A.). Aniline and formamide were products of Merck & Co. O^.S.A.). Phenol and chloroform were purchased from USB (JJ.K.). The RNA marker was purchased from Promega O^.S.A.). Photographic fihn model 667 was purchased from Polaroid (U.S.A.). Tritiated leucine was purchased from Amersham ^U. K.). Rabbit reticulocyte lysate was obtained from rabbits raised at the University Animal House, the Chinese University of Hong Kong (C.U.H.K.). All other reagents were of analytical or molecular biology grade, tf possible, they were autoclaved. Precautions for RNA work were taken throughout the process.
33 2.2.2 RIP isolation Preparation of crude extract Li a typical extraction, 5 g of decorticated seeds were ground by pestle and mortar until they became powder form. The powder was than homogenized in 40 ml of 10 mM NaH2PO4, pH 7.0 by Polytron (Kinematics). The slurry was stirred at 4°C for 3 hours. It was then centrifuged at 30,000 g for 1 hour at 4°C in a high- speed centrifuge (Beckman model J2-MC). The pellet was discarded and the supernatant was filtered through cheesecloth. It was then dialyzed at 4°C overnight against 10 mM NaH2PO4, pH 7.0 in a dialysis tubing (SPECTRAyTRO ®) of 3,500 molecular weight cut-off (MWCO) in a cold room.
Affinity chromatography on Aff!-Gel blue gel Crude extract from the seeds was mixed well with Affi-Gel blue gel (�8 ml) which was equilibrated with 10 mM NaH2PO4, pH 7.0. They were mixed well by shaking. It was then centrifuged at 3,800 g at 4°C for 5 minutes at a table- top centrifuge QBeckman). The supernatant (�35 ml) was collected as unadsorbed (breakthrough) proteins and stored at 4°C. The gel was then further washed in the buffer fivemor e times with a total volume of 240 ml. The gel was then packed in an Econo colunm (BioRad) to form an Affi-Gel blue gel column (1 x 8.6 cm). The gel was washed with 72 ml buffer to further remove any unbound proteins. It was then eluted with 0.5 M NaCl in the same buffer.
34 The salt-eluted fractions (�23 ml) were pooled and dialyzed against 2 mM NaH2PO4, pH 7.5 using a 3,500 MWCO dialysis tubing. It was concentrated through a 3,000 MWCO filter membrane in a concentrator (Amicon, U.S.A.).
Chromatography on Mono S The concentrated sample (�12 ml) was filtered through a 0.22 ^M filter. The adsorbed fraction from the Affi-Gel blue gel column was filtered and then applied to a Mono S HR 5/5 fast protein liquid chromatography OFPLC) column fPharmacia, Sweden) previously equilibrated with 2 mM NaH2PO4, pH 7.5. The Mono S column was connected to a Waters 600S/486 (U.S.A.) HPLC and UV monitor system, ti a typical run, 1 ml of the filtered sample was injected into a 1 ml sample loop and loaded on the column at a flow rate of 1 mVmin. The breakthrough protein peak was collected. Afterward, another 1 ml sample was injected. The method included washing for 10 min. with buffer and a linear gradient of 0-60 mM NaCl in the buffer for 60 min. The eluted protein peaks were dialyzed against distilled water in a 10,000 MWCO dialysis tubing.
2.2.3 Characterization Sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) Electrophoresis of protein samples was performed according to the procedure previously reported (Laemmli and Favre, 1973). The resolving gel was composed of 15 % (w/v) acrylamide, 0.375 M Tris, pH 8.8，0.1 % (w/v) SDS, 1 %
35 ammonium persulfate and 6.6 X 10_5 % TEMED. The stacking gel was composed of 5 % (w/v) acrylamide, 6.25 mM Tris, pH 6.8, 0.1 % (w/v) SDS, 1 % ammonium sulphate and 7 x 10'^ % TEMED. Analysate were diluted with a buffer which was composed of 1/4 v of stacking gel buffer, 20 % glycerol (w/v), 10 % SDS (w/v), 0.005 % bromophenol blue, and lmM EDTA, pH 7.5. The diluted sample was boiled for 5 min. The electrophoresis buffer was composed of 0.025 M Tris, pH 8.3, 0.192 M glycine and 0.1 % SDS (w/v). Electrophoresis was carried out at a constant current of 16 mA and room temperature for about 1.5 hr. The gel was stained with 0.115% Coomassie Brilliant Blue R-250 in a staining solution composed of 8% (v/v) acetic acid and 25 % (v/v) ethanol for 1 hr, and then destained in a solution of acetic acid: ethanol: water (1:3:4, v:v:v) until the background was clear.
Amino acid sequence analysis Proteins for amino acid sequencing were first separated by SDS-PAGE using 15 % polyacrylamide gel as described above. The gel was rinsed with double-distilled water and soaked in transfer buffer (25mM Tris and 192 mM glycine，pH 8.3 with 15% methanol) for about 5 minutes. Four filter papers of about the same size as the gel were also soaked in the transfer buffer. Meanwhile, a polyvinylidene difluoride (PVDF) membrane, which was also about the size of the gel, was soaked in 100% methanol for a few seconds and then soaked in the transfer buffer for about 10 to 15 minutes.
36 The polyacrylamide gel was laid above the PVDF membrane and they were sandwiched between the buffer-soaked filter papers (2 above and 2 below them) in a blotting cassette (BioRad, U.S.A.). The blotting was carried out at a constant voltage of 22V for lhr. Afterward, the PVDF membrane was stained with 0.1% Coomassie Blue for 1 hr and destained ovemight. The desired protein band was cut out and rinsed thoroughly with double- distilled water. It was then applied to an amino acid sequence analyser (Hewlett Packard, U.S.A., by courtesy ofDr. Wing-Shing Ho, the Chinese University of Hong Kong).
Determination of protein concentration Protein concentration was determined by the method previously reported (Lowry et al., 1951) using bovine serum albumin (BSA) as a standard. To each 0.1 ml sample, 1 ml solution A (1% CuSO4.5H2O : 2 % Na3C6H5O7.2H2O : 2 % Na2CO3 in O.lN NaOH, 1:1:100，v:v:v) was added. The sample was allowed to stand at room temperature for 10 min. Then 0.2 ml solution B (1 to 1 dilution of Folin's reagent in water) was added immediately followed by mixing. After standing at room temperature for another 45 min, absorbance at 750 nm was measured using a Hitachi U2000 spectrophotometer.
37 iV-glycosidase assay Sample 10 |il was added to a mixture composed of 70 ^1 rabbit reticulocyte lysate and 60 ^il buffer A (25 mM Tris-HCl, pH 7.6，25 mM KC1, 5 mM MgCl2). The negative control was 10 |il DEPC-H2O instead of sample. The positive control was 10 ^1 trichosanthin (TCS, a type I Rff, final concentration, 0.33 |jM). The reaction mixture was incubated in a dry bath OEppendorf) at 37°C for 30 min. The reaction was stopped by adding 260 ^1 0.77 % SDS and chilling in an ice-bath. RNA was extracted with phenol:chloroform equilibrated with 0.1 M Tris- C1, pH 8.0. An equal volume of phenoVchlorofonn (equilibrated with 0.1 M Tris- C1, pH 8.0) was added to the SDS-treated reaction mixture in an Eppendorf vial followed by gentle mixing until an emulsion was formed. The aqueous layer was separated by centrifugation in an Eppendorf centrifuge at 14,000 rpm, 4°C for 2 min. The aqueous layer was re-extracted using the same procedure. RNA obtained from the phenoVchloroform extraction step was mixed with 2 volumes of absolute ethanol and 0.1 volume of 3M sodium acetate, pH 5.2. The mixture was well vortexed and kept at -20°C for 30 min. It was centrifuged at 14,000 rpm, 4°C for 25 min. in an Eppendorf centrifuge. The ethanol layer was decanted. The RNA pellet was washed with 70 % ethanol and centrifuged for 5 min. The RNA pellet was resuspended in formamide. After washing, the
38 precipitated RNA was resuspended in 16 |il foraiamide. Half of the resuspended • V RNA was incubated with 10 volumes of 10 % (v/v) distilled aniline (working aniline) in a dry bath at 37°C for 3 min. The other half was labelled as untreated (-ve) RNA and stored at -70°C. Working aniline was prepared from distilled aniline, acetic acid and DEPC-H2O (60: 90: 444，v:v:v). Ten volumes of working aniline were mixed with the extracted RNA. The mixture was mixed well and incubated at 60°C for 3 min. Afterward, the mixture was chilled on ice and then spun down for a few seconds. The aniline-treated RNA was then precipitated as described above. The RNA was resuspended in 8 jid formamide. The aniline-treated (+ve) and untreated (-ve) RNA were analysed by formamide gel electrophoresis. Formamide gel (1.2 %) was prepared. Agarose (0.48 g) was melted in 16 ml DEPC-H2O in a microwave oven. Then 4 ml of electrophoresis OEP) buffer (36 mM Tris, pH 7.6, 30 mM NaH2PO4, 2mM EDTA- Na) and 20 ml of 100 % formamide were added. After mixing and cooling, the mixture was poured into a plastic gel cast. An aliquot of 3 ^1 precipitated RNA was mixed with 1 ^1 EP buffer, 6 ^il 100 % formamide and 1 ^1 gel loading buffer (0.5M EDTA and 0.5 % (w/v) bromophenol blue in glycerol). The mixture was incubated in a dry bath at 65°C for 5 min. Afterward, the sample mixture was chilled on ice and spun down for a few seconds.
39 Before sample loading, the agarose gel was placed in a gel tank filled with electrode buffer (10 % EP buffer, 50 % foraiamide, 40% DEPC.H2O). Samples were then loaded. Electrophoresis was run at 80V for 1.5 hour. The gel was stained with ethidium bromide (0.5 [xg/ml) for 10 min. and then destained for 15 min. in distilled water. The gel was photographed with polaroid 667 instant fihn. To prevent RNase contamination, gloves were wom throughout the whole process. Whenever possible, sterile disposable plasticware was used. Reusable plasticware was soaked in a 0.1 % (v/v) solution ofDEPC followed by rinsing in DEPC-treated distilled water prior to use. Glassware was heated at 180°C overnight. Water and buffer were stirred in 0.1 % (v/v) DEPC overnight at room temperature and then autoclaved before use. Tris buffer was non-autoclavable. It was prepared from the DEPC-treated water and autoclaved. Wherever possible, all other reagents were ofRNase-free quality.
Cell-free translation inhibition assay ^REP assay) Rff assay was based on a previous method OPeUiam and Jackson, 1976) with modification. The sample was diluted with 0.1 mg/ml BSA. Ten microlitres ofthe diluted sample were mixed with 10 ^1 hot mixture (500 mM potassium chloride, 5 mM magnesium chloride, 130 mM phosphocreatine and 1 ^Ci L-[4,5- 3R]leucine) and 30ul of working lysate (0.1 ^iM hemin; 5 |ig ofcreatine kinase in rabbit reticulocyte lysate). The reaction mixture was incubated at 37°C for 30
40 min. The reaction was stopped by addition of 330 |il lM NaOH containing 0.2% H2O2. The mixture was further incubated at 37°C for 10 min. The reaction mixture was divided into triplicates of 100 |il and pipetted into a 96-well plate (Nunc, U.S.A.). The reaction mixture was mixed with an equal volume of acidified caesin (40% (w/v) trichloroacetic acid and 2% (w/v) casein hydrolysate). It was kept in an ice-bath for 10 min. The precipitate was then collected by filtration on a circle of glass fibre filter paper (Whatman GF/A, U.K.) through a cell harvestor which was washed with absolute ethanol. The filter paper was suspended in scintillant (67% toluene, 33% Triton X- 100, 4 g/1 PPO and 0.4 g/1 POPOP) and counted in a liquid scintillation counter ¢.56500 Beckman, U.S.A.).
Preparation of rabbit reticulocyte lysate Rabbit reticulocyte lysate was used to provide rRNA for iV-glycosidase assay and cell-free translation apparatus for RP assay. Rabbits rendered anaemic by phenyUiydrazine was prepared as described previously (Maniatis, 1982). PhenyUiydrazine (1.2 %; neutralized to pH 7.5 with 1 M HEPES, pH 7.0) was injected subcutaneously into a 2-3 kg rabbit. Different volumes, 1.0, 1.6，1.2，1.6, and 2.0 ml, were injected on days 1，2，3, 4 and 5 respectively. The rabbit was bled on days 7 and 8. Blood was collected directly from the heart chamber with a heparin-rinsed syringe, mixed with normal saline containing 0.02% heparin, and
41 then centrifuged at 2,000g for 5 min. in a Beckman model J2-21 centrifuge. The packed cells were washed with normal saline three more times. The last centrifugation was carried out at 17,000 g for 30 min. The lysate thus formed was aliquoted and stored at -70°C with a stability for at least 6 months.
2.3 Results Crude extract from the seeds of Momordica charantia was mixed with Affi-Gel blue gel. Afterward, unadsorbed proteins were washed off from the column and the gel was then packed into a column. The profile of elution from the column is presented in Fig. 2.2. Adsorbed proteins were eluted by 0.5 M NaCl as a single peak designated Affi_S (Affi_Seed extract) after it had been demonstrated that no ftoher unadsorbed proeins were eluted from the column by washing with 10 mM NaH2PO4 buffer, pH 7.0.
42 3.50「， , 、 、 ， ^ . ‘ ‘ V ‘ ‘‘ ‘‘ . ； V. ‘ ,/ ' g:iS|;;iii;^ii:i;i;;iiiiiliSiiii;i$^|i V iiiM^iWMSf^^
5?;¾;;¾¾:;;¾^:^¾¾:;¾¾':¾¾-:-:¾¾¾¾¾¾:¾iiililliiiiW*il^^^ ^ 3 00 ” ’ � ， ，''" / ‘ ‘‘ , , ‘‘ ‘‘ ‘ ‘ ‘ ,: ',� , ‘ :� —:, V:'/,-' ‘iAl6Bi^S ”” ‘ ^^^^^^^^^Si^m^9^mrn^^^^^^^^^m. w^^^^^^^^^m^m^9^m^^^^ ^^^i^^m^^m^m^^^^^^^^^^^^^^^^^^^^^^^-錄書缘丨麗:纏:丨顯丨::難顯丨:丨_靈丨鬆_丨_丨纏丨丨翁辟鎮:,霧;丨黨:__纖:蕩,鎮養 2.50 ^ ‘� “ , f :,' ‘ ‘‘ ‘ , \ ''''� ：,' '' “ “':，',、:7:‘ ____誦騰____聽_謹___誦 ___讓_____議___麗____1_ iii:i;i!^;;g;p;e:^;i;i;i^is:;i;:^li!^^ ^ 欲務丨_丨丨!；_欲丨鍵謹欲麵|絲珍;;_竊_獲漏::_:參総丨靠珍资麟丨_丨獲浴錄_;:纏丨丨縫穆法麗丨丨丨凝;務丨丨丨務:丨缀縫终纖忽丨_黎丨_丨丨総_;建務 ：；：：：：：：< ：；；- ^!:::Sy gs:®i^^w^5SS：；：；：；> ®；：：：：?! ：：：：- ^^；；>> ^^ 0 2.00 - Mfiivt^I 00 • I • S V ‘
0 1.50 ^¾^;¾¾;;¾;!^¾;¾!¾;::!:¾- , ¾ . 丨銷控丨丨鐵: I毅鎮凝_ 丨餘頃丨丨懲丨丨丨益丨§:;: ‘ 擬丨按_ 取殺彩；:: , ,丨^ 睡怒総務麵法移努錄敏丨凝紹; , \ 兹/ 殺丨_ 麵鈔如發丨滋; 丨丨: 蕩; 恣; 丨; 游兹数齒丨; 丨滋丨丨丨怒思凝; 凝丨圾《: 丨: 丨:::;: 绅:; 形;: 《麵鋒丨丨摄:;) 丨, 注 i:Sw;iiS;;iS;i;^i:i:gi:;S:s;;iS 凝_蔡_錄_丨;_丨丨德海黎羅丨議缀_1 :|纖丨顯缀丨丨觀鐵:3 _錄_!_鏃毅 iiiiiii;;|Sllii;ii:^|ii^:-:¾;¾;;-::;:::;>¾:;:::;¾!:;.^:!:;;¾';':-:¾':'^^ ^ ^¾::¾¾-:::::::¾¾;%¾ 1.00 - :. . \；：:‘ .”:.'' ：^；：'； 0.50 ：：：：：；：-; ;¥ , ：：；; ; ：；：：：^ ! ：；$ ：：：! ^ ；^! > !^ -- ', ‘‘0.5MNaC ‘ ‘ , l ^;?;%::<::::¾;!!^:¾;:¾¾::¾', ,/V \¾ - ：‘‘ “‘ ,':',.'" , - ‘ ‘ ‘, ‘ ",,X ‘‘‘.‘‘.‘ ‘, 懸:麗_1_纖,雅_麵縫_麵^ _纖禱灘縫凝 ,_____丨毅麵______鐘_纖_聽灣:霧____纖麵_ ^m^^^^^^^m^^^^^^^^^^mm‘ � , , ‘ « iiiw‘ i/ ‘ ‘ ,'^s ^^Ss ^ ‘‘,‘ 0.00 丨丨丨- ~" …“…丨丨， 3L-l——‘‘丨^-^,"""•-1,,||,|||| ‘ < • 0 10 20 30 40 50Ehdo 60 n Vokon70 80 e 9(mT0 )10 0 110 120 130 140 Fig 2.2 Profile of elution ofMomordica. charantia seed extract from Affi-Gel blue gel column (1x8.6 cm). Refer to section 2.2.2 for details. Briefly, crude extract from 5 gM charantia seeds was mixed with Affi-Gel blue gel equilibrated in lOmM NaH2PO4, pH 7.0. Unadsorbed (breakthrough) proteins were obtained by several times ofwashing. Afterward, the gel was loaded to the column and further washed with the buffer. The column was finally eluted with 0.5 M NaCl in buffer (indicated by the arrow). Fractions of 1 ml each were collected. The salt-eluted fractions (elution volume from 81 to 110 ml) were pooled to form Affi-S. After concentration and dialysis, AfFi-S was loaded to a Mono S column. The Affi-S was resolved by the Mono S column into ten peaks (Fig. 2.3). The first peak (peak 1) was the unbound (breakthrough) peak which was eluted before salt-elution began. The other nine peaks were eluted after salt elution begins. They were designated as peaks E to X with retention times of 13, 15, 24, 27, 41, 43，53 and 63 min., respectively.
43 OD280 concentration 丽 ofNaCl (mM) i /- 60 0.35- : i / ‘ ‘� ‘ ‘ ‘‘ , v^ ‘ � � � ‘‘ ‘‘ ‘ ‘！ ‘ ‘ , . , ‘ jT ‘‘ ‘ X ^ • 50 0.30- vn v^ ; 0.25 • A ‘ y^ ‘ ‘ :- 40 ‘： 0.20 • “ '''. V , i^.‘‘‘ ‘ oi5_ ‘ ：厂，/^ ?' " 'A "““ ：二、.' :Zf.\:/:f\ 1 0.05-\ ‘ u^wv VI I \ m I \ - 10 V Ak y^,w�J 、一入」\ 0.00 �..,..I'‘I f"�r - 1 丨 .(~I r^ 00 0 10 20 Elutio30 n tim4e0 (min.) 50 60 70 Fig. 2.3 Profile of elution ofAffi-S from a Mono S column. Refer to section 2.2.2 for details. Briefly, the column was washed with 2mM NaH2PO4, pH 7.5 and eluted with a linear gradient of 0 - 60 mM NaCl in the same buffer at a flow rate of 1 ml/min. Peaks I to X were collected.
Among the peaks, peaks V, Vm and X were shown to have 7V-glycosidase activity (data not shown). By molecular weight determination and comparison of the results of a- and p-MMC isolation (Go et al,, 1992), peak Vm and peak X were identified as a- and p-MMC, respectively (data not shown).
44 1 2 3 4 M Molecular weight j^^^^i;^^^�jt^^^^^j^i^^^^^^^^^^^^^^^^^^^^^a^;:^^y^gi^^^^^g^j;^j^j^^^;>^;;^>^;^^j^j^|>.^^^>^^^^^gg^;>^y~^»^^^^;^y^^^^^^gjg *""* 醒觀^^^^^^^^^^^^^^^^^^^^® rkDa^ ^^^^^^m^m^^^^^^^^^^m^^^^^^m^^m^m-^¾¾¾^^^^^^^^^^^^^^^^^^^^*^^¾^¾ ： \x^aj ^^^^^^^^^^*^^^^^^^»«»*#^»^•^8«^^^^ i ^^^^^^^^^^^^^^-::< •、： '、,/ ；•；- : . . . ；H , "^^ ~ — 94.0 67.0 ^^^^^^m0M^^^^^^^^^^^^^^^^^^^mM^^Kw^^^^^^^^^^^m^m ^^^^^^^^^^^B^^^^^^^B i~ ^^^^iii^^^^^^^^H 43.0 ^^^^^^H^^^^^^^^^^^^^^^^WI^^^^^;鐘__種蒙!_^^^^_^^寒攀|(^^^^^^^__|^^^^8^^^®^^^^H® ^^^^^^^^?^^^響禱働^^^^^^^^__^^^^^^^耀藥爆黎^^^^^^^^^^攀^^穩尊|__1^^^ ^<»£&»»«"胁1 ^^^^^^^^te� ^ >^di^&tfitoViMMP^^ W9"*ww - ，fti__fni^哪w,Mr ^ ^^HM^Mi * ^*"“^ 30.0 ,7^'^x?>V0'\,' '\' '?� ':^:i^t''j������ 2'r'i^^^0^''怎>(n\��斤'vS^^h'^(�� *^ ' ,''/' ^'V' ' ''^'^' XS >j"''^N^^v ^t^^ ^^^^_S^iP^^^^P^^^^^S^^^^M^^^Si^P^W^_^^^Ki^^^B |^^g-,�"?V '^''V�� %^^'^^:^^i^|^\�^ ~》"^<、公《^|V^^' V\/ C"'^^>^^ A^*^'^ 4's'^' '%- ； sV"/"'* ^/^^^'>^i^''^'i%'&iMi S M^^^^M^^^^^^^^^H ._ 81¾^¾^¾¾^^^^^¾:!¾^¾¾!^^^^^¾¾¾¾¾^^^¾¾!^¾¾^¾^^^^^^^Pi*^^^^^^^^^^^^^!!^^^^^•糖翻|凝_逾徵1麵麵總纖徵翻鏺您鄉燃,^1_^^^_纖纖鬆^^5:|^錄丨麵_察_驗磁H^^^¾ ^ 20.1 ^^P^^^M^^^^^^g^^^^i^^^^^__IPI^^i^^^i_i"^;M_^^:^^l>^^^^^^^^l^^i^^^^^^^_^^f^l^Plfii"^3^l^l_^_^B ^ ^
^|1?^纖::怨羅暖》赛§$5丨緯,議 14.4 %^^^fc^^^®j^^t^^^^ ^½''^/'" ^^^¾ ^' ^1^¾ Fig. 2.4 SDS-PAGE of breakthrough peak and peak V (6-MMC) from Mono S column. Lane 1，trichosanthin (TCS, 2 ^ig); lane 2，a-momorcharin (a- MMC，2 ^g); lane 3，peakV(5-MMC) from Mono S column (1.5 ^ig); lane 4， peak I (breakthrough) from Mono S column (0.5 ^ig); lane M, protein marker (Pharmacia). a-MMC (29 kDa) and TCS (26 kDa) are standard proteins. Peak V was also analysed by electrophoresis. In Fig. 2.4, peak V and peak I (breakthrough peak) from the Mono S column were run in SDS-PAGE with the standard type I Rffs of a-MMC and TCS (Dr. T. B. Ng, Biochemistry Department, Chinese University ofHong Kong). Peak V was shown to be a single protein. The molecular weight of this single protein was found to be higher than those ofboth a-MMC (29 kDa) and TCS (26 kDa). It was estimated to be about 30 kDa. Peak I consisted of a number of different proteins but no 30 kDa protein as the peak V. 45 1 2 M 3 4 5 6 7 8 9 10 11 12 RNA marker W|^^||^^^^^||^^^|^^g^^^^^^^m^a^^^^^^^mK|^^m||^^^^|m ^^^^^^^^^^^^^^^K^^^^^^^^^^^^^^^^M 6583 •• ^^^^^HB|^^^^^^^^^H 4981 ^ ^^^^W^^H^HH^^^^^^^^^H iH_:MWMiwr™
1383 . 0 •••^••^rRNA
HB. flH^ J| flK>JBJ^^yi —Endo's band ^^'^Hm^^H
Fig. 2.5 A^-glycosidase activity of peak V (6-MMC) from Mono S. Assay method was described in section 2.2.3. Sixty nanogram ofrRNA from rabbit reticulocyte lysate was loaded to each sample lane. Lanes 1 & 2: negative control of DEPC-H2O; lanes 3 & 4: positive control of trichosanthin (TCS, 0.33 ^iM); lanes 5 & 6，7 & 8，9 & 10，11 & 12: peak V from Mono S column (2.38 nM, 2.38 x 10'^ nM, 2.38 x 10"^ nM and 2.38 x 10"^ nM, respectively). The "+"，，sign indicates aniline treatment while "-" sign indicates no aniline treatment.
Fig. 2.5 shows the iV-glycosidase activity of peak V. When rRNA from rabbit reticulocyte was treated with peak V at concentrations from 2.38 nM to 2.38 X 10'3 nM, followed by aniline treatment, the characteristic Endo's band ofabout 400 bases was observed. The positive control of the type I REP, TCS, also
46 produced the Endo's band when aniline-treatment was done. For both peak V and
TCS, no Endo's band was seen when there was no aniline treatment ("-"’，lanes). There was also no Endo's band in the negative controls.
100 n 90 ^ T 80— jX^^^ 录 ： A § 7oi / 1 60 — / •^^ - Z 名 50- 7f 0) 二 丁 / ; 1 ( yK i S ： ^x 2 30 - ^y^ 0. :O-"-"-""""^^ 20 ^ 10 "i
- .•.. • • 0 I 111 j 1—I—I~I"•I 1111—‘—I—I—I“I I 1111 1—I—I“I""""^I 111 10-2 10-1 10� 101 Concentration of peak V (5-MMC) from Mono S column (nM)
Fig. 2.6 Inhibitory effect of peak V (5-MMC) from Mono S column on cell- free translation. Please refer to section 2.2.3 for details of the assay. 1 |iCi of tritiated leucine was added to each reaction tube. Data represent mean 土 standard deviation (N=3).
47 The inhibitory effect of peak V on the cell-free translation inhibition was also determined and shown in Fig. 2.6. The 50 % inhibitory concentration (IC50) was estimated to be 0.15 nM. The first seven amino acids at the iV-terminal of the peak V protein were shown to be DVNFGLA in an amino acid sequence analysis. This sequence is not identical to the iV-teraiinal sequences of the previously isolated M charantia RJDPs. The yields of the proteins at different steps of isolation and purification of Rff from M. charantia seeds are presented in Table 2.1. From 5 g ofseeds, 34 ^g of peak V proteins were isolated. It represented 0.01 % ofthe crude extract. Table 2.1 Protein yields at different stages of isolation and purification of RIP from five grams of Momordica charantia seeds. Purification stage 1 Amount ofProtein I Protein yield (%) Crude extract 230.0 mg 100 Affi-S 36.9 mg 16.1 Mono S peak V (5-MMC) I 34 呢 j ^ 2.4 Discussion In the attempt to isolate new RJPs fromth e seeds of Momordica charantia. Peak V from the Mono S column was isolated. This peak V shows the characteristics of type I Rffs. Firstly, it is a single protein as shown in SDS- PAGE (Fig. 2.4). Type H and JM RCPs have two or more subunits linked by disulphide bond and non-covalent interactions, tfpeak V is either ofthem, two or more proteins might have been seen in the SDS-PAGE. Meanwhile, as proteins in SDS-PAGE were run under denaturing conditions (with SDS and p_ mercaptoethanol), it is unlikely that undissociated subunits exist in the single
48 protein of peak V. Thus, it is also unlikely that peak V is type E or TV REPs. Secondly, the molecular weight of the peak V protein is about 30 kDa (Fig. 2.4). This is consistent with the molecular weights of other type I RPs. The N- glycosidase activity OFig. 2.5) of the peak V protein confirmed that it is a RJCP. Li repeated isolation of peak V from the seeds of M charantia, it was found that the retention time of peak V was the same at about 27 min. with a corresponding NaCl concentration of about 17 mM. It is different from the elution concentration of 35 and 53 mM NaCl reported for a- and P-MMC, respectively, in a similar isolation procedure (Go et al., 1992). Jn fact, from the result of SDS-PAGE (Fig. 2.4), it was confirmed that the peak V protein possessed a molecular weight higher than that of a-MMC. And, since P-MMC (28 kDa) is lighter than a-MMC (29 kDa), the peak V protein is unlikely to be either a- or P-MMC. It is unlikely to be y-MMC (11.5 kDa) either. Since the peak V protein is similar in molecular weight to MAP30, a N- terminal amino acid sequencing was carried out to investigate whether the peak V protein is MAP30. It was found that one out of the first seven A^-teraiinal amino acids was different between the peak V protein and MAP30. It is glycine at the fifth position of the peak V protein instead of asparatate in MAP30. Thus, the peak V protein is also not MAP30. As the peak V protein is not a-MMC, P-MMC, y-MMC or MAP30 but it showed iV-glycosidase activity as other RTPs，it is considered as a new Rff and
49 designated as 5-momorcharin (5-MMC) after the previously isolated momorcharins. 5-MMC is a type I RP isolated from the seeds of M. charantia. Like other RBPs, 5-MMC inhibited cell-free translation at a concentration as low as nanomolar level. From Fig. 2.6, it was found that the 50 % inhibitory concentration (IC50) of 5-MMC was 0.15 nM. Its effect was as high as a-MMC (IC50 = 0.12 nM) and p-MMC ^Cso = 0.11) (Yeung et aL, 1988). Nearly 90 % inhibition occured at a concentration of 1 nM. This showed that 5-MMC is as potent as other M. charantia seed Rffs. The yield of5-MMC was low (Table 2.1). Similarly, Rffs with low yields have been encountered in the isolation of a family of RDPs of saporins from Saponaria officinalis OFerreras et al., 1993). Jn a similar isolation procedure, 6.1 and 3.4 mg of a- and P-MMCs were isolated, respectively OFong et al., 1996). The yield of 5-MMC was 100 folds lower than oc- and p-MMC. Similar wide I differences in the yield of Rffs from the same tissue was also observed in the Saponaria officinalis (Ferreras et al., 1993). The yield of the four different saporins from the seeds varied from 20.7mg/5 g seeds to 0.75 mgy^5 g seeds. As will be discussed in the general discussion, this may be due to a tissue-specific expression. Li some other parts of the plant, the yield of 6-MMC may be higher.
50 Chapter 3 Study on a new ribosome-mactivating protein ^RIP) from Momordica charantia fruits
3.1 Introduction The fruit of Momordica charantia has been a traditional Chinese medicinal plant since ancient time. Despite the intensive studies of RPs from the seeds of M. charantia, there are few studies ofRffs from its fruits. Li an investigation of the hypoglycemic action of M. charantia fruit extract, it was observed that the extract induced uterine bleeding (Sharma et aL, 1960). This suggested that the fruit might have some abortifacient components. Two abortifacient proteins (a- and P-MMCs) from M. charantia were confirmed to be RJPs O^eung et al； 1988). But they were isolated from the seeds, not from the fruits. At present, one RJP was isolated from the fruits of M. charantia. It is the MAP30. This RDP exists in both the seeds and fruits of the plant ^.ee-Huang et cd； 1990). MAP30 may not be the only RJP in the fruit of M charantia. It was claimed that there may be more RJDPs in M. charantia 0-ee-Huang et aL, 1995a). A new Rff was purified from the fruits of M. charantia in the present investigation.
51 A method similar to that for the seeds was adopted with minor changes (Fig. 3.1). The seeds and the pulp were first removed. The fruit extract was then subjected to the precipitation by OMH4)2SO4. The proteins from the precipitate formed by addition of 30 % to 90 % saturated O^)2SO4 were applied to an Affi- Gel blue gel affinity column. The adsorbed proteins were eluted by 0.5 M NaCl. After dialysis, these proteins were resolved by FPLC on the cation-exchange Mono S column.
Fruit with seeds and pulp removed (i) Extract with 1 OmM NaH2PO4 buffer, pH 7.0; (ii) Precipitation by 30 % to 90 % saturated OSnftO2SO4inbuffer. N^ Crude extract
vi' Affinity chromatography on Affi-Gel blue gel 人 Breakthrough Salt-eluted proteins (i) Concentrate by ultrafiltration; (ii) Dialyze against 2mM phosphate buffer, 、， pH7.5. FPLC ion-exchange chromatography on Mono S
\y NewRTP Fig. 3.1 Procedure for the isolation of a new ribosome-inactivating protein fRBP) from the fruits ofMomordica charantia fruits.
52 3.2 Materials and methods 3.2.1 Materials Fresh unripe fruits of Momordica charantia (bitter gourd) were purchased from a local market. They were stored at -20°C before use. Before extraction, the fruits had their seeds and pulp removed. Affi-Gel® blue gel (100-200 mesh size), Mini-PROTEAN® n electrophoresis cell for SDS-PAGE and agarose (high strength analytical grade) were products of Bio-Rad O^.S.A.). Mono S HR 5/5 FPLC column was purchased from Pharmacia LKB (Sweden). SPECTRAyTOR® dialysis membranes were purchased from Spectrum (U.S.A.). Ammonium sulphate, diethyl pyrocarbonate QDEPC), bromophenol blue, ethidium bromide, hemin, creatine kinase and creatine phosphate were purchased from Sigma OJ.S.A.). Aniline and formamide were products of Merck & Co. (U.S.A.). Phenol and chloroform were purchased from USB OU.K.). The RNA marker was purchased from Gibco O^.S.A.). Photographic film model 667 was purchased from Polaroid O^J.S.A.). Tritiated leucine was purchased from Amersham 0^. K.). Rabbit reticulocyte lysate was obtained from rabbits raised at the University Animal House, the Chinese University of Hong Kong (C.U.H.K.). All other reagents were of analytical or molecular biology grade. Jf possible, they were autoclaved. Precautions for RNA work were taken throughout the process.
53 3.2.2 RIP isolation Preparation of crude extract Li a typical isolation, 500 g of unripe and deseeded fruits were blended in two volumes of 10 mM NaH2PO4, pH 7.0. After 1 hr., the extract was filtered through cheese cloth. It was then centrifuged at 9,500 rpm for 30 min. in a high speed centrifuge OBeckmann, J2MC) at 4°C. The supernatant was retained. Ammonium sulphate was then added to the supernatant to make a 30 % saturated solution. After stirring for 30 min. at 4°C, the solution was centrifuged at 9,500 rpm again at 4°C for 15 min. The supernatant was collected and ammonium sulphate was added to reach a final saturation of 90 %. It was stirred at 4°C for 30 min. and again centrifuged at 9,500 rpm and 4°C for 15 min. The precipitate was resuspended in 5 ml buffer. It was dialyzed against the buffer using a dialysis tubing with a molecular weight cut-off (MWCO) of3,500.
Affinity chromatography on Affi-Gel blue gel The dialyzed sample was loaded on an Affi-Gel blue gel column (1 x 7.2 cm), which had been equilibrated with 10 mM NaH2PO4, pH 7.0. The column was eluted with the same buffer at a flow rate of 1 nJAnin. After all the unbound proteins had been washed off (-25 ml), the column was eluted with 0.5 M NaCl in the buffer at 1 mVmin. Fractions of 1 ml each were collected. The fractions corresponding to the eluted peak were pooled and dialyzed against 2 mM
54 NaH2PO4, pH 7.5 using a dialysis tubing with a MWCO of 3,500. The dialyzed sample was concentrated through a filter membrane with a MWCO of 10,000 in an Amicon concentrator.
Chromatography on Mono S column The concentrated sample (�12 ml) was filtered through a 0.22 mM filter. The adsorbed fraction from the Affi-Gel blue gel column was filtered and then applied to a Mono S HR 5/5 fast protein liquid chromatography (FPLC) column (Pharmacia, Sweden) previously equilibrated with 2 mM NaH2PO4, pH 7.5. The Mono S column was connected to a Waters 600S/486 (U.S.A.) HPLC and UV monitor system, ti a typical run, 1 ml of the filtered sample was injected into a 1 ml sample loop and loaded on the column at a flow rate of 1 nJAnin. The breakthrough protein peak was collected. Afterward, another 1 ml sample was injected. The method included washing for 10 min. with buffer and a linear gradient of 0-60 mM NaCl in the buffer for 60 min. The eluted protein peaks were dialyzed against distilled water in a dialysis tubing with a MWCO ofl0,000.
55 3.2.3 Characterization Characterization using SDS-PAGE, iV-glycosidase assay and cell-free translation inhibition fRP) assay was carried out as described under section 2.2.3 in chapter 2.
3.3 Results During affinity colunm chromatography on Affi-Gel blue gel, most of the unbound proteins in the crude extract were washed off the colunm after elution with 25 ml buffer (Fig. 3.2). When the column eluted with 0.5 M NaCl in the buffer, a smaller peak was eluted. It was designated as the peak Affi-F (Affi-fruit extract). The Affi-F was resolved on the Mono S column (Fig. 3.3) into two peaks. The breakthrough peak was designated peak A. Another peak was the salt-eluted peak and it was designated peak B. Peak B was eluted at about 23 min at a NaCl concentration of about 13 mM. Peaks A and B were analysed by SDS-PAGE (¥ig. 3.4). Peak A was not homogeneous. It consisted mainly of proteins with a molecular weights ranging from above 20.1 kDa to about 64 kDa. Two major bands between 30 and 43 kDa were observed.
56 3.5 n:;::¾¾;;:;¾::^¾:¾;;;¾:¾;¾¾¾:;;^¾::;::: , % , - . ‘ ' “‘‘‘ ‘ .'‘,,, _廳______國讓___麗_咖^^^ ：；：: :¾^;:::::>:::::¾::::¾:;::>;;:;::;_____：：；：：：：；：：< ¥ ：：：：：：；：：：；：；；：：；：：> ：：：：：：? ：；：：：：：：：；：：：：：：：；< ：>> > I^iiiliiilii* ：；：：：> ^ ^ \ iliiiiiiiiliiiiiiiiiiiilii ^^^^^K^^m^^^^^^^^^^^^^^^^^^^^^^, );:i藝議懸靈i_«__iiI編_義iiiiili;i1丨_廳麗_騰•,_i_黎:_鍵i__il___p_l____^rt___ 安從丨漆丨丨翅：：丨热没丨丨：丨丨錢器； 骇丨丨:缀^K^^^^^^^^^^^^^^^^^^^^^_獲發緯::::丨凝丨欲溪凝鬆毅丨丨:缀毅錢凝:丨^ :Wf^
•*••口 , ^ ‘ ‘ , ‘ ‘ , “ , , ‘ ‘ ‘ ,,,, wm;^mM ^m0m^ _忿_;丨:丨丨丨§丨资丨激丨:丨丨:纏§_终騰丨丨丨譯餘_黎丨凝餘灘§錢丨丨:丨辨::丨丨丨戀沒兹箱窓丨;_:丨丨丨凝丨§:擬餘凝§ iliiiiiiiliiii 謹iiiiii__i_ _|___ iiiiipiiii*^漏_義^_____ 圖 v|v ::::::¾¾¾¾:':.:-:::!>: ¢-5;¾-:-:-:¾-: ;-:-:< ： ：•；•：•••：•：<•? . <-p'.-<::<'^<^^^y:i^^ .:•:. :::::.:;.:.:.:.:.:.:.:.?:.:.x.> ；•：；. •：•：•：•< ::与 :.:.:•::::::: .::::>::::>:。:投::::::好:.:.:::::::::::::::::::兴辞::::级:::；％::.々‘:；：知 .:::.:.:.x-次:.:-: 1 , ‘‘ � ’,’； OJMNaGl •'''' 1 _ ；：：：；：；：；^ ；：：：：：；：；- ：：；：. ^!1 丨丨: 丨: 这丨$:.;::: 吃:::; 丨::: ；:;::::::: 访丨妖; ：；:: ：> ：；：；：| ；> ：：^ ：«^ ：；5 ：：；：；; ；：；; ；- ® 1 “ 场_丨:丨丨丨丨丨:豸:丨教^^丨资§ :¾:¾;:¾;;;^:;;¾;;¾:¾;¾^ iPil 'MMHimwi i:»«»i 1：™»^^ __:錢聽____纖漏___釋^^ 毅丨：經：；：；：毅；：：；! iy ；：：；：：；：! : ：；; ；；：; ；? ；; ： ：；® ； ；? * ；：：：! ：：%$ ：! ；；« ；；：；：? ；^ «5 g;S ;®S; S;:"::;iS;AS:x^x>:x：；? 纖丨鐵切纖丨錢丨髮騰丨：凝：：潘毅丨理绍鼓丨_翁:::丨騰;凌翁丨:丨凝丨丨:丨§纖裔激丨丨丨凝!!丨凝翁翁遂_ ^¾¾^ ;¾:;:¾;:¾¾¾¾¾^ 丨縫；资漏囊MiMMmMi !!!：丨：驗毅證丨！！资錄资缀逾錄凝縫凝珍錄凝；：；丨；發翁錄:；：^；;§；；|^|；;；;|；|：||^ 05- ::..:��/'乂(广 Affl^ ,”.: •/'' \f*^ ‘ ‘ >"*v' ,J ^^^^^^^^^^^>>^^_X/^ ^^^ 0 -~I “‘ -—...-'i-'- I I �’""…. “‘I I 0 10 20 30 40 50 60 EMon vohttne (mT) Fig 3.2 Profile of elution of Momordica charantia fruit extract from Affi-Gel blue gel column (1 x 7.2 cm). Refer to section 3.2.2 for details. Briefly, crude extract from 500 g M charantia fruits was loaded to the Affi-Gel blue gel column. After washing with 25 ml of 10 mM NaH2PO4, pH7.0, the column was eluted with 0.5 M NaCl in buffer (indicated by the arrow). Fractions of 1 ml each were collected. The salt-eluted fractions (elution volume from 30 to 45 ml) were pooled to form Aff!-F. 57 OD28_^^__0 ConcentratioT^ ^,.,,n. of NaCl (mM) .:;y!:::::T:-ii^ '::::::::s::,,__:_ ！.=jjipw/j：!：^ 60 0.55_ / PeakA / 0.50— /
0.45— / J 0.40^ ^ /
一40 0.35- /i , --：-• • ,:. • ‘ • ./ •. 、'’-'、---- 0.30- /M
0.25-[ / 0.20- 二： , . . / / \ V ‘‘ / -20 0.15-j ,� / 謹-’ / PeakB 0.05“ / 0 |M 0
0 10 20 30 40 50 60 70 80 Elution time (min.) Fig. 3.3 Profile of elution ofAffi-F from Mono S column. Refer to section 3.2.2 for details. Briefly, the column was washed with 2mM NaH2PO4, pH 7.5 and eluted with a linear gradient ofO _ 60 mM NaCl in the same buffer at a flow rate of 1 ml/mln. Peaks A and B were collected.
58 Molecular weight (kDa) M 1 2 3 4 94.0^-~''� 67.0 _^ mm^ • ;,:;.:_:: .-_:.::.::::::.. .,.;, . .. - - ••;::;. ':.;.,:.:;,;::•::;•,:: 43.0 —• 4MMli i>^W!WSS!lwS^SwS
^:.¾'¾':;¾:^^ • ••：；••. :....,:. •... -• • ,. ：•：• yV'>:V >',' .•.•：：：: :; ： .•；：> • ''M••^<•l ':¾';':'¾';'¾<:¾¾ ••：••. . ；• .；• ••:::：•：••••：• '::<^^/'r'.i:m<::ym .•••-mmm^rc-i^;::::v:t:v':::#^iv;^^>o-T-,-^-^.-. ：.： .:.•...:. . ;::;;:%^vV.,y ‘:...:....;•:.::::::?;.::;”:::;-:•::.:.....,：••：； : •:^7：；：^：；：；\：；：&：；-：- ..;;..:.::.:.:.:,.;::.;;;:::.:.| ;::;::';:C;;V^;;;::'v" - • . . .::::-. ..:: ：••••；：•; ：；：••：• .:.::..:::..:.:.:.:.？：:..:.…^^^^^
Fig. 3.4 SDS-PAGE of peaks A and peak B (e-MMC) eluted from Mono S. The gel was 15 % acrylamide gel. Lane M, protein marker (Pharmacia); lane 1，peak A (breakthrough) from Mono S column (0.5 ^ig); lane 2，peak B from Mono S column (1.0 ^g); lane 3，a-momorcharin (a-MMC, 2 ^ig); lane 4，trichosanthin (TCS, 2 ^g). a-MMC (29 kDa) and TCS (26 kDa) are standard proteins.
Peak B consisted of a single protein which was 24 kDa in molecular weight. This was confirmed by comparison with the patttems of a-MMC and trichosanthin (TCS) in SDS-PAGE. The latter two proteins were 29 kDa and 26 kDa，respectively, in molecular weight. Peak B obviously had a lower molecular weight.
59 Fig. 3.5 shows the result of the iV-glycosidase assay. Li lane 2, the type I Rff trichosanthin (TCS) after aniline treatment released the characteristic Endo's band of about 400 bases from rRNA. No such band was observed for the lane without aniline treatment (lane 1). A similar result was observed for peak B (lanes 4 and 3). The 28 S rRNA band in lanes 2 and 4 disappeared. No such disappearence ofband was observed in lanes 1 and 3 without aniline treatment.
M 1 2 3 4 RNA marker - + - + (kilobase,kb) ^^^||||^^^^||^^^^| 4.42307 ~~^|^^^^H^^|^| DtntH^ 28S
"^^Hi^ff^ "Endo's band 0.24 ~^LnJHIHHHIIHHI Fig. 3.5 iV-glycosidase activity of peak B (s-MMC) from Mono S. Assay method was described in section 3.2.3. Sixty nanogram of rRNA from rabbit reticulocyte lysate was loaded to each sample lane. Lanes 1 & 2: positive control of trichosanthin (TCS, 0.33 mM); lanes 3 & 4: peak B from Mono S column (0.4 ^M). Lane M, RNA marker (Gibco). The "+"，，sign indicates aniline treatment while "-" sign indicates no aniline treatment.
60 The cell-free translation-inhibitory activity of peak B was shown in Fig. 3.6. The 50 % inhibitory concentration C^C50) of peak B on the translation in rabbit reticulocyte lysate was 0.17 ^M (between 0.15 and 0.2 _•
75 "]
g65- yf^" 广 / ^ 55- / ^ T / �5 0 - / I 45 / 8 45 - y^ i 53 T ^ 40 - I y/ 35 - ^^ 30 I I "~I I I I I I iiiiimi I 1 ‘ I I I I I I iiiiiiiii I 1~~I I 10� 1Q1 1Q2 103 104 Concentration of peak B (e-MMC) from Mono S column (nM)
Fig. 3.6 Inhibitory effect of peak B from Mono S column on cell- free translation. Please refer to section 3.2.3 for details ofthe assay. 1 ^iCi of tritiated leucine was added to each reaction tube. Data represent mean 土 standard deviation (N=3).
61 The protein yields in the isolation and purification process are presented in Table 3.1. From 500 g deseeded fruits, 51.8 mg of crude extract was obtained. There were 3.2 mg protein adsorbed on the Affi-Gel column. About 94 % ofthe crude extract did not bind to the Affi-Gel colunm and were eluted as the breakthrough fraction. Peak B from the Mono S colunm contained 62 ^g material. More than 90 % of the Affi-F protein did not adsorb on Mono S.
Table 3.1 Protein yields at different stages of purification ofRff from five hundred grams of Momordica charantia fruits. Purification step | Amount ofProtein Protein yield (%) Crude extract 51.8 mg 100 Affi-F I 3.2 mg 6.18 Mono S peak B 62 ^ig 0.12
3.4 Discussion Fruits usually have only a little amount of proteins. This may explain why not many RDPs have been isolated from the fruits of a plant. Despite this, a new Rff has been isolated fromth e fruits of Momordica charantia in the present study. It was designated 8-momorcharin (s-MMC). ii the running of the Mono S column, only peaks A and peak B were obtained. The purity of the proteins from these two peaks was investigated by
62 SDS-PAGE. It was found that peak A was composed of several proteins while peak B consist of only a single protein. To investigate whether peak B is a RJDP，it was examined for A^-glycosidase activity. The molecular weight of the bands for peak B is similar to that of the known type I Rff, trichosanthin (TCS, Fig. 3.5). A characteristic Endo's band was observed for both proteins. The concomitant fainting of the 28 S rRNA band was due to cleavage of the rRNA and release of the Endo's band. For both proteins, there was no Endo's band in the non-aniline-treated "-" lanes. So, they were not RNases but specific iV-glycosidases. These results showed that peak B was a RCP and we designated it as s-momorcharin, s-MMC. s-MMC was not as potent as the other REPs. From Fig. 3.6, it was found that it had a 50 % inhibitory concentration (IC50) in the cell-free translation inhibition assay of 0.17 ^M. This is distinct from many other type I Rffs which had IC50 values in a similar assay ranging from 10 nM to 0.01 nM (Table 1.6). The significant deviation of the activity of s-MMC from other type I RPs may reflect some structural differences fromother s as will be discussed in chapter 4. The yield of s-MMC was not high. Only 62 ^g of s-MMC, was obtained from 500 g M. charantia fruits. The existence of a weak REP at a very low level in the fruit seems difficult to be explained. It seems ineffective for s-MMC to be functional at such a low level. The isolation of s-MMC may provide some clues about the physiological fimction of REPs in plants.
63 Chapter 4 General Discussion and Conclusion
4.1 General discussion The objective of this study was to isolate new RJDPs from Momordica charantia which could be utilized in the future for the production of new immunotoxins. As a result of the study, two new Rffs have been isolated from Momordica charantia. They were designated as 5-MMC and s-MMC respectively, after the previously isolated a-, P- and y-MMC. As discussed in the previous two chapters, these two MMCs were not the Rffs previously isolated from M. charantia.
Differences between 5- and s-MMC The coincidence that the retention times for 5- and s-MMC in the Mono S colunm were similar (within the range from 23 min. to 27 min.) has prompted the question whether the two MMCs were actually the same protein. Jn repeated isolation of 5- and s-MMC, it was found that the retention times for both ofthem were still the same (at 27 and 23 min. for 5- and s-MMC, respectively). When 5- and s-MMCs were subjected to SDS-PAGE, the difference became obvious. Their molecular weights were 30 kDa and 24 kDa, respectively (Fig. 2.5 and 3.4). In the investigation of the cell-free translation-inhibitory activity of these two
64 MMCs, a wide difference in activity was also observed OFig. 2.6 and 3.6). So, the two MMCs were actually distinct proteins.
Relationship of 5- and s-MMC with other Momordica charantia REPs 6- and s-MMC seemed to exist exclusively in the seeds and fruits, respectively. When M. charantia seeds extract was applied to Mono S column, three out of ten peaks were identified to be RPs. Two of them were a- and p_ MMC as isolated in a previous study (Go et al., 1992). The remaining one was 6- MMC. A similar procedure was adopted for the isolation of fruit Rffs in M charantia. There was no peak for 5-MMC in the Mono S column elution profile when fruit extract was loaded (Fig. 3.3). On the other hand, the peak corresponding to s-MMC was also absent in the Mono S column elution profile for the seed extract (Fig. 2.3). The seemingly exclusive existence of 5- and s-MMC in the seeds and fruits, respectively, but not vice versa, may be due to differences in the gene product. Li a distribution study on saporins (type I REPs) from Saponaria officinalis, it was found that saporins from this plant could be categorized into leaf, root and seed Rffs (Terreras et al., 1993). Within the same tissue, isoforais of RJPs have similar amino acid compositions, iV-terminal amino acid sequences and even iV-glycosidase activities towards ribosomes. For example, the values of 50 % inhibitory concentration (IC50) of saporin-S5 and saporin-S6 from seeds in a
65 cell-free translation inhibition assay (Rff assay) using the ribosome from rabbit reticulocyte lysate were 0.05 and 0.06 nM, respectively OFerreras et al” 1993). Those for saporin-Rl, saporin-R2 and saporin-R3 from the roots were 0.86, 0.47 and 0.48 nM, respectively. Meanwhile, the IC50 of saporin-R3 from the roots in another RJP assay (using ribosomes from the plant of Vicia sativa) was 0.02 nM while that of saporin-L2 from the leaves was 20.91 nM ^'erreras et al., 1993).
3 The difference was 10 folds. These results led to the suggestion that the expression of saporins was tissue-specific and different saporins originated from different genes of the same family ^'erreras et al., 1993). This is supported by the fmding that saporin genes form a multigene family (Benatti et al., 1991). By the same token, 5- and s-MMC may also be the products of different genes and their expression may be tissue-specific. The most striking evidence was that there were large differences between the activities of these two new MMCs in the RH> assay (Fig. 2.6 and 3.6). The IC50 of5-MMC in this assay was 0.15 nM while that of s-MMC was 0.17 ^M. The difference was as large as 10^ folds.
Yields of 5- and s-MMC The yields of 6- and s-MMC were quite low as well. From 1 g of seeds, 6.8 jig of 6-MMC were obtained. It was 0.01 % ofthe crude protein extract (46 mg，100 %) (Table 2.1). From 1 g of fruits, 0.12 ^ig (0.1%) of s-MMC was obtained out of 103.66 昭(100%) crude extract (Table 3.1). Thier yields were
66 relatively low. Li a similar experiment for the isolation of a- and P-MMC, the yields were respectively 1.24 mg and 0.68 mg per gram of tissue (Fong et al., 1996). This reflects the low abundance of 5- and s-MMC in the tissue. There were also low abundance of some REPs among others in a tissue. For example in Saponaria officinalis, there were at least seven RBPs (saporins) from different tissues. Li the seeds, there were four different saporins. Their yields varied from 2.23 mg to 0.08 mg per gram of seeds. There was an approximately 30 folds difference. This difference may stem from the tissue-specific expression of different members of a multi-gene family as discussed above. The low yields of5 -and s-MMC may be due to expression of different genes. The expression of Rffs in the fruits seems to be a weak process. Apart from the low yield of s-MMC, the yield of a RJP from the fruits ofanother plant is also low. Pepocin is a type I Rff isolated from the fruits of Cucurbita pepo (Yoshinari et al., 1996). From 200 g of fruits from Cucurbita pepo, only 5.4 mg ofpepocin were isolated (27 ^g / g tissue). There is no data available on the seed RJDP in Cucurbita pepo. But, at least by comparing the yield of seed RP e.g. a- and p-MMC from M, charanta (1.24 mg and 0.68 mg / g tissue, Fong et cd., 1996)，it is shown that the yields for seed Rffs is much higher. This is consistent with the notion that the Rffs from different tissues are originated from different genes. This may also apply to Rffs in M. charantia.
67 Possible physiological function The low yield offruit Rffs in comparison with that of seed Rffs may have implications about the physiological function ofRffs. It was suggested that RPs may have metabolic function in the germination of seeds (Barbieri et aL, 1993). The relatively low yield ofs-MMC and the high yields of a- and P-MMC in seeds indirectly support this notion. But the exact role of RPs in seeds is still a mystery. Meanwhile, RIPs not only exist in seeds, they also exist in different parts of the plant, hi some cases, the activities of REPs in tissues other than seeds are even higher than those of seed RBPs (Barbieri et al., 1993; Ferreras et aL, 1993; Gasperi-Campani et al., 1985). The possible function in seed germination may not be the only function of RDPs. Further studies on the distribution of Rff will provide more information on the function ofRffs. On the other hand, the physiological fimction ofRffs may be related to the defense of plants (Stirpe and Barbieri, 1986). It was hypothesed that different Rffs may play different roles to defend the plant as an immune system ^"erreras et aL, 1993). There is no report about whether the concentration of RDPs will increase on encounter with pathogens. Further studies in this direction will throw more light on the physiological role ofRffs.
t
68 Chromatographic behavior on Affi-Gel blue gel During the isolation process, MAP30 and y-MMC were not isolated. Among the ten peaks in the Mono S column elution profile, only the peaks corresponding to a-, P_ and 5-MMC ( peak X，VEI and V，respectively, in Fig. 2.3) possessed iV-glycosidase activity. It was suspected that MAP30 and y-MMC might have co-eluted with other proteins unadsorbed on the Mono S colunm. But SDS-PAGE for the breakthough of Mono S column ^Fig. 2.4) indicated that there were no protein bands with a molecular weight of 30 kDa or 11.5 kDa (corresponding to the molecular weights of MAP30 and y-MMC, respectively). There was also no evidence for the existence ofMAP30 or y-MMC in the Mono S colunm breakthrough when the fruit extract was applied (Fig. 3.3). As the sample loaded onto the Mono S colunm was derived from the salt- eluted fractions from the Affi-Gel blue gel column, MAP30 and y-MMC might have been eluted with the breakthrough proteins of that affinity colunm. to a study of the binding of RJPs to Affi-Gel blue gel, a correlation between glycosylation of RJP and binding onto this kind of affinity gel was observed OVIunoz et al., 1990). y-MMC is non-glycosylated (Pu et al., 1996). This property might have prevented it from binding to the Affi-Gel blue gel. Deglycosylation of ricin A-chain did not affect the binding of this subunit of type n Rff to bind to Affi-Gel blue gel (Munoz et al., 1990). Some other unknown properties might also prevent some Rffs from binding to the Affi-Gel
69 blue gel. This might explain why MAP30 was not detected in the salt-eluted fractions fromth e affinity column. Since the breakthrough proteins were mixed with different proteins like RNase，it was unable to identify any REPs inside the breakthrough fraction by the A^-glycosidase assay. All the 28S/18S RNA or Endo' s band was digested and appeared as a smear (data not shown). From the above discussion, it seems likely that, with the method used in this project, some unknown RIPs may have been missed just like MAP30 and y- MMC. It was claimed that eight isoforais of genes ofRffs were identified in M charantia leaves (Lee-Huang et al” 1995a). Three ofthem were claimed to be a-, P-MMC and MAP30. The identities of the remaining five isoforms were not available. The y-MMC isolated recently and the 5-MMC and s-MMC isolated in this project may be members of the eight isoforms. Jf it is true, there are at least two more RJPs which can be isolated from M. charantia.
Possible use in the construction of immunotoxins The presence of more than one form of Rff in M charantia will help circumvent the problems which arise from the use of Rff-conjugated hnmunotoxins (ITs). Jn the therapies using Rff-conjugated ITs, immune responses against the YT will make prolonged treatment impossible. An effective solution is to switch to new RJPs after prolonged IT therapy (Battelli et aL, 1996). This would prevent immune responses against the Rff moiety.
70 New Rffs with low IC50 values will be useful. It is because a lower dose of ITs can produce an effect high enough to kill diseased cells in the body. This minimizes the risk of immune response against the IT. Table 4.1 shows that the IC50 of 6-MMC is one of the lowest among the REPs isolated from M. charantia. It is worthwhile to investigate whether this strong activity can be used to compact tumour, or other undesirable cells.
The activity of s-MMC is quite low (IC50 at 0.17 ^M). However, since the values ofIC50 varies in different assay systems (Ferreras et aL, 1993)，it is possible that s-MMC may be very active in some other systems and effective in killing some types of cells.
Table 4.1 The inhibitory activities ofREPs from Momordica charantia in cell- free translation inhibition assay. M. charantia RIPs IC50 (nM) References a-MMC ^ O^eungetal., 1988) P-MMC 0.11 O^eimg et aL, 1988) Y-MMC 55.00 (P« 过这1； 1996) MAP30 3.30 OLee-Huang et al.’ 1995a) 5-MMC 0.15 This thesis s-MMC l]0 This thesis IC50： 50 % inhibitory concentration.
In the future, experiments such as cytotoxicity studies using human cell lines should be done to characterize the effects of 5- and s-MMC and evaluate the feasibility for them to form ITs.
71 To sum up, the isolation of 6- and s-MMC has provided new possible conjugates for the production of immunotoxins. Meanwhile, they may also provide important information for the gene structure and so the evolution of RIPs in M charantia. They are also important for the study on the distribution ofRffs in plants (at least in M charantia).
4.2 Conclusion Momordica charantia contains several ribosome-inactivating proteins (RBPs). The previously isolated M. charantia REPs are a-momorcharin (a-MMC), P-momorcharin (P-MMC), y-momorcharin (y-MMC) and Momordica anti-HW protein 30 (MAP30). fc this project，two new RTPs，namely 5- and s-momorcharin (6- and s-MMC), have been isolated from the seeds and fruits of Momordica charantia, respectively. 5-MMC is similar to other M. charantia RBPs in many aspects. It is a single-chain protein with a molecular weight of 30 kDa. It has A^-glycosidase and cell-free translation-inhibitory activities which are as strong as those of the other M. charantia RH^s previously isolated. But the yield of 5-MMC is low. s-MMC is also a single-chain protein as 5-MMC. Its molecular weight is 24 kDa. Despite these similarities with other M. charantia RTPs, the N- glycosidase and cell-free translation-inhibitory activities of s-MMC are less active
72 than other M. charantia RH>s. Similar to 5-MMC, the yield of s-MMC is also low. The presence of more than one form of RPs provides alternatives for the production of different Rff-based immunotoxins containing different Rffs. This would prevent immune responses against the RP moieties in the immunotoxins. The exclusive existence of 5- and s-MMC in the fruits and seeds of M. charantia may indicate the presence of tissue specific expression of RBPs in M. charantia. Meanwhile, the presence of low yield RJPs (5- and s-MMC) among other RJPs and the expression of low activities of s-MMC in M charantia may provide some clues about the physiological role, such as germination and defense role, ofRJPs in M. charantia and other plants.
73 References
Ali，L., Khan, A.K.A., Mamun, M.LR., Mosihuzzaman, M., Nahar, N., Nur-e- Alam, M. and Rokeya, B. (1993). Studies on hypoglycemic effects offruitpulp, seed, and whole plant of Momordica charantia on Normal and diabetic model rats. Planta. Med., 59, 408-412. Barbieri, L., Battelli, M.G. and Stirpe, F. (1993). Ribosome-inactivating proteins from plants. Biochim. Biophys. Acta, 1154,237-282. Barbieri, L., Bolognesi, A., Cenini, P., Falasca, A.L, Garofano, L., Guicciardi, A., Lappi, D., Miller, S.F., Min^etti, A. and Stirpe, F. (1989). Ribosome- inactivating proteins from plant cells in culture, Biochem. J., 257, 801-807. Barbieri, L., Aron, G.M., frvin, J.D. and Stirpe, F. (1982). Purification and partial characterization ofanother form of the antiviral protein from the seeds ofPhytolacca americana L (pokeweed). Biochem. J., 203, 55-59. Battelli, M.G., Polito, L., Bolognesi, A., Lafleur, L., Fradet, Y. and Stirpe, F. (1996). Toxicity ofribosome-inactivating proteins-containing immunotoxins to a human carcinama cell line. Lit. J. Cancer., 65, 485-490. Benatti, L., Nitti, G., Solinas, M., Valsasina, B., Vitale, A., Ceriotti, A. and Soria, M. (1991). A saporin-6 cDNA containing a precursor sequence coding for a carboxyl'terminal extension. FEBS Lett., 291, 285-288. Bourinbaiar, A.S. and Lee-Huang, S. (1996). The activity of plant-derived antiretroviral proteins MAP30 and GAP 31 against herpes simplex virus infection in vitrio. Biochem. Biophys. Res. Comm., 219，923-929. Bradley, J.L., helder, S.M. and McGuire, P.M. (1987). Depurination ofyeast 26S ribosomal RNA by recombinant ricin A chain. Biochem. Biophys. Res. Commun.，149, 558-593. Cammarano, P., Teichner, A.，Londei, P., Acca, M., Nicolaus, B., Sanz, J.L. and Amils，R. (1985). Insensitivity of archaebacterial ribosomes to protein synthesis inhibitors. Evolutionary implications. EMBO J., 4,811-816. Chan，W.Y., Ng, T.B. and Yeung, H.W. (1992). P-momorcharin, a plant glycoprotein, inhibits synthesis ofmacromolecules in embryos, splenocytes and tumor cells. M. J. Biochem.，24，1039-1046.
74 Chan, W.Y., Tam, P.P.L., So, K.C. and Yeung, H.W. (1985). The inhibitory effects of P-momorcharin on endometrial cells in the mouse. Contraception, 31, 83-90. Chan, W.Y., Tam, P.P.L. and Yeung, H.W. (1984). The termination of early pregnancy in the mouse by ^-momorcharin. Contraception, 29, 91-100. Citores, L., Ferreras, J.M., Iglesias, R., Carbajales, M.L., Arias, FJ., Jimenez, P., Rojo, M.A. and Girbes, T. (1993). Molecular mechanism of inhibition of mammalian protein synthesis by some four-chain agglutinins. FEBS Lett., 329, 59-62. Claflin，A.J., Vesely, Dl.，Hudson, J.L., Bagwell, C.B., Lehotay, D.C., Lo, T.M., Fletcher, M.A., Block, N.L. and Levey, G. (1978). Inhibition ofgrowth and guanylate cyclase activtiy of an undifferentiated prostate adenocarcinoma by an extract ofthe balsam pear (Momordica charantia abbreviata). Proc. Natl. Acad. Sci. USA., 75, 989-993. Dinota, A.，Barbieri, L., Gobbi, M.，Tazzari, P.L., Rizzi, S., Bontadini, A., Bolognesi, A., Tura, S. and Stirpe, F. (1989). An immunotoxin containing momordin suitable for bone marrow purging in multiple myeloma patients. Br. J. Cancer, 60，315-319. Endo, Y., Huber, P.w. and Wool, I.G. (1990). RNA-protein interaction: an analysis with RNA oligonucleotides of the recognition by alpha-sarcin of a ribosomal domain critical for function. J. Biol. Chem., 258, Endo, Y. and Wool, I.G. (1982). The site of action of alpha-sarcin on eukaryotic ribosomes. The sequence at the alpha-sarcin cleavage site in 28S ribosomal ribonucleic acid. J. Biol. Chem., 257, 9054-9057. Falini, B., Falini, B., Bolognesi, A., Flenghi, L., Tazzari, P.L., Broe, M.K., Stein, H., Durkop, H., Aversa, F., Comeli, P., Pizzolo, G., Barbabietola, G., Sabattini, E.，Pileri, S., Martelli, M.F. and Stirpe, F. (1992). Response of refractory Hodgkin's disease to monoclonal anti-CD30 immunotoxin. Lancet, 339, 1195- 1196. Fando，J.L., Alaba, I.，Escarmis, C., Femandez-luna, J.L., Mendez, E. and Salinas, M. (1985). The mode of action of restrictocin and mitogillin on eukaryotic ribosomes. Inhibition of brain protein synthesis, cleavage and sequence ofthe ribosomal RNA fragment. Eur. J. Biochem., 149, 29-34. Ferreras, J.M., Barbieri, L., Girbes, T., Battelli, M.G., Rojo, M.A., Arias, FJ., Rocher，M.A., Soriano, F., Mendez, E. and Stirpe, F. (1993). Distribution and
75 properties ofmajor ribosome-inactivating proteins (28 S rRNA N-glycosidases) of the plant Saponaria officinalis L. (Caryophyllaceae). Biochim. Biophys. Acta, 1216,31-42. Fletcher, M.A., Caldwell, K., Claflin, AJ. and Malinin, T. (1980). Further characterization of a tumor cell growth inhibitor from the balsam pear. Proc. Fed. Am.Soc. Exp. Biol., 39，414. Fong, W.P., Poon, Y.T., Wong, T.M., Mock, W.Y., Ng, T.B., Wong, R.N.S., Yao, Q.Z. and Yeung, H.W. (1996). A highly efficient procedure for purifying the ribosoome-inactivating proteins a- and fi-momorcharins from Momordica charantia seeds, N-terminal sequence comparison and establishment of their N-glycosidase activity. Life Sci., 59, 901-909. Gao, W., Ling, J., Zhong, X., Liu, W., Zhang, R., Yang, H., Cao, H. and Zhang, Z. (1994). Luffin S-a small novel ribosome-inactivating protein from Luffa cylindrica. FEBS Lett., 347，257-260. Gasperi-Campani, A., Barbieri, L., Battelli, M.G. and Stirpe, F. (1985). On tthe distribution of ribosome-inactivating proteins amongst plants. J. Nat. Prod” 48,446-454. Gasperi-Campani, A., Berbieri, L., Lorenzoni, E. and Stirpe, F. (1980). Seed extracts inhibiting protein synthesis in vitro. Biochem. J., 186，439-441. Gasperi-Campani, A., Barbieri, L., Lorenzoni, E., Montanaro, L., Sperti, S., Bonetti, E. and Stirpe, F. (1978). Modeccin, the toxin of Adenia digitata. Purication, toxicity and inhibition of protein synthesis in vitro. Biochem. J., 174，491-496. Gelfi, C.，Bossi, ML.，Bjellqvist, B. and Righetti, P.G. (1987). Isoelectric focusing in immobilized pH gradients in the pH 10-11 range. J. Biochem. Biophys. Methods, 15, 41-48. Girbes, T., Citores, L., Ferreras, J.M., Rojo, M.A., Iglesias, R., Munoz, R., Arias, FJ., Calonge, M., Garcia, J.R. and Mendez, E. (1993a). Isolation and partial characterization of nigrin b, a non-toxic novel type 2 ribosome-inactivating protein from the bark of Sambucus nigra L. Plant Mol. Biol., 22, 1181 -1186. Girbes, T., Citores, L., Iglesias, R., Ferreras, J.M., Munoz, r., Rojo, M.A., Arias, FJ., Garcia, J.R., Mendez, E. and Calonge, M. (1993b). Ebulin 1’ a nontoxic novel type 2 ribosome-inactivating protein from sambucus ebulus L Leaves. J. Biol. Chem., 268, 18195-18199.
76 Go, T.T.M., Yeung, H.W. and Fong, W.P. (1992). Deoxyribonucleolytic cativity of a- and P-momorcharins. Life Sci., 51, 1347-1353. Habuka, N., Akiyama, K., Tsuge, H., Miyano, M., Matsumoto, T. and Noma, M. (1990). Expression and secretion of Mirabilis antiviral protein in Escherichia coli and its inhibition of in vitro eukaryotic and prokaryotic protein synthesis. J. Biol. Chem., 265,10988-10992. L:vin, J.D. (1975). Purification and partial characterization of the antiviral protein from Phytolacca americana which inhibits eukaryotic protein synthesis. Arch. Biochem. Biophys., 169, 522-528. frvin, J.D., Kelly, T. and Robertus, J.D. (1980). Purification and properties ofa second antiviral protein from Phytolacca americana which inactivates eukaryotic ribosomes. Arch. Biochem. Biophys., 200，2. Kanellos, J., Mckenzie, I.F. and Pietersz, G.A. (1989). Intratumour therapy of solid tumours with ricin-antibody conjugates. Lnmunol. Cell. Biol., 67，89-99. Kataoka, J., Miyano, M., Habuka, N., Masuta, C. and Koiwai, A. (1993). A genomic gene for MAP, a ribosome-inactivating protein from Mirabilis jalapa, contains an intron, Nucleic Acids Res., 21, 1035. Kataoka, J., Habuka, N.，Furuno, M.，Miyano, M., Takanami, Y. and Koiwai, A. (1991). DNA sequence of Mirabilis antiviral protein (MAP), a ribosome- inactivating protein with an antiviral property, from mirabilis jalapa L and its expression in Escherichia coli. J. Biol. Chem., 266, 8426-8430. Kubota, S., Yeung, H.W. and Yang, J.T. (1986). Conformation ofabortifacient proteins: trichosanthin, a-momorcharin and P-moomorcharin. Biochim. Biophys. Acta, 871，101-106. Laemmli, U.K. and Favre, M. (1973). Maturation of the head of bacteriophage T4. J. Mol. Biol., 80，575-599. Lau, C.F. (1983) Histopathological and ultrastructural changes of mouse placenta caused by two plant abortifacient proteins: trichosanthin and a-momorcharin, M. Phil Thesis. Chinese University ofHong Kong, Hong Kong. Lee-Huang, S., Huang, P.L., Chen, H.C., Huang, P.L., Boimnbaiar, A., Huang, H.I. and Kung, H.F. (1995a). AntUHW and anti-tumor activities of recombinant MAP30 from bitter melon. Gene, 161，151-156.
77 Lee-Huang, 8丄.’ Huang，Pl.，Huang, P1., Bourinbaiar, A.S, Chen, H.C. and Kung, H.F. (1995b). Inhibition of the integrase of human immunodeficiency vinis (HIV) type 1 by anti_mVplant proteins MAP30 and GAP 31, Proc. Natl. Acad. Sci. USA, 92, 8818-8822. Lee-Huang, S.，Huang, Pi” Kung, H.F., Q., L.B., Huang, P.L., P., H, Huang, H.I. and Chen, H.C. (1991). TAP 29: an anti-human immunodeficiency virus proteinfrom Trichosanthes kirilowiL Proc. Natl. Acad. Sci. U. S. A., 88，6570- 6574. Lee-Huang, S., Huang, P.L., Nara, P.L., Chen, H.C., Kung, H.F., Huang, P., Huang, H.I. and Huang, P1. (1990). MAP30: a new inhibitor ofMY-l infection and replication. FEBS Lett., 272, 12-18. Leung, K.N., Yeung, H.W. and Leung, S.0. (1986). The immunomodulatory and antitumour activities of trichosanthin-An abortifacient protein isolatedfrom Tien-hua-fen (Trichosanthes kirilowii). Asian Pacif. J. Allergy tomun., 4，111- 120. Leung, S.0., Yeung, H.W. and Leung, K.N. (1987). The immunosuppressive activities oftwo abortifacient proteins isolated from the seeds ofbitter melon (Momordica charantia). ^nmunopharmacology, 13，159-171. Lord, M.J. (1985a). Precursors of ricin and Ricinus communis agglutinin. Glycosylation and processing during synthesis and intracellular transport. Eur. J. Biochem., 146，411-416. Lord, M.J. (1985b). Synthesis and intracellular transport of lectin and storage protein precursors in endosperm from castor bean. Eur. J. Biochem., 146, 403-409. Lowry, O.H., Rosebrough, NJ., Farr, A.L. and Randall, RJ. (1951). Protein measurement with the Folinphenol reagent. J. Biol. Chem., 193，265-275. Mock, J.W.Y., Ng, T.B., Wong, R.N.S., Yao, Q.Z., Yeung, H.W. and Fong, W.P. (1996). Demonstration of ribonuclease activity in the platn ribosome- inactivatingproteins alpha- and heta-momorcharins. Life Sci., 59, 1853-1859. Monzingo, A.F. and Robertus, J.D. (1992). X-ray analysis ofsubstrate analogs in the ricin A-chain active site. J. Mol. Biol., 227，1136-1145. Munoz，S.M., Caldera, M., Canevan, S., Tosi, E., Cogliati, T.，I.，C.M. and Conde, F.P. (1990). Fractionation ofthe ribosome-inactivating protein preparations yvith triazine dyes. Biochem. Biophy, Res. Comm., 173, 554-560.
78 Ng, T.B., Liu, W.K., Sze, S.F. and Yeung, H.W. (1994). Action of a- mornorcharin, a rihosome-inactivating protein, on cultured tumor cell lines. Gen. Pharmac., 25, 75-77. Ng, T.B., Chan, W.Y. and Yeung, H.W. (1992). Proteins with abortifacent, ribosome-inactivating, immunomodulatory, antitumor and anti-AIDS activities from Cucurbitaceaeplants. Gen. Pharmac., 23, 575-590. Ng, T.B., Wong, C.M., Li, W.W. and Yeung, H.W. (1987). Peptides with antilipolytic and lipogenic activties from seeds of teh bitter gourd Momordica charantia (family Cucurbitaceae). Gen. Pharaiac., 18，275-281. PeUiam, H.R.B. and Jackson, RJ. (1976). An efficient mRNA-dependent translation system from reticulocyte lysates. Eur. J. Biochem., 67，247-256. Prestle, J., Schonfelder, M., Adam, G. and Mundry, K.W. (1992). Type I ribosome-inactivating proteins depurinate plant 25S rRNA without species specificity. Nucleic Acids Res., 20, 3179-3182. Pu, Z, Lu, B.Y., Liu, W.Y. and Jin, S.W. (1996). Characterization of the enzymatic mechanism of y-momorcharin, a novel ribosome-inactivating protein with lower molecular weight of 11,500 purified from the seeds of bitter gourd (Momordica charantia). Biochem. Biophys. Res. Comm., 229, 287-294. Ramakrishnan, S., Fryxell, D.，Mohanraj, D., Olson, M. and Li, B.Y. (1992). Cytotoxic conjugates containing translational inhibitory proteins. Annu. Rev. Pharmacol. Toxicol., 32, 579-621. Roberts, L.M. and Lord, J.M. (1981). The synthesis of Ricinus communis agglutinin, cotranslational and posttranslational modification of agglutinin polypeptides. Eur. J. Biochem., 119，31-41. Sharma, V.N., Sogani, R.K. and Avora, R.B. (1960). Some observations on hypoglycemic activity ofMomordica charantia. M. J. Med. Res., 48, 471-477. Srivastava, Y., Venkatakrishna-Bhatt, H. and Verma, Y. (1988). Effect of momordica charantia linn, pomous aqueous extract on cataractogenesis in murrin alloxan diabetics. Pharmacol. Res. Comm., 20, 201-209. Stirpe, F., Wawrzynczak, EJ., Brown, A.N.F., Knyba, R.E., Watson, GJ., Barbieri, L. and Thorpe, P.E. (1988). Selective cytotoxic activity of immunotoxins composed of a monoclonal anti-Thy 1.1 antibody and the ribosome-inactivating proteins bryodin and momordin. Br. J. Cancer, 58, 558- 561.
79 Stirpe, F. and Barbieri, L. (1986a). Ribosome-inactivating proteins up to date. FEBS Lett., 195,1-8. Stirpe, F., Barbieri, L.，Battelli, M.G., Falasca, A.L, Abbondanza, A., Lorenzoni, E. and Stevens, W.A. (1986b). Bryodin, a ribosome-inactivating protein from the roots ofBryonia dioica L (white bryony). Biochem. J., 240，659-665. Stirpe, F., Barbieri, L., Abbondanza, A., Falasca, A.L, Brown, A.N.F., K., S., Olsnes, S. and Pihl, A. (1985). Properties of volkensin, a toxic lectin from Adenia volkensii, J. Biol. Chem.., 260, 14589-14595. Stirpe, F., Saperi-Campani, A., Barbieri, L., Falasca, A.L, bbondanza, A. and Stevens, W.A. (1983). Ribosome-inactivating proteins from the seeds of Saponaria officinalis L (soapwort), of Agrostemma githago L (corn cockle) and of Asparagus officinalis L. (asparagus), and from the latex of Hura crepitans L. (sandbox tree). Biochem. J., 216, 617-625. Strockbine, N.A., Marques, L.R.M., Newland, J.W., Smith, H.W., Hohnes, R.K. and O'Brien, A.D. (1986). Two toxin-converting phages from Escherichia coli 0157:H7 strain 932 encode antigenically distinct toxins with similar biologic activities. Mect. Lmnun. 53, 134-140. Takemoto, DJ., Dunford, C. and McMurray, M.M. (1982). The cytotoxic and cytostattic effects of the bitter melon (Momordica charantia) on human lymphocytes. Toxicon, 20, 593-599. Terenzi, A.，Bolognesi, A., Pasqualucci, L.，Flenghi, L., Pileri, S., Stein, H., Kadin, M., Bigema, B., Polito, L., Tazzari, P.L., Martelli, M.F., Stirpe, F. and Falini, B. (1996). Anti-CD30 (BER-H2) immunotoxins containing the type-l ribosome-inactivating proteins momordin and PAP-S(pokeweed antiviral protein from seeds) display powerful antitumour activity against CD30+ tumour cells in vitro and in SCID mice. Br. J. Haematol., 92, 872-879. Toyokawa, S.，Takeda, T. and Ogihara, Y. (1991). Isolation and characterization of a new abortifacient protein, karasurin, from root tubers of Trichosanthes kirilowii Max. var. japonicum Kitam. Chem. Pharm. Bull., 39, 716-719. Wawrzynczak, EJ., Cumber, AJ., Henry, R.V., May, J., Newell, D.R., Pamell, G.D., Worrell, N.R. and Forrester, J.A. (1990). Pharmacokinetics in the rat of a panel ofimmunotoxins made with abrin A chain, ricin A chain, gelonin, and momordin. Cancer Res., 50, 7519-7526.
80 Yeung, H.W., Li, W.W. and Ng, T.B. (1991). Isolation of a ribosome- inactivating and ahortifacient protein from seeds ofLuffa acutangula. Lit. J. Pept. Protein Res., 38，15-19. Yeung, H.W., Li, W.W., Feng, Z., Barbieri, L. and Stirpe, F. (1988). Trichosanthin, a-momorcharin and P-momorcharin: identity of abortifacient and ribosome-inactivating proteins. M. J. Peptide Protein Res., 31, 265-268. Yeung, H.W., Ng, T.B., Li, W.W. and Cheung, W.K. (1987). Partial chemical characterization of alpha- and beta-momorcharins. Planta Med., 1987，164- 166. Yeung, H.W., Li, W.W., Chan, W.Y. and Ng, T.B. (1986). Alpha and Beta momorcharins. Lit. J. Peptide Protein Res.’ 28, 518-524. Yoshinari, S., Yokota, S., Sawamoto, H., Koresawa, S., Tamura, M. and Endo, Y. (1996). Purification, characterization and subcellular localization of a type-l ribosome-inactivating protein from the sarcocarp of Cucurbita pepo. Eur. J. Biochem.，242, 585-591. Zheng, P., Lu, B.Y., Liu, W.Y. and Jin, S.W. (1996). Characterization ofthe enzymatic mechanism of y-momorcharin, a novel ribosome-inactivating protein with lower molecular weight of 11,500 purified from the seeds of bitter gourd (Momordica charantia), Biochem. Biophy. Res., 229, 287-294.
81 . «i
,
‘：…
* .,
：：11?-.¾.
.:.
‘；•
fi i'
.
. , ,.:..:-:-r
/
-
.11 f a w i i . i i i 一 - i f . . •—.— . $ t i m & ^
CUHK Libraries 1 _l__i i DD35fiTSE7 i