THE JOURNAL OF BIOLOGICAL CHEMISTRY (c) 1991 by The American Society for Biochemistry and Molecular Biology, Inc

Vol. 266, No. 13, Issue of May 5,pp. 8426-8430,1991 Printed in U.S.A.

DNA Sequence ofMirabilis Antiviral Protein (MAP), a Ribosome-inactivating Protein with an Antiviral Property, from Mirabilis jalapa L. and Its Expression in Escherichia coli*
(Received for publication, December 17, 1990)

Jiro KataokaS, Noriyuki Habuka, MasahiroFuruno, Masashi Miyano, Yoichi Takanami, and Akira Koiwai
From Life Science Research Laboratory, Japan Tobacco Inc., 6-2 Umegaoka, Midori-ku, Yokohama, Kanagawa 227, Japan

We cloned a cDNA for Mirabilis antiviral protein (MAP), a ribosome-inactivating protein (RIP), which inhibits the mechanical transmission of plant virus and the in vitro protein synthesis of both prokaryotes and eukaryotes. The cDNA consisted of 1066 nucleotides and could encode 278 amino acids. The major part of the amino acid sequence (from Ala2*to Ser2) was identical with the sequence of native MAP as determined by protein sequencing. An NH2-terminal extrapeptide (28 amino acid residues) of MAP was comparable with the signal peptides of plant proteins accumulating in the vacuole. A stable hairpin structure was predicted in the 3noncoding region of the cDNA. Tandem repeated sequences were found downstream from the hairpin structure. They were composed of triple complete repeats of a heptanucleotide with preceding and following hexa-nucleotide repeats. The cDNA was expressed in Escherichia coli based on the T7 expression system. The product encoded by the cDNA was confirmed to beMAP precursor by Western blotting followed by immunological analysis. The growth of the transformants was inhibited by the expression of the gene. MAP precursor also seemed to inhibit the protein synthesis of E. coli just as native MAP has been observed to do.

Mirabilis antiviral protein (MAP) isolated from the roots of Mirabilis jalapa L. induces the plants systemic resistance against themechanical transmission of plant viruses, such as tobacco mosaic virus and cucumber green mottle mosaic virus (1-3). Such antiviral proteinsor glycoproteins have also been found in other higher plants, e.g. Phytolacca americana (4), Dianthus caryophyllus (5), Chenopodium sp. (6), and Boerhaavia diffusa(7). MAP has been purified to homogeneity and its complete amino acid sequence determined (8). It is composed of 250
* The costs of publication of this article were defrayed in partby the payment of page charges. This article must therefore be hereby marked aduertisernent in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The nucleotide sequence($ reported in thispaper been submitted has t o the GenBankTM/EMBL Data Bank with accession numbeds)

amino acids having a combined molecular weight of 27,833. The total synthetic gene of MAP has been synthesized and expressedin Escherichia coli using tac promoter (8). The growth of the transformant was inhibited by the expression of the gene and theyield of the productwas very low (3.6 pg/ liter). The amino acid sequence of MAP has 24% homology with that of ricin A-chain. Ricin inhibits the in vitro protein synthesis of theeukaryotesystem by inactivation of the ribosomes (9, 10) and is thus called a ribosome-inactivating protein (RIP)(11, 12). RIPs have been isolated from a variety of plants, e.g. ricin from Ricinus communis (13), tritin from Triticum aestiuum (14), andpokeweed antiviral protein from Phytolacca americana (15). Although the ribosomes of E. coli had been considered to be insensitive to the action of RIPs (16), MAP was found to inhibit the vitro protein synthesis in of the prokaryote system as as the eukaryote system well (16). These factssuggest that MAPalso inhibits thein vivo protein synthesis of E. coli. Recently, the nucleotide sequences of cDNAs for several have RIPs, e.g. ricin (17), trichosanthin (18), and saporin (19), been determined. By analysisof the sequences, it was shown that the proteins were translated as precursorforms, and the extrapeptides resembled a consensus secretory signal peptide (20). So, it has been suggested that the proteins are secreted and compartmentalized (18).In the case of ricin, it was also found that ricin A- and B-chains are generated during maturation by the cleavage of a single peptide. These things suggest that the analysis of acDNA for a RIP could reveal the maturation process of the RIPprecursor. (8), Since native MAP lacks the NHn-terminal methionine MAP is expected to be translated as a precursor form. T o clarify the maturation process of MAP precursor, we cloned cDNA of the MAP gene from M. jalapa L. In this paper, we report the nucleotidesequence of MAP precursor and its expression in E. coli. The function of the signal peptide of MAP precursor isalso discussed.
EXPERIMENTAL PROCEDURES

Materials

Restrictionendonucleasesand nucleicacidmodifyingenzymes were purchased from Boehringer Mannheim Yamanouchi (Tokyo, 090347. Japan), Pharmacia Japan (Tokyo, Japan), Bethesda Research Labo$ To whom correspondence should be addressed: Life Science Reratories (Tokyo, Japan), and Takara Shuzo (Kyoto, Japan). Nylon search Laboratory, Japan Tobacco Inc., 6-2 Umegaoka, Midori-ku, Yokohama, Kanagawa 227, Japan. Tel.: 045-972-5901; Fax: 045-972- membrane and reagents for oligonucleotide synthesis were from Japan Poul (Tokyo, Japan) and Applied Biosystems Japan (Tokyo, 6205. Japan), respectively. [cY-~P]~CTP from Du Pont-New England was The abbreviations used are: MAP, Mirabilis antiviralprotein; used RIP, ribosome-inactivating protein; IPTG, isopropyl-1-thio-p-D-ga- Nuclear (Tokyo, Japan). The other reagents in the experiments lactopyranoside; ELISA, enzyme-linked immunosorbent assay; ER, were from Wako Pure Chemical (Tokyo, Japan) and Kanto Science (Tokyo, Japan). endoplasmic reticulum.

8426

cDNA Sequence of MAP and Its Expression in E. coli

Bacteriophages, Plasmids, and Bacterial Strains
X cloning vectors, XZAP (21) and XgtlO (22), and plasmidvectors,

a427

pBluescript KS(-) and KS(+), were purchased from Stratagene (Tokyo, Japan). E. strains, XL1-Blue (recAl, endAl,gyrA96, coli lac-, thi, hsdR17, supE44, relAl, IF'proAB, ladQ,lacZAM15, TnlOJ) and C600hfl- (supE44, thi-1, leuB6, lacY1, tonA21, L-Irk+, mk+JmcrA-, mcrB', hfl-) were used for the construction of cDNA libraries and JM109 (recAl, endAl,gyrA96, thi, hsd supE44, relAl, X-, A(lacR17, proAB),IF',traD36,proAB, lacIQlacZAM15J) was used for the subcloning of cDNAs. The E. strain JM109(DE3)@ purchased coli was from Promega (Tokyo, Japan) and usedfor the expression of the cDNA. It is almost the same as JM109 except that it has a cloned gene of T 7 RNA polymerase under the control of lac promoter in its chromosome. It expresses aforeigngene under the control of T7 promoter with IPTG (23). Methods

Extraction of Poly(A)' RNA from M. jalapa L.-A suspension culture of M. jalapa L. was homogenized in extraction buffer (0.1 M Tris-HC1, pH 9.0, 10 mM EDTA, 0.1 M NaCl,and 1.0% sodium dodecyl sulfate)andcentrifuged (7,000 rpm for10 min at room temperature). The supernatantwas subjected to phenol/chloroform/ isoamylalcohol (25:24:1) extraction (7,000 rpm for 20 min a t room temperature) four times. All the nucleic acids were precipitated from the aqueous phase by the addition of 1/20 volume of 5 M NaCl and 2.5 volumes of ethanol (24). After centrifugation (7,000 rpm for 30 min at 4 "C), thepellet of nucleic acid was dissolved in sterile water. The solution was laid on 5.7 M CsCl solution, and all the RNA was recovered as a pellet after ultracentrifugation (28,000 rprn for 24 h at RESULTS AND DISCUSSION 16 "C) (25). Poly(A)+ RNA purified using an oligo(dT)-cellulose was column or oligo(dT)-Latex (26). Nucleotide Sequence of MAP Gene-The total synthetic Construction of cDNA Libraries and Screening of MAP GeneMAP gene was designed based on the codon usage of E. coli, cDNAs were synthesized according to the methods of Gubler and not on that of plants (8). However, the synthetic gene is Hoffman (27). A cDNA library was constructed (28) using XZAP@ expected to be at least 60% homologous with its messenger cloning vector (21). ThesyntheticMAP gene (8) wasused as a RNA as estimated on the basis the universal genetic code. of template DNA for the synthesis of a uniformly labeled DNA probe nucleotide sequence of using randomoligonucleotide primers (29). Plaque lifting and hybrid- In fact, in the present experiment, the ization were carried out according to the methods recommended by the syntheticgene was found to be 74% homologous with that the supplier of the nylon membrane. The probe was hybridized with of MAP850 on average. The longest identical region was 14 20 plaques from a total of approximately 150,000 plaques. By analyz- bases long (data not shown). The labeledprobefrom the ing the cDNAs of the 20 plaques, part of the nucleotide sequence of synthetic gene hybridized properly in thegene cloning. MAP gene was determined and thelongest cDNA was designated as Fig. lA shows the nucleotidesequence of CM7 and the MAP850 (Fig. 1A). cDNAs were newly synthesized by priming a synthetic oligonucleotide; the position of the primer was 208 bases deduced amino acid sequence of the long open readingframe. downstream from the 5"terminal of MAP850 (Fig. 1A). A second A stop codon (TAA) was found at the 15th base upstream cDNA library was constructed, and MAPN-15 was screened out by from the first start codon of the open reading frame, and the the same procedures used for MAP850 except that a labeled DNA major part of amino acid sequence (AlaZ8-Serz7')is identical probe was synthesized usingMAP850 DNA as a template. A plausible initiation codon could not be found in the sequence of MAPN-15. with that of native MAP, which was determined by protein Since Moloney murine leukemia virus reverse transcriptase could sequencing (8).These data lead to the conclusions that MAP synthesize a longer first strand than avian myeloblastosis virus reprecursor is encoded by the open reading frame, and CM7 is verse transcriptase according totheinstructions of thesupplier a nearly full-length cDNA clone of MAP gene. The sequence (Bethesda Research Laboratories), cDNAs were synthesized using also revealed the presence of an extrapeptide (28 residues) at Moloney murine leukemia virus reverse transcriptase insteadof avian of myeloblastosis virus reverse transcriptase. Thirty plaques from the NH2-terminal, and most the amino acid residues comhydrophobic (Fig. 2 A ) . This isa feature among approximately 150,000 plaques were screened outby the same posing the peptide are procedures used for MAPN-15 except thatg t l O (22) was usedinstead common to the signal peptides of both eukaryotes and proX of XZAPO. By analyzing each cDNA of the 30 plaques, the longest karyotes. Although no consensus amino acidsequence has cDNAs (CM7 and CM31) found and then were subcloned in a plasmid been reported among the signal peptides, several signal pepvector (pBluescript) at the EcoRI site, generating pCM7 and pCM31, tides of plants have sequences comparable with that MAP, of respectively. Determinationand Analysis of Nucleotide Sequence for MAP for example sporamin, the storage protein of sweet potato, vacuole (33), tomato protease inhibGene-The nucleotide sequences of the four cDNAs were determined which is deposited in the by the dideoxy chain termination method (30). Overlapping deletion itor, which is also transported to the vacuole (34), and ricin, clones were constructed for each cDNA by the combination of exo- the seed proteinof R. communis, which is transported to the nuclease 111 andmung beannuclease (31).TwoDNAanalyzing protein body, one mode of vacuoles (35). Since about amino 20 programs, Microgenie" (Beckman, version 6.0) and DNASIS@ (Hiacidresidues of signal peptide from the NH,-terminal are tachi, version6.0), were usedfor theanalysis of the nucleotide enough to import the following protein into theendoplasmic sequence. Expression and Detection of MAP Precursor in E. coli-The E. coli reticulum (ER) lumen (33), the first 20 amino acid residues strainJM109(DE3)@ (23) was transformed by pCM7,pCM31, or of the signal peptides are aligned(Fig. 2B). Twodistinct pBluescriptas a control,andthe respective transformants were regions, A and B, were observed. Region A is variable except designated DE3/pCM7, DE3/pCM31, and DE3/pBs. o examine the T effect of IPTG on the transformants, each transformant was pre- for the leading lysine, the amino acid residue common to all signal peptides.Region B is conserved there aretwo adjacent cultured for 3 h in 5 ml of Luria broth medium with 50 pg/ml of ampicilin. Cells were harvested by centrifugation (5,000 rpm for 10 aromatic residues and another one (phenylalanine,tryptomin) and suspendedin 1.0 ml of the medium. Half of the suspension phan, tyrosine) and preceded by hydroxylated residue(s)

(0.5 ml) was transferred to5 ml of the medium, and the remainder to a medium containing 0.5 mM of IPTG. The growth of each transformant was monitored by its absorbance a t 550 nm (Fig. 4). The gene product was purified using DE3/pCM7. An overnight culture of DE3/pCM7 was used as a seed according to the methods recommended by the supplier of the E. coli strain. But, the culture showed an abnormal growthprofile and a long lag phase (more than 1 2 h) in the main culture (data not shown). Then, the main culture was carried out in sucha way as to maintain the logarithmicgrowth of the transformant. DE3/pCM7 was pre-cultured in 10 ml of Luria broth medium with 50 pg/ml of ampicilin and 2% (w/v) glucose at 37 "C for 3 h. The pre-culture was incubated in 100 ml of the same medium and finally in 1,000 ml a t intervals of 3 h. IPTG was added to a final concentration of 0.5 mM and the culturewas incubated for an additional 3 h. Cells were harvested by centrifugation and lysed by sonication. Cell debris wasremovedby ultracentrifugation at 30,000 rpm for 2 h. The precipitate of 30-90% ammonium sulfate saturation was dissolved in the buffer (10 mM sodium phosphate, pH 8.0) and dialyzed against the same buffer. The solution was applied on a column of Mono S. After washing the column with the buffer, proteins were eluted with a 25-ml liner gradient of sodium chloride (0-0.5 M ) in the buffer. The eluate was monitored by its absorbance a t 280 nm, and the presence of MAP precursor was measured for each fraction (1 ml) by enzyme-linked immunoabsorbent assay (ELISA) using the antibody against MAP Each fraction contain(8). ing MAP precursor was analyzed with sodium dodecyl sulfate-polyacrylamide gel electrophoresis(32),andthe gel was subjected to Western blotting followed by immunoblotting analysis (8).

cDNA Sequence of MAP and Its Expression in E. coli

A
50
110 170
1
IICGI\GCTTTTCTTCTCTC==~G~~~G~~~==~C=~RR~AC~*~~RRG~C 19

G A C
0

*-TGCTTI\CRRCTACGIGGTGTTTTTTCTCCTATTGACGA~TG~ATCACTTGG~*CGCG

109

A T T G T G R R T C C A C R R T C R R G G G C A G C ~ C C G A C A C T C G ~ C R R T ~ G C A T C ~ ~ ~ 169 C ~ ~ G~C I V N P Q S R A A P T L E T I A S L O L

C
A:U

u
U:A A:U
U:A

R R T I ~ C C C I I C C R C A T I \ T C T A T C * ~ ~ = ~ T 729 I C C N N P T T Y L S F I T N I R T K V A D K
A C C G A G C I T G T A C A A T T C A G W T R R G C W C A T T T A C T C ~ G A T ~ T T C A T A C A T C 289 T E Q C T I Q X I S R T F T Q R Y S Y I
149
409
469

230

290
350 410
470

529
iaq

530 590

-!&
619
709

G:C G:C G:C aA J G V P W F W F W - - N w

-Poly (A)

650
110

69

FIG. 3. Secondary structure of the 3noncoding region of MAP gene. Deduced polyadenylation signals are underlined. Arrows indicate each unit of tandem repeats and theirdirection.

770

829
889

plaques were found in a totalof 150,000plaques, MAP mRNA must share only 0.015-0.02% of the total mRNA. However, a 890 919 substantial amount of MAP was recovered from the suspen950 :oc9 sion culture. It seems that the long lifetime of MAP mRNA 1010 enabled it to be translated many times due to the action of B the repetitive extragenic palindromic-like sequence. Tandem repeated sequences were found downstream of the hairpin structure.The repeats were composedof triple perfect T7 promoter MAP gene lac promoter repeats of hepta-nucleotides (5AAUGUGU3)with preceding FIG. 1. A , nucleotide sequence of MAP gene (CM7) and its de- and following repeats of hexa-nucleotides, 5AUGUGU3 and duced amino acid sequence. The position of primer used for the 5AAUGUU3, respectively. Tandem repeated sequences have synthesis of MAPN-15 is underlined. The asterisk, solid circle, and been reported in other plant genes, such as sporamin in sweet plus sign indicate the 5terminal of MAP850, MAPN-15,and CM31, potato (38) and zein in Zea mays (39), and a complementary respectively. E , schematic representation of pCM7 and pCM31. Bold arrows indicate the transcriptional direction of the promoters. The repeated sequence was also found in each 5-noncoding region. thin line and shaded box indicate the coding regions of MAP precursor It has been suggested that a plausible base pairing between direct repeats inthe 5- and3noncoding regions affects the and thesignal peptide, respectively. translational efficiency (39) or the stability of the mRNA (38). These thingsmight suggest the presence of the complePmscuiq site mentary repeated sequence upstream of the 5terminal of MAP gene. Expression and Detection of MAP Precursor-Eight independent clones of pCM7 and ten of pCM31 were analyzed and each cDNA was inserted in the direction inverted to the N H Z - ~ T T T ~ F L L L T T I ~ I ~ A I V N- Q S P lacZ promoter (Fig. 1B). It is known that the regulation of B Rg n A ei o mion B lacZ promoter is relatively leaky (8),and thetranslation of a chimeric mRNA could start from the initiation codon of the MAP cDNA (40, 41). A transformant with a recombinant plasmid, in which each cDNA was inserted in the right direction for SPO-B the lacZ promoter, could express MAP precursor, which inhibits the growth of the host through its toxicity. Then, an attempt was made to express the gene product in the E. coli strainJM109(DE3)@(23) using the T7 promoter present upstream from the 5-terminal of MAP gene on the recomFIG. 2. A , amino acid sequence of the preceding peptide of MAP precursor. The hydrophobicity index was calculated according to binant plasmids (Fig. 1B). The expression of a foreign gene Chou and Fasman (49). B , alignment of amino acid sequences of in a T7 expression system is expected to be controlled more signal peptides. SPOE and TPZ mean sporamin B and tomato pro- stringently than in a lacZ promoter system by the two regutease inhibitor, respectively. latory systems. (i) The gene of T7 polymerase is controlled under lac UV5 promoter. (ii) No RNA polymerase of the host (threonine or serine). The similarities of the sequences suggest recognizes the T7promoter sequence (42). Fig. 4 shows the effect of IPTG on the growth of DE3/ that MAP is also imported into ERlumen and transported to pCM7, DE3/pCM31, and DE3/pBs. No difference was obthe vacuole like the other proteins. In the 3noncoding region of the cDNA, a stable hairpin served between the growth profiles of DE3/pBs in the presstructure was predicted by computer analysis (Fig. 3). The ence and absence of IPTG. DE3/pCM7 and DE3/pCM31 in hairpin consists of a 13-base pair stem separated by an 11- the absence of IPTG grew in a way similar to DE3/pBs, but base loop (AG = -14.6 kcal/mol, -613.2 kJ/mol). A hairpin the growth of the transformantsin the presence of IPTG was structure in the 3noncoding region is known to be a termi- distinct from that of DE3/pBs. The growth of the transformnator for the precise termination of transcription in prokar- ants stopped within 1.5 h after the addition of IPTG. Since yote genes (36) and a repetitive extragenic palindromic se- inhibition of the growth of E. coli was also observed when the quence for the stability of mRNA (37). Since 20-30 positive synthetic MAP gene was expressed in the organism (81, the
830

Id cDNA Sequence of M A P ar; Its Expression in E. coli

8429

smaller(approximately 2,300) thanthe detlucedprecursor withthesignalpeptide.Thediscrepancy in themolecular weights might arise because a part of the signal peptide of M A P is cleaved off in E . coli as has been observed to happen in sporamin, in which the first 20 amino acid residues ol' the signal peptide are cleaved in E , coli ( 3 3 ) .T h e region is also off processed in a n in vitro translation system in the presence of dog microsomal membrane and is thought to he cleaved off during importation into the ER lumen ( 3 3 ) . The similarity between the signal peptidesof MAP precursor and sporamin ".. , seems to support the suggestion that MAP is imported into I 8 0 100 200 300 as t h e E R l u m e n described above. Duration of culture (min.) Ricin cleaves a n N-glycosidic bond at A"" in rat liver 2% S h e are FIG. 4. Growth curvesof transformants.+ and - indicate t h e rRNA (11). A"""' andthecorrespondingadenine located in a highly conserved region of 'LR S-like rRNAs o f presence or ahsence of I P ' I K , respertivelv. various origins (43). 23 S rRNA (28 S-like rRNA of I-,'. coli) in intact ribosomes resistant to ricin, and only deproteinized is 23 S rRNA is cleaved by a high concentration of ricin A-chain (44). Therefore, we suggested a prevention mechanism for the ribosomes by which r-protein(s) of E . coli protect the rRNA fromanattack by ricin (16). Some Rll's, c . , ~pokeweed . antiviral protein (45) and tritin (46), have been reported not to inhibit thein vitro protein synthesis of the plants t o which R I P S could inactivate not theyarenative,thatis,these ribosomes protected by the r-proteins their own plants. of On the other hand, the ribosomes of H. communis arc j u s t n I0 20 as susceptible to a high concentration of ricin as rihosomes Number of fraction from different species 4 7 ) . Ricin is produced in large amounts ( FTG.5. Absorbance of ELISA for each fraction eluted from in the seed of I<. communis (more than 0.Sr; of thegross Mono S column. The solid circlrs and the thin linr indirate the weight) ( 3 5 ) . Therefore,anothermechanismmustexist to ahsorhance o f I51,ISA and the concentration of sodium chloride, protect the host's ribosomes from ricin. A P is a unique 1111' M respectively. which inhibits the uitrn protein synthesisof eukaryotes and in prokaryotes (16). Since MAP seems to inactivate the rib1 2 3 somes of eukaryotes and prokaryotes in spite the presence of of r-proteins, MAP might inactivate the ribosomeso f its own 43plant, which produces a large amount (16.Y; of soluhle protein) of M A P in t h e root. Thus, another mechanism is also required to protect the host's ribosomes fromM A P . I t is worthy of note that all 1111's. including MAP, have a 30signal peptide and seemto be transported to the vacuole ( 18, 47). Proteins in the vacuole are known to he translated at ribosomes binding to t h e ER and to be imported into the Eli lumen co-translationally (48). According to this mechanism, 20t h e host.'s ribosomes would he kept fromRIP produced hy the ER membrane and could synthesize a large amount of R I P by compartmentalization of the RIP. This mechanism may FIG. 6. Western hlotting analysis M A P precursor. / , a m 1, be a strategy common to host plants producing an R I P for of indicates native MAI' from the Mirnhilis plant; lane 2 indicates MAP minimizing the risk of ribosomes being inactivated by their precursor produced in I)F:/CM7: and lnnr 3 is for a mixture of lanrs own RIP. Thisdoesnotexcludethepossibilitythatthe I and 2. ribosomes of M. jalapa L. are insensitive toM A P . In order t o verify the proposed mechanism of a host's ribosomes being transformants are thought to produce the protein. A single protected from RIP by compartmentalization, it is important band was observed only in the lanes of the crude extractsof to study theeffect of MAP on the ribosomesof its host plant, DE3/pCM7 and DE3/pCM31 cultured with IPTG when the M. jalapa L. primary analysisof Western blotting was performed (data not shown). The transformants were confirmed to produce the protein. The protein was purified from the culture of DEB/pCM7 as described under "Experimental Procedures." The gene product eluted from t.he column of Mono S was analyzed by ELISA using the antibody against MAP (Fig. 5). T h e t,otal recovery 2.5 was estimated to be pg/liter. The product was eluted with 170 mM NaCI. Western blotting analysis showed that the molecular weight of the protein was slightly larger (approximately 1,000) t.han t,hat of native MAP from the Mirabilis plant (Fig.6). Since the molecular weight the signal peptide of was calculated to be 3315, the product seemed to be slightly
1

8430

cDNA Sequence of M A P alnd Its Expression in E. coli

28. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) in Molecular Cloning:ALaboratory Manual, 2nd Ed.,pp. 8.73-8.75, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY 29. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) in Molecular Cloning: A Laboratory Manual, 2nd Ed, p. 10.13, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY 30. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Sci. U. S. A. 7 4 , 5463-5467 31. Yaniscb-Perron, C . , Vieira, J., and Messing, J. (1985) Gene (Amst.) 3 3 , 103-119 32. Laemmli, U. K. (1970) Nature 2 2 7 , 680-685 33. Hattori, Ichihara, T., S., and Nakamura, K. (1987) Eur. J. Biochem. 1 6 6 , 533-538 34. Graham, J. S., Pearce, G., Merryweather, J., Titani, K., Ericsson, L., and Ryan, C. A. (1985) J.Biol. Chem. 260,6555-6560 35. Jansen, F. K., Bourrie, B., Casellas, P., Dussossoy, D., Gros, O., Vic, P., Vidal, H., and Gros, P. (1988) in Immunotoxins (Frankel, A. E., ed) pp. 97-122, Kluwer Academic Publishers, Boston 36. Lewin, B. (1985) in Genes, 2nd Ed, pp. 206-211, John Wiley & Sons, Inc., New York 37. Newbury, S. F., Smith, N. H., Robinson, E. C., Hiles, I. D., and Higgins, C. F. (1987) Cell 4 8 , 297-310 38. Murakami, S., Hattori, T., and Nakamura, K. (1986) Plant Mol. Biol. 7,343-355 39. Spena, A., Krause, E., and Dobberstein, B. (1985) E M B O J. 4 , 2153-2158 40. Hattori, T., Nakagawa, T., Maeshima, M., Nakamura, K., and Asahi, T. (1985) Plant Mol. Biol. 5,313-320 41. Nakamura, K., Hattori, T., and Asahi, T. (1986) FEBS Lett. 1 9 8 , 16-20 42. Chamberlin, M., Mcgrath, J., and Waskell, L. (1970) Nature 228,227-231 43. Chan, Y.-L., Endo, Y., and Wool, I. G. (1983) J. Biol. Chem. 2 5 8 , 12768-12770 44. Endo, Y., and Tsurugi, K.(1988) J. Biol. Chem. 263,8736-8739 45. Owens, R. A., Bruening, G., and Shepherd, R. J. (1973) Virology 56,390-393 46. Coleman, W. H., and Roberts, W. K. (1981) Biochim. Biophys. Acta 654,57-66 47. Harley, S. M., and Beevers, H. (1982) Proc. Natl. Acud. Sci. U. S. A. 79,5935-5938 48. Alberts, B., Bray, D., Lewis,J., Raff, M., Roberts, K., and Watson, J. D. (1989) in Molecular Biology of T h e Cell (Robertson, M., ed) 2nd Ed., pp. 433-451, Garland Publishing Inc., New York 49. Rose, G. D., and Roy, S. (1980) Proc. Natl. Acad. Sci. U. S. A. 77,4643-4647

4. Wyatt, S. D., and Shepherd, R. J. (1969) Phytopathology 5 9 , 1787-1794 5. Ragetli, H.W. J., and Weintraub, M. (1962) Virology 1 8 , 241248 6. Smookler, M. M. (1971) Ann. Appl. Biol. 6 9 , 157-168 7. Verma, H. N., Awasthi, L. P., and Saxena, K. C. (1979) Can. J. Bot. 5 7 , 1214-1217 8. Habuka, N., Murakami, Y., Noma, M., Kudo, T., and Horikoshi, K. (1989) J. Biol. Chem. 264,6629-6637 9. Endo, Y., Mitsui, K., Motizuki, M., and Tsurugi, K. (1987) J. Biol. Chem. 262,5908-5912 10. Endo, Y., and Tsurugi, K. (1987) J . Biol. Chem. 262,8128-8130 11. Endo, Y. (1988) in Zmmunotoxins (Frankel, A. E., ed) pp. 75-89, Kluwer Academic Publishers, Boston 12. Barbieri, L., and Stipe, F. (1982) Cancer Suru. 1, 489-520 13. Olsnes, S., and Pihl, A. (1982) in T h e Molecular Action of Toxins and Viruses (Cohen, P., and van Heyningen, S., eds) pp. 51105, Elsevier Science Publishing Co., Inc., New York 14. Roberts, W. K., and Stewart,T. S. (1979) Biochemistry 18,26152621 15. Irvin, J . D. (1975) Arch. Biochem. Biophys. 1 6 9 , 522-528 16. Habuka, N., Akiyama, K., Tsuge, H., Miyano, M., Matsumoto, T., and Noma, M. (1990) J. Biol. Chem. 265,10988-10992 17. Lamb, F. I., Roberts, L. M., and Lord, J. M. (1985) Eur. J. Biochem. 1 4 8 , 265-270 18. Chow, T. P., Feldman, R. A., Lovett, M., and Piatak, M. (1990) J. Biol. Chem. 265,8670-8674 19. Benatti, L.,Saccardo, M.B., Dani, M., Nitti, G., Sassano, M., Lorenzetti, R., Lappi, D. A., and Soria, M. (1989) Eur. J. Biochem. 1 8 3 , 465-470 20. von Heijne, G. (1981) Eur. J . Biochem. 1 1 6 , 419-422 21. Short, J. M., Fernandez, J. M., Sorge, J . A,, and Huse, W. D. (1988) Nucleic Acids Res. 1 6 , 7583-7600 22. Huynh, T. V., Young, R. A,, and Davis, R. W. (1985) in D N A Cloning: A Practical Approach (Glover, D. M., ed) Vol. I, p. 49, IRL Press, McLean, VA 23. Studier, F. W., and Moffatt, B. A. (1986) J. Mol. Biol. 1 8 9 , 113130 24. Hattori, T., Iwasaki, Y., Sakajo, S., and Asahi, T. (1983) Biochem. Biophys. Res. Commun. 1 1 3 , 235-240 25. Glisin, V., Crkvenjakov, R., and Byus, C. (1974) Biochemistry 13,2633-2637 26. Wolf, S. F., Haines, L., Fisch, J., Kremsky, J. N., Dougherty, J. P., and Jacobs, K. (1987) Nucleic Acids Res. 1 5 , 2911-2926 27. Gubler, U., and Hoffman, B. J. (1983) Gene (Amst.)25,263-269