CA2050593A1 - Recombinant trichosanthin and coding sequence - Google Patents

Abstract

ABSTRACT
Disclosed are the entire protein and nucleic acid coding sequences for unprocessed and mature trichosanthin protein from Trichosanthes kirilowii. A recombinant tri-chosanthin protein produced from the coding sequence, and the trichosanthin protein with amino-terminal and/or carboxy-terminal extensions are also described. Primers derived from the coding sequence are disclosed for use in obtaining the coding sequences of ribosome-inactivating-proteins which have regions of amino acid sequence identical to those of trichosanthin. Further, a multi-gene family of ribosome-inactivating-protein encoding genes of Trichosanthes kirilowii is also described.

Description

.3 RECOMBINANT TRICHOSANTHIN AND CODING SEQUENCE
This application is a continuation-in-part of co-- 5 owned, co-pending U.S. Patent Application No. 333,184.

1. Field of the Invention The present invention relates to recombinantly produced trichosanthin and DNA coding sequences there-fore.

2. Referenc~s Asano, K., et al., Carlsberg Res Commun, 51:129 (1986).
Barbieri, L., et al., Riochem J, 203:55 (1982).
Bullock, W.O., et al., Biotechniquesl 5(4):376 (1987).
Calderwood, S. B., et al., Proc Nat Acad Sci USA, 8~:4364 (1987).
Casellas, P., et al., Eur. J. Biochem. 176:581 (1988).
Chaudhary, V.K., et al., Nature, 335:369 (1988).
Coleman, W.H., et al., Biochem Biophys Acta, 696:239 (1982).
Crowe, S., et al., Aids Research and Human Retroviruses, 3(2):135 (lg87).
Cumber, J.A., et al., Methods in Enzymology, 112:207 (1985).
Duncan, R.J.S., et al., Anal Biochem, 182:68 (1983).
Falasca, A., et al., Biochem J, 207:505 (1982) .

- .

~ ~ r ~

Funatsu, G., et al., Agric Biol Chem, 52(4):1095 (1988).
Gasperi-Campani, et al., FEBS Lett, 76(2):173 (1977).
- 5 Grasso, S., et al., Phytopathology, 68:199 (1978).
Gu, Zi-wei, et al., Acta Chemica Sinica, 43:943 (1984).
Habuka, N., et al., J. Biol. Chem. 264:6629 (1989).
Halling, K.C., et al., Nuc Acids Res, 13:8019 (1985).
Hsu, K.J., et al., Acta Zool Sin, 22:149 (1976).
Hwang, Y.N., Chinese J Integrated Trad and Western Medicine, 7:154 (1987).
Irvin, J.D., Arch Biochem Biophys, 169:522 (1975).
Kao, H., et al., Acta Biol Exp Sin, 11:253 (1978).
Kuo-Fen, C., et al., Obs and Gyn, 59(4):494 (1982).
Lamb, F.I., et al., Eur J Biochem, 148:265 (1985).
Law, L.K., et al., J Reprod Fert, 69:597 (1983).
Lifson, J.D., et al., Science, 232:1123 ~1986).
Lin, J.Y., et al., Toxicon, 16:Ç53 (1978).
Maddon, P.J., et al., Cell, 42:93 (1985).
Maraganore, J.M., et al., J Biol Chem, 262(24):11628 (1987).
McGrath, M. et al. (1989) Proc. Natl. Acad. Sci.
USA 86:2844.

.. . .

.
. .
;

.

3 ~ ~ 3 ~ ~

Murray, H.G. et al., Nuc Acids Res, 8:4321 (1980).
Ohtsuka, E., et al., ~ Biol Chem, 260(5):2605 (1985).
Olsnes, S., Nature, 328:474 (1987).
Olsnes, S., et al., in Molecular Action of Toxins and Viruses, (Elsevier, 1982), Chapter 3.
Pan, ~., et al., Scientia Sinica (Series B) 30(4):386 (1987).
Spreafico, F., et al., Int J Immunopharmoc, 6(4):335 (1983).
Takahashi, Y., et al., Proc Nat Acad Sci, USA, 82:1931 (1985).
Taylor, B. et al, BRL Focus, 4(3):4 (1982).
Till, M.A., et al., Science, 242:1166 (1987).
Wang, Yu, et al., Pure & Appl Chem, 58~5):789 (1986).
Xiong, Y.Z., et al., Acta Zool Sin, 11:236 (1976).
Xuejan, Z., et al., Nature, 321:477 (1986).
Yeung, H. W. et al., Int J Peptide Protein Res, 27:325 (1986).

3. Background of-the Invention Trichosanthin (TCS) is a plant protein which is obtained from the Trichosanthes kirilowil root tuber.
The protein, which is also known as alpha-trichosan-thin (Law) and Radix trichosanthis (Kuo-Fen), is a basic, single-chain protein having a molecular weight of about 25,000 daltons. An incorrect protein se-quence of TCS has been reported (Gu; Wang), and a molecular model has been derived from X-ray analysis . .

.. ~

(Pan).
It has been shown that TCS is a potent inhibitor of protein snythesis in a cell-free lysate system (Maraganore). This activity is consistent with the observed homology in amino acid sequence between TCS
and the A chain of ricin, a ribosome-inactivating protein (RIP) which shows amino acid homology with a number of other RIPs, including abrin A chain (Olnes, 1982, 1987) and modeccin (Olsnes, 1982), and various single-chain ribosome-inactivating proteins, such as pokeweed anti-viral protein (PAP) (Irvin), RIPs from a variety of other plants (Coleman; Grasso; Gasperi-Campani) and the A subunit of Shiga-like toxins from E. Coli (Calderwood).
TCS, or plant extracts containing TCS, have been used in China as an abortifacient agent in humans, particularly during midtrimester (14 to 26 weeks). As such, the drug has been administered by intramuscular, intravenous, or intraamniotic routes, typically at a single dose of between about 5-12 mg. The phenomenon of mid-term abortion has been attributed to the selective destruction of placental villi. Other studies indicate that the syncytiotrophoblast is pre-ferentially affected (Hsu; Kao) and that secretion of hCG may be impaired (Xiong). TCS has also been shown to have a suppressive e~fect on human choriocarcinoma, and the protein appears to be able to pass the blood/brain barxier (Hwang).
It has recently been shown that TCS has a selective inhibitory effect on viral expression in human T cells and macrophages infected with human immunodeficiency virus (HIV). This is evidenced by , 2 ~ '.1 3 nearly complete inhibition of HIV-derived antigen in infected cells treated with the protein, as well as selective inhibition of protein and DNA synthesis in the infected cells. Similar results were also - 5 discovered for momorcharin, a basic ~lycoprotein obtained from the seeds of the bitter melon plant (Falosia; Spreafico; Lin; Earbieri~. These findings, and applications of the two proteins for the treatment of HIV infection, are detailed in U.S. Patent No.

4,795,739 for "Method of Selectively Inhibiting HIV".
Particularly in view of the ability of TCS to inhibit viral expression in HIV-infected human T cells and macrophages, it would be desirable to produce a relatively pure, invariant preparation of TCS, for use as a human therapeutic agent. Methods of preparing TCS from the roots of T. kirilowi1 have been repprted (Yueng). Analysis of the purified TCS produced by earlier-disclosed known methods indicates that the protein is only partially purified, and in particular, contains hemagglutinating contaminant protein(s). A
more recent purification method described in co-owned patent application for "Purified Trichosanthin and Method of Purification", U. S. Application No.
07/333,181, filed 4 April 1989, yields a highly purified TCS preparation which is substantially free of protein contaminants, including hemagglutinating proteins.
Additionally, it would be desirable to produce TCS by means of recombinant DNA technology. Synthesis of the protein by recombinant methods would avoid the difficulty of obtaining T. kirilowii roots in fresh form, since at present the tuber roots are available '. ';' ' ' :

,' ~ 3~ ~3 only from certain regions of the Orient. Recombinant production of TCS would also avoid the problem of variations in primary amino acid sequence in TCS
obtained from natural root material from different - 5 geographic areas.
Recombinant production of TCS would also facili-tate the production of peptide derivatives of TCS, in-cluding bioactive peptide portions of TCS, and bioac-tive portions of the protein fused with functional peptides which confer, for example, enhanced target-cell specificity.

4. Summary of_the Invention It is therefore one object of the present inven-tion to provide a recombinant TCS protein capable of selectively inhibiting viral expression in HIV-infec-ted human T cells or macrophages.
It is a related object of the invention to pro-vide the coding sequence for TCS from T. kirilowii.
Still another object of the invention is to provide sets of degenerate primers corresponding to spaced amino acid regions of TCS which are homologous to spaced amino acid regions of RIPs, for use in selectively amplifying plant-derived genomic sequen-ces which code for such RIPs.
In one aspect, the invention includes a cloned nucleic acid molecule which encodes a trichosanthin protein having the functional properties of TrichosantheS-Obtained trichosanthin. The nucleic acid molecule is included in the sequence:

- : ` ~ ' ' :

`
, ' ' ' EcoRl 1 GAATTCAAATATTTTCTGAATAAATATAAAATTCATI'GTAGAGAAATGATGAGAA 55 111 GCCCGAATTTATTATCA~AAGCGA~AGTTAATAATATCT~AAAAAAAACTATT 163 164 AccTTATAAGAAGcTATTAccTAGATGGcATAAGATcATAcTTTTATTTTTGATT 218 274 ATACCTCTCTATA~AAACCACAGCTTGAGATGCTCCAATGGCATCCAAATTCCTC 328 499 CTG TAC GAT ATC CCT CTG TTA CGT TCC TCT CTT CCA GGT TCT 5~0 lOg5 GTT GTG CTT ATA AAT GCT CAA AAC CAA CGA GTC ATG ATA ACC 1086 where basepairs 409 to 114g encode the mature form of TCS
isolated from Trichosanthes kirilowii.
The nucleic acid of the invention may include:
(a) basepairs 409 to 1149 which encodes mature TCS from T. kirilowii,;

, (b) in addition to (a), basepairs 340-408, which en-codes an amino t~_rminal extension of the mature form of TCS from T. kirilowii;
(c) in addition to (a), basepairs 1150 to 1206 which encodes a carboxy terminal extension of the mature form of TCS from T. kirilowii; and (d) a TCS codin~ sequence joined with a ligand peptide coding sequence, encodin~ a fused protein having a ligand peptide which confers cell-surface recognitiOn properties on the fused protein.
The invention also includes the coding sequence for TCS
from T. kirilowii in combination with an expression vector. One preferred expression vector construction contains a promoter, a ribosome binding site, an ATG
start codon positioned adjacent the amino-terminal codon of TCS, and a stop codon positioned adjacent the carboxy terminal codon of mature TCS.
In another aspect, the invention includes a primer mixture for use in selectively amplifying a genomic fragment coding for first and second spaced regions of TCS from T. kirilowii DNA, by repeated primer-initiated strand extension. The primer mixture includes a first set of sense-strand degenerate primers, and a second set of anti-sense primers, where each set contains substantially all of the possible coding sequences corresponding to the first and second region of known trichosanthin amino acid sequence, respectively. That is, each degenerate primer set includes at least one primer species which is effective to hybridize with the coding sequence of the corresponding amino acid region.
In a preferred embodiment, the primers in the first and second primer sets are designed to hybridize to first and second coding regions, respectively, which encode TCS
amino acid sequences that are ~omologous in amino acid , : :
:: :

, 3~

sequences to first and second amino-acid sequences in a variety of RIPs, such as ric:in A chain, abrin A chain, pokeweed antiviral protein, and barley ribosome inhibitor. The two primer sets may be used to obtain genomic co~ing sequences for the corresponding RIPs, by repeated primer-initiated strand extension.
Also forming a part of the invention is a recombinant trichosanthin protein having the functional properties of mature trichosanthin (a) derived from T. kirilowii and (b) having the sequence:

Asp Val Ser Phe Arg Leu Ser Gly Ala Thr Ser Ser Ser Tyr Gly Val Phe Ile Ser Asn Leu Arg Lys Ala Leu Pro Asn Glu Arg Lys Leu Tyr Asp Ile Pro Leu Leu Arg Ser Ser Leu Pro Gly Ser Gln Arg Tyr Ala Leu Ile His Leu Thr Asn Tyr Ala Asp Glu Thr Ile Ser Val Ala Ile Asp Val Thr Asn Val Tyr Ile Met Gly Tyr Arg Ala Gly Asp Thr Ser Tyr Phe Phe Asn Glu Ala Ser Ala Thr Glu Ala Ala Lys Tyr Val Phe Lys Asp Ala Met Arg Lys Val Thr Leu Pro Tyr Ser Gly Asn Tyr Glu Arg Leu Gln Thr Ala Ala Gly Lys Ile Arg Glu Asn Ile Pro Leu Gly Leu Pro Ala Leu Asp Ser Ala Ile Thr Thr Leu Phe Tyr Tyr Asn Ala Asn Ser Ala Ala Ser Ala Leu Mat Val Leu Ile Gln Ser Thr Ser Glu Ala Ala Arg Tyr Lys Phe Ile Glu Gln Gln Ile Gly Lys Arg Val Asp Lys Thr Phe Leu Pro Ser Leu Ala Ile Ile Ser Leu Glu Asn Ser Trp Ser Ala Leu Ser Lys Gln Ile Gln Ile Ala Ser Thr Asn Asn Gly Gln Phe Glu Thr Pro Val Val Leu Ile Asn Ala Gln Asn Gln Arg Val Met Ile Thr A n Val Asp Ala Gly Val Val Thr Ser Asn Ile Ala Leu Leu Leu Asn Arg Asn Asn Met Ala The recombinant TCS protein may further include an amino-termin~l extension having the sequence:

Met Ile Arg Phe Leu Val Leu Ser Leu Leu Ile Leu Thr Leu Phe Leu Thr Thr Pro Ala Val Glu Gly;

and/or a carboxy-terminal extension having the sequence:

~..,.,~

.: ,, ..

- :

Ala Met ASp Asp Asp Val Pro Met Thr Gln Ser Phe Gly Cys Gly Ser Tyr Ala Ile.

The in~ention further includes a recombinant process for the production of a trichosanthin protein having the functional properties of Trichosanthes-obtained TCS.
This recombin~nt process involves inserting a DNA
sequence encoding the TCS protein into an expression vector, transforming a suitable host with the vector, and isolating the recombinant protein expressed by the vector.
The invention further comprises nucleic acid and protein coding sequences for several unique members of the ribosome-inactivating-protein multi-gene family of TrichosantheS kirilowii.
These and other objects and features of the invention will become more fully apparent when the following detailed description of the invention is read in con~unction with the accompanying drawings.
Brief Description of the Drawings Figure 1 shows the amino acid sequence of mature TCS
isolated from T. kirilowii as determined herein (upper line) and as reported previously (lower line);
Figure 2 illustrates the steps in the method used to obtain cloned TCS coding sequences;
FigureS 3A and 3B show the DNA sequence from an amplified yenomic fragment containing a portion of the TCS coding sequence, and the corresponding amino acid . 30 sequences in the three possible reading frames in both directions;
Figure 4 shows the nucleotide sequence of the TCS
coding region from T. kirilowii and adjacent 5'- and 3'-end sequences;

:

" ' ' - :

Figure 5 illustrates the steps in the method used to express mature TCS in a bacterial system;
Figure 6 shows plots of percent inhibition of HIV
antigen (p24) production as a function of culture - 5 concentration of plant-derived TCS (closed boxes) and rTCS (open boxes);
Figure 7 shows plots of percent inhibition of 3H-leucine incorporation into trichloroacetic acid precipitable protein as a function of concentration of plant derived TCS (closed boxes) and rTCS (open boxes) in a cell free rabbit reticulocyte lysate protein synthesizing system.
Figure 8 illustrates the steps .in a method for producing a fused TCS protein containing a CD4~ peptide moiety; and Figure 9 compares the amino acid sequence of TCS with those of exemplary RIPs.
Figure 10 shows a hydropathy index computation for the entire coding sequence of trichosanthin including: the amino-terminal extension, the sequence encoding the mature protein, and the carboxy-terminal extension.
Figure 11 illustrates the synthetic gene designed for expression of TCS. The bold restriction sites are those flanking the individual synthetic fragments as o~erlapping sticky ends; for example, KpnI-HindIII, HindIII-NsiI, etc.
Figure 12A shows the correct orientation of the polylinkers in the construction of pPS200; Figure 12B
the incorrect orientation.
Figure 13 outlines a generalized schematic diagram of the method used for the condensation of cloned synthetic gene fragments to form a complete synthet.ic gene.
Figure 14 shows a schematic of the steps taken to clone and condense the synthetic gene for ~-trichosanthin.

.

Figure 15 shows the ability of the ~-trichosanthin protein, synthesized from the synthetic gene, to inhibit ln vitro translation reactions, relative to mutant forms of the protein.
- 5 Figure 16 shows the nucleic acid sequence of the cloned insert of pQ30E and corresponding protein coding sequence.
Figure 17 shows the nucleic acid sequence of cloned insert of pQ24 and the corresponding protein coding sequence.
Figure 1~ shows the nucleic acid sequence of cloned insert of pQ2 and the corresponding protein coding sequence.
Figure 19 shows the nucleic acid sequence of cloned insert of pQ3 and the corresponding protein coding sequence. Figure 20 shows the nucleic acid sequence of cloned insert of pQ12 and the corresponding protein coding sequence.
Figure 21 shows an alignment of protein sequences corresponding to the nucleic acid sequences of cloned inserts of pQ2, pQ3, pQ12, and of pQ21D and pQ30E.
Figure 22 shows an alignment similar to Figure 21 where the protein coding sequence corresponding to the nucleic acid sequence of cloned insert 2 is used as a standard sequence and amino acid substitutions representing the other protein coding sequences are listed in vertical columns for each amino acid residue -- the hyphens act as space holdèrs to allow easier alignment and an asterick indicates a site where omission of an amino acid is possible.

Detailed Description o~ the Invention I. Definitions _ . ' ~ . .

, t ~ ~ J

The terms below have the following meanings as used herein:
A "trichosanthin protein" is a protein having at least about 90% amino acid sequence identity with alpha-- 5 trichosanthin obtained from ~r. kirilowwii.
A trichosanthin protein has the functional properties of Trichosanthes-obtained trichosanthin if it has (a) the ability to selectively inhibit expression of HIV-antigen in HIV-infected T-cells or monocyte/macrophages, and/or (b) protein-synthesis-inhibitory activity.

II. Producing Recombinant TC5 This section describes methods for obtaining a genomic region containing the coding sequence for TCS from T._ kirilowii, and for expressing mature TCS protein in a bacterial expression system.

A. TCS Amino Acid Sequence TCS was purified by a novel method which is detailed in co-owned patent application for "Purified Trichosanthin and Method of Purification~, U. S. Application No.
07/333,181, filed 4 April 1989, and outlined in Example 1. The protein was at least about 98% pure as judged by HPLC and gel electrophoresis analysis.
The primary amino acid sequence of the purified trichosanthin was determined under contract with the Protein Chemistry Services at Yale University School of Medicine. The sequence is shown in Figure 1 (upper line) along with the previously published sequence (lower line) of TCS (Gu; Wan~). Variations between the two sequences are noted by double underlining.
As seen from Figure 1, the present sequence differs substantially from the published sequence. Most significant, as compared to th~ published sequence, the - ~ .
:, .
. - .

-2~3 present TCS sequence lacks a block of 10 amino acids atposition number 6~ and contains an additional sequence of 21 amino acids at position number 222. The present sequence agrees closely with X-ray diffraction data on - 5 crystalized TCS, and resolves inconsistancies between X-ray diffraction data and the previously published TCS
sequence. The new sequenca, particularly including the 21-amino acid addition, also provides greater sequence homology with a number of RIPS, such as ricin A chain and abrin A chain (see below) than the earlier published sequence.

B. TCS Coding Sequence Fi~ure 2 outlines the steps described below for ob-taining the complete coding sequence of TCS from T.kirilowii. The actual procedure used is given in Example 2.
With reference to the figure, genomic DNA isolated from T. Xirilowii is mixed with at least two sets of degene-rate primers in a reaction mixture designed for carryingout selective amplification of a TCS coding sequence.
In preparing the sets of degenerate primers, two spaced amino acid regions of TCS were selected for coding sequence targeting. The two amino acid sequences which were selected are overlined in Figure 1 and relate to a 35-mer degenerate primer for the sequence denoted A and to a 32-mer degenerate primer for the sequence denoted B.
Each set of degenerate primers were designed such that at least one primer sequence is effective to hybridize with the DNA sequence coding for the corresponding amino acid sequence. Deoxyinosine nucleotides were incorporated in order to generate probes longer than 20 nucleotides of manageable complexity (Ohtsuka;
Takahashi). One of the two prlmer sets is designed for , .

; J~ 3 hybridization with the anti-sense strand of one coding region, and the other primer set, for hybridization with the sense strand of the second coding region.
The primer set corresponding to the 35-mer includes 128 - 5 isomers and is of the general sequence:

GA~A,G)GCIGCIAA(A,G)TA(C,T)Gl'ITT(C,T)AA~A,G)GA(C,T)GCIATG
(A,C)G -3' where bases placed in parantheses indicate a mixture and I is inosine. This set is designated MPQP-1, and was designed for binding to the anti-sense strand of the TCS
coding region. The other two primer sets, designated MPQP-2 and MPQP-3, each consist of 128 isomers and together comprise all potential coding sequences of the 32-mer and are of the general sequences:

5'- CG(C,T)TTICCIAT(C,T)TG(C,T)TG(C,T)TCIAT(A,G)AA(C,T)TT
(A, G) TA -3' and 5'- CT~C,T)TTICCIAT(C,T)TG(C, T) TG (C, T)TCIAT(A,G)AA(C, T) TT
(A,G~TA -3', respectively. They were designed for binding to the sense strand of the TCS coding region, and were typically used in a primer mixture designed MPQP-2/-3.
A DNA amplification reaction was carried out by repea-ted primer initiated strand extension, using a commercially suppLied kit (Perkin-Elmer/CetuS) and according to methods supplied by the manufacturer as outlined in Example 2. The product of the DNA
amplification step was isolated by agarose gel electrophoresis, and by polyacrylamide gel electro-phoresis, with detection by ethidium bromide fluorescence .
, - :
.
."

.

and/or autoradiography. A major product of about 255 base pairs was detected.
Figures 3A and 3B show the r)NA sequence of the ampli-fied material, and the amino acid sequences corresponding - 5 to all three reading frames i.n both directions. The underlined translation shows a sequence that is homolgous to amino acids 128 through 163 in TCS. This sequence is within the region predicted to be amplified and confirmed that a TCS or TCS-like coding region was amplified.
Southern blot analyses were performed on the DNA pre-pared from the plant tissue to assess the organization and the complexity of TCS genes in the total DNA
background. The Southern blots were probed separately with 32P-labelled MPQP-l and MPQP-2/ 3. The results (not shown) suggested that there might be several TCS-related genes, and that the overall complexity of the plant genome is on the order of a mammalian genome and could be effectively screened using standard lambda-phage banks~
With continued reference to the method outlined in Figure 2, the amplified coding sequence from above was used as a probe to identify one or more T. kirilowii - genomic library clones containing TCS coding sequences.
The genomic library clones were prepared and probed conventionally, as outlined in Example 2. Two clearly positive plaques were picked, amplified and converted to plasmids, according to protocols suppiied by the manufacturer of the cloning system. One clone, designated pQ21D, contained an approximate 4kb insert;
the other, designated pQ30E, contained an approximate 0.6 kb insert. The pQ21D vector has been deposited with The American Type Culture Collection, 12301 Parklawn Drive, Rockville, MD, 20852, and is identified by ATCC No.
67907. Partial sequence analysis showed that the 4 kb insert contained sequences that coded for a protein :' : : ~ , ,q~ 3 ~3~

having substantially the same amino acid sequence shown for plant-derived TCS in Figure 1. The 0.6 kb insert was found to contain sequences encoding a peptide homologous to, but not identical with, plant-derived TCS.
- 5 The complete sequence for the insert of pQ21D was determined and is shown in Figure 4, along with the cor-responding amino acid sequence of TCS. As seen in the figure, the sequence encodes a protein that contains a continuous amino acid se~uence identical to that of plant-derived TCS except for two conservative changes --a Thr for a Ser substitution at amino acid position 211 and a Met for a Thr substitution at position 224.
The minor differences between the two sequences are presumably related to ~ariations between dif~erent T.
kirilowii strains. The purified TCS was obtained from T.
kirilowil roots from the Canton region of China; the genomic DNA was obtained from T. kirilowli leaves from Korea.
These conservati~e sequence variations illustrate strain-related DNA sequence variations which result in functionally equivalent trichosanthin proteins.
A comparison of the amino acid sequence of mature plant-derived TCS (Figure 1) and that encoded by the DNA
in Figure 4 shows that TCS is likely produced as a secreted protein that undèrgoes post-translational processing at both the amino and carboxy ends.
Specifically, nucleotides 340 through 408 code for a putative secretory signal peptide having the sequence:

Met Ile Arg Phe Leu Val Leu Ser Leu Leu Ile Leu Thr Leu Phe Leu Thr Thr Pro Ala Val Glu Gly.

As can be seen from Figure 10 (Example 5), these first 23 amino acids of the trichosanthin coding sequence have the characteristiC hydrophobicity of a secretory signal. Ac-,, - ....

cordingly, the nucleic acids encoding this sequence will be useful when expressing trichosanthin proteins in plants cell since the sequence will provide a homologous secretory signa:L. Further, recombinant trichosanthin - 5 protelns produced in heterologous expression systems, which retain the 5' leader sequence, can be used as a substrate in assays to identify the leader sequence processing enzyme activity. The 23 amino acid sequence itself can be used as an ant:igen to generate antibodies to examine, for example, the steps of ln vivo protein processing in plant cells.
Nucleotides 1150 through 1206 code for a putative car-boxy terminal extension that is not present in the mature protein, and which has the sequence:
Ala Met Asp Asp A.sp Val Pro ~et Thr Gln Ser Phe Gly Cys Gly Ser Tyr Ala Ile.

As mentioned above for the amino-terminal extension, the carboxy-terminal extension will also be useful for expression of the trichosanthin protein in plant cells.
Although the role of the carboxy-terminal extension has not yet been determined, it is possible that this peptide functions to neutralize the ribosome inhibiting activity of the peptide prior to cellular secretion:
trichosanthin proteins retaining this sequence, or variations of this sequence, will be useful in specifically defining the function of this protein extension. Also, the carboxy-terminal sequence itself can be used as an antigen to generate antibodies to examine, for example, the steps of ln vivo protein maturation in pla~lt cells.
According to one aspect, the invention includes a nu-cleic acid which encodes for a trichosanthin protein .

.: :

, 1 9 ~ C~ J tD
which has the functional properties of Trichosanthes-obtained TCS. The nucleic acid preferably has the sequence shown in Figure 4, where basepairs 409-1149 of the sequence code for mature TCS from T. kirilowii. The - 5 nucleic acid of the invention may include:
(a~ basepairs 409 to 1149 which encodes mature tri-chosanthin from T._kirilowii;
(b) ln addition to (a), basepairs 340-408, which en-codes a putative amino terminal extension of the mature form of trichosanthin from T. kirilowli;
(c) in addition to (a), basepairs 1150 to 1206 which encodes a putative carboxy terminal extension of the mature form of trichosanthin from T. kirllowll; and (d) a TCS coding sequence joined with a ligand coding sequence, encoding a fused protein having a ligand peptide which confers cell-surface recognition properties on the fused protein.

C. Expressin~ Recombinant TCS Protein Recombinant TCS was produced using the above TCS coding sequence, following the steps outlined in Figures 4 and 5, and described in Example 5. With reference to Figure 5, plasmid pQ21D from above was digested with EcoRI and NcoI, releasing a 1.2 kb fragment insert containing the complete coding sequence for TCS. This TCS-coding fragment was cloned into plasmid pKK233-2 which was previously digested with EcoRI and NcoI. After replication the recombinant plasmid, designated pQ21D/pKK233-2, was divided into two samples. One sample was digested with EcoRI and SalI, and and the second sample with SalI and Ncol to generate an EcoRI/SalI amino portion fragment and a SalI/NcoI carboxy portion fragment. The two fragments were cloned into M13 phage vectors for site speci~ic muta~enesis, to place a NcoI

, ~

~3~ 3 site containin~ an ATG start codon at the amino terminal end of the mature TCS coding sequence, and a double T~A
translation stop sequence plus a HindIII cloning site after the carboxy end of the mature sequence, as - 5 illustrated in Fi~ure 5.
The modified sequences were excised from the mutagenized clonas, and cloned together into a pKK233-2 expression vector (Pharmacia) which contains a synthetic trp/lac promoter positioned appropriately ahead of a ribosome binding slte that is also positioned appropriately ahead of an ATG start codon contained within an NcoI site. Several clones were characterized and verified to contain the modified insert in the correct orientation. The DNA sequences of the modified regions were directly verified for one clone, designated pQR19.
More generally, the pQRl9 expression vector is exemp-lary of a TCS coding sequence operatively placed in an expression vector for TCS expression in a suitable host.
In a preferred embodiment, and as exemplified by pQR19, the expression vector construction contains a promoter, a ribosome binding site, and an ATG start codon positioned before and adjacent the amino terminal codon of mature TCS, and a stop codon positioned after and adjacent at the carboxy terminal codon of mature TCS.
For expression of recombinant TCS (rTCS), plasmid pQRl9 and similar clones were propagated in an appropriate E.
coli host strain that carries a lacIq gene for regulation of the synthetic trp-lac promoter. The host strain XL-1 Blue (Bullock) was employed. Its relevant genotype is recA1, endA1, gyrA96, thi, hsdR17 (rk-, mk+), supE44, relA1, ~-, lac-[F', proAB, lacIqZ~M15, TnlO (tet~)].
Induction of promoter activity may be achieved by adding 5 mM IPTG ~ opropylthiogalactoside). Under culture :: ' ~ 3 conditions described in Example 3, cells carrying pQR19 and similar plasmids were induced and, at a selected cell density, the cells were harvested and disrupted by sonication. Aliquots of total cell material, of material - 5 pelleted at 15,000 x g for 5 min, and of material remaining in solution at 15,000 x g for 5 min were analyzed by polyacrylamide gel electrophoresis and sub-sequently by Western blot analysis. The Western blot was probed with rabbit anti-TCS sera.
The results showed an immunoreactive product that co-migrated with authentic TCS in the total cell and soluble cell fractions from pQRl9~XLl-blue induced cells, but not in the insoluble fraction from the same cells, nor in any fraction from pKK233-2(vector)/XLl~blue induced cells, i.e., cells containing the pKK233 expression vector without the TCS coding insert.
The pQR19 expression vector which contains the TCS cod-ing sequence, and which expresses rTCS in a sultable bac-terial host has been deposited with The American Type Culture Collection and is identified by ATCC No. 67908.
Clarified cell extract material was fractionated using the steps described in Example 1, yielding rTCS with a purity, as judged by gel band staining with Coomassie blue on SDS polyacrylamide gels, of greater than 90~.
About 9 mg of purified rTCS were obtained from nine liters of culture.
The rTCS protein produced is exemplary of an rTCS pro-tein derived from the amino acid sequence shown in Figure 4. More generally, the rTCS protein of the invention in-clud~s a recombinant protein containing the entire aminoacid sequence for mature TCS, as described above, and a recombinant TCS protein containing an amino-terminal extension having the sequence:
_ - .:
.

.

22 ~ t~
Met Ile Arg Phe Leu Val Leu Ser Leu ~eu Ile Leu Thr Leu Phe Leu Thr Thr Pro Ala Val Glu Gly:

and/or a carboxy-terminal extension having the sequence:

Ala Met Aqp Asp A~p Val Pro Met Thr Gl~ Ser Phe Gly Cy5 Gly Ser Tyr Ala Ile.
The invention thus further includes a recombinant process for the production of a trichosanthin protein having the functional properties of Trichosanthes-obtained trichosanthin. The method includes the steps of inserting a DNA sequence encoding said protein into an expression vector, transforming a suitable host with the vector, and isolating the recombinant protein expressed by the vector.
In one preferred embodiment, the expression vector is pQR19 and the host is E. coli.

D. Bioactivity of Recombinant TCS
As pre~iously described in above-cited U.S. Patent No.
4,795,739, TCS obtained from T. kirilowii is a potent and selective inhibitor of HIV antigen expression in HIV-in-fected T cells and monocyte/macrophages. The inhibitory effect of rTCS on expression of HIV-specifiC antigens in HIV-infected T cells can be demonstrated as follows.
Acutely HIV-infected human T cells were treated with varying concentrations of rTCS. After four da~s culture, the amount of HIV p24 antigen present in cPll free culture supernatants was quantitated using a commercially available antigen capture immunoassay (Coulter).
Inhibition was determined by comparison of results for treated cultures and untreated cultures.
The viral inhibition studies detailed in Example 4A
compared the inhibitory ac~ivity of plant-produced TCS
with the above rTCS protein. ~he plots in Figure 6 show . .

23 ~ ~ r~
percent inhibition of p24 HIV antigen production as a function of culture concentration of plant derived TCS
(closed boxes) and rTCS produced as above (open boxes).
As seen, both proteins gave substantially the same level of inh~bition at higher protein concentrations, although the plant-derived protein was more ef~ective at the lowest protein concentrations.
Also, as mentioned above, it has been shown that plant-produced TCS is a potent inhibitor of protein synthesis in a cell-free lysate system. The protein-synthesis inhibitory properties of both plant-produced TCS and rTCS
were compared in a rabbit reticulocyte lysate system, as outlined in Example 4B. The plots in Figure 7 show percent inhibition of 3H-leucine incorporation as a function of concentration of plant-derived TCS (closed boxes) and rTCS (open boxes) in the rabbit reticulocyte system. The plots show that both plant-produced and recombinant TCS have substantially the same specific protein synthesis inhibitory activity.
E. Structure-function studies of a-trichosanthin throu~
manipulation of a synthetic gene.
A synthetic gene for ~-TCS has been constructed ~Example 6, Figure 1~) to facilitate mutational analyses of ~-TCS, aimed at elucidating structure-function relationships, and to better understanding the ~-TCS mechanism of action in blocking HIV replication. The synthetic g~ne contains unique restriction sites spaced 20 to 90 bp apart (Figure 11), thus allowing convenient introduction of mutations by cassette replacement. The nucleic acid sequence ~or the synthetic gene was created by assigning nucleic acid codons corresponding to the primary amino acid sequence of the mature ~-trichosanthin protein sequence.
Accordinyly, the translation pr~duct of the synthetic .. . ~ , .

`: ' gene corresponds to the mature a-trichosanthin (~~TCS) protein sequence.
An energy minimi~ed molecular model for ~-TCS has been generated by fitting its primary amino acid se~uence to - 5 the known crystallographic st:ructure for ricin A-chain, a related RIP. Guided by this model and by peptide sequence homology alignments with ricin A-chain and other RIPs (Figure 9), RIP-invariant residues residing in a putative active site cleft were determined. Two of these sites, Glu160 and Argl63 of the -TCS peptide sequence (Example 6, Figure 11), have been modified to assess their relationship to the translation-inhibitory and anti-HIV-1 replication activities of -TCS. These mutations altered Glu160 and Argl63 to Asp and Lys, respectively. Proteins containing one or both of these alterations were expressed in E. coli, and were purified to approximately 95% homogeneity (Example 6). These mutant pxoteins were compared to the unmodified protein made from a synthetic gene (KQS; also expressed in E.
coli), for their ability to inhibit ln vitro translation (IVT) in a rabbit retlculocyte system and to reduce production of p24 antigen in HIV-1 infected T-cells.
The doubly-modified variant (DK12) was found to be al-most 3 logs less active at inhibiting in vitro translation ~Figure 15) and more than one log less active at inhibiting p24 production, when compared to the unmodified protein. The singly-modified variants showed intermediate activities, for both inhibition of translation and p24 production, relative to DK12 and KQS.
Additional mutant proteins are being produced that will allow investigation of the function of other RIP-invariant residues in the -TCS molecule.
The synthetic gene described above, provides a tool to generate variants of the ~-triFhosanthin protein. These . ~:

~, :

2 ~

proteins can be screened, as described above, for less active and more active variants which affect the ribosome inhibitory and/or HIV-I inhibitory activities of the wild-type protein.

F. TCS Fusion Protein In another aspect, the invention includes TCS fused at its amino or c~rboxy end with a ligand peptide to form a fused ligand/TCS protein. The TCS making up the fused protein is preferably rTCS or bioactive portion thereof, as described above.
Where TCS is used to inhibit viral expression in HIV-infected human cells, the protein may be advantageously fused with a soluble CD4 peptide, which shows specific binding to the HI~-related gpl20 antigen present on the surface of HIV-infected cells (Till), or with a monoclonal antibody specific against an HIV-specific cell surface antigen.
The fused TCS protein may be ~ormed by chemical con-jugation or by recombinant techniques. In the formermethod, the peptide and TCS are modified by conventional coupling agents for covalent attachment. In one exemplary method for coupling soluble CD4 to TCS, recombinant CD4 ~rCD4) is derivatized with N-succinimidyl-S-acetyl thioacetate ~Duncan), yielding thiolated rCD4. The activated CD4 compound is then reacted with TCS derivatized with N-succinimidyl 3-~2-pyridyldithio) propionate (Cumber), to produce the fused protein joined through a disulfide linkage.
As an alternative method, recombinant TCS (rTCS) may be prepared with a cysteine residue to allow disulfide coupling of the rTCS to an activated ligand, thus simplifying the coupling reaction. The TCS expression vector used for production of rTCS can be modified for : :

2~ J ~3 lnsertion of an internal or a terminal cysteine codon according to standard methods o~ site-directed mutagenesis.
In a preferred method, the fused protein is prepared - 5 recombinantly using an expression vector in which the coding sequence of the fusion peptide is joined to the TCS coding sequence. Figure 8 illustra~es the construction of an exemplary expression vector for a fused TCS/CD4 protein.
Briefly, an EcoRI-StuI DNA fragment containing the coding region for the first 183 amino acids of mature CD4 peptide, which may effectively bind gpl20, (Maddon) is inserted into an M13MP19 phage between SmaI and EcoRI
sites and the vector, in a single~strand form, is then subjected to primer mutagenesis. SpecifiCally, the amino-terminal portion of the CD4 gene is modified with primer MP101 (5~-ccAGcAGccATGGAGGGAAAcAAAG -3'); and the carboxy portion of the gene is modified with primer MP102 (5'-CATCGTGGTGCTAGCT-CCACCAC Q CcAccAccAccAccAccAcccATGGAGGcATGcAAGcTTG -3~).
These modifications place an NcoI site containing an ATG
st~rt codon at ~he beginning of the mature CD4 peptide coding sequence, and a string of proline codons terminating at an NcoI cloning site after amino acid 180 in the CD4 sequence, as illustrated in Figure 8.
The NcoI fragment ~rom the phage vector is inserted into the pQRl9 expression vector from above previously cut with NcoI. Successful recombinants are confirmed by restriction analysis for proper orientation of the CD4 sequence insert.
An expression vector formed as above, and designated pQR19/CD4 in Figure 8, contains (a) a synthetic tr ~ c promoter positioned appropriately ahead of a ribosome binding site that is also positioned appropriately ahead ,, : . :

s~
~7 of an ATG start codon contained within an NcoI site, (b) the CD4 coding sequence, (c~ a spacer coding sequence coding for 10 proline residues, which spaces the CD4 and TCS protein moieties, (d) the coding sequence ~or mature - 5 TCS and (e) a stop codon positioned adjacent the carboxy-terminal codon of mature TC~;. The method generally follows that used in fusing a soluble CD9 to domains 2 and 3 of pseudomonas exotoxin A, as described previously (Chaudhary).
Plasmid pQR19/CD4 is analysed for expression of fused TCS protein as above. Briefly, the expression vector is cultured in a suitable bacterial host under IPTG
induction conditions to a desired cell density. Tha cells are harvested, ruptured by sonication, and the cell material is clarified by centrifugation. The clarified material is tested for (a) binding to gp120 antigen, to confirm CD4 ligand binding activity, and (b) for ribosome inhibition acti~ity, to confirm TCS enzymatic activity.
The protein may be purified by molecular-sieve and ion-exchange chromatography methods, with additional puri-fication by polyacrylamide gel electrophoretiC separation and/or HPLC chromatography, if necessary.
It will be appreciated from the above how other ligand/TCS-Containing fusion proteins may be prepared.
One variation on the above fusion is to exchange positions of the CD4 and TCS molecules in the fusion protein.

III. Genomic Clonin~ of Ribosome-inacti~atin~ Proteins .
deriving from a Multi-gene Family in Trichosanthes kirilowii Maxim.
Two Type I ribosome-inactivating proteins (RIP) have been isolated from the plant Trichosan~hes kirilowii Maxim~ trichosanthin, i~olated from the root . ~ , .
. .

tubers (Law, Yeung~; and, ~2) trichokirin, isolated from seeds ~Casellas). As described above, two cloned sequences related to ~-trichosanthin, designated pQ21D
and pQ30E, ha~e been isolate~ from the genome of T.
- 5 kirilowii. In an effort to :isolate a full length clone for pQ30E, an SpeI lambda Z~?II genomic library was constructed (Example 7). Th:Ls library was probed with the entire EcoRI insert from clone pQ30E (Figure 16) undsr low stringency conditions. Of 21 cloned inserts identified by a positive hybridization reaction (Example 7), 5 clones showing different restriction digest patterns were selected for complete DNA sequence analysis; two of these cloned inserts had identical sequences. Four unique cloned inserts, designated pQ2, pQ3, pQ12 and pQ24, were also identified; each insert encoded a putative full length RIP protein (Figures 17-20).
DNA sequence and protein alignment analysis showed that the insert coding sequences fall into 3 sub-groups with amino acid sequences similarities ranging from 58%-75%
between the subgroups, and >90~ within each subgroup - (Figure 21). By comparison and alignment with pQ21D, the insert coding sequences contain the following characterist~CS: l) each insert sequence encodes a pre-pro protein with a putative signal peptide of 23 aminoacids, and a carboxy terminal extension consisting of 1, 10 or l9 amino acids; 2) each insert sequence contains amino acid residues which are characteristic for RIPs;
3) the corresponding proteins possess between 0 and 4 potential N-linked glycosylation sites and have calculated pI's of 9.6 to 9.9; 4) the predicted size of the proteins are 247 or 248 amino acid residues; and, 5) the proteins are translated from unspliced messages --there are no introns in any of ~he clonesO

_ ~ 3~ 3 From the data presented above the genome of Tricho-santhes kirilowii Maxim appears to contain a multi-gene family encoding multiple ribosome-inactivating proteins;
this is similar to an RIP family in, for example, Ricinus communis.
The coding sequences of the members of this multi-gene family can be used as probes to identify, for example, additional members of the gene family in Trichosanthes or to identify homologous genes in other plants. The proteins corresponding to the multi-gene family coding sequences can be expressed in bacterial expression systems tas described above for ~-~richosanthin), and the proteins used to ~l) examine the proteins' ribosome-inactivating and anti-HIV-I properties, and, (2) generate antibodies which can be used to examine expression patterns of the proteins in Trlchosanthin tissues.
Figure 22 is a representatiOn of the ribosome-inacti-vating protein multi-gene family of Trichosanthes kirilowii. The vertical columns of amino acid residues at each site show acceptable substitutions at that site based on comparisons of the coding sequences of all the known genes of the family (Example 7, Figure 21). For example, in Figure 22 the first amino acid residue is M, followed by I or N, followed by R, followed by F, followed by L, P, or S, etc. Nucleic acid coding sequences can be determined for any protein coding sequence derived from Figure 22 and probes or synthetic genes synthesized for the chosen protein coding sequence as described above in Section II-E for ~-TCS. The probes can be used to identify, for example, additional members of the ribosome-inactivating protein gene family in Trichosanthes or to identify homologous genes in other plants. The synthetic gene can be used to recombinantly express the chosen protein coding sequence: this , ' , . .

2 ~

protein's ribosome-inactiva~ing and anti-HIV-I properties can then be examined.

IV. Obtaining RIP Coding Sequences As described above, the cocling sequence of TCS was obtained by selective amplification of a TCS coding region, using sets of degenerate primers for binding to spaced coding regions of a TCS coding sequence in genomic DNA. This section describes the use of such ssts of degenerate primers for selective amplification of coding sequences for a variety of RIPs.
In selecting suitable primer sets, the amino acid sequences of TCS and one or more RIPs are examined for regions of sequence homology, i.e., regions where the amino acids sequences are identical or differ at most by one or two amino acid residues. Typically, the length of the regions being examined should contain at least about 7 amino acids, i.e., at least about 20 nucleotides, although it is appreciated that longer oligonucleotide primers axe preferred, even though overall complexity is increased.
Figure 9 shows the complete amino acid sequences of TCS
(top line), and three RIPs whose sequences have been published. The RIPs are ricin A chain kicA; Lamb), ~brin A chain tabrA; Funatsu) and barley protein synthesis inhibitor (BPSI; Asano), and Mirabillis anti-viral protein (MIRA; Habuka). The amino acids are indicated by conventional one-letter codes. Amino acid matches among the four proteins are shaded.
As seen from the figure, there are several regions, each containing at least seven amino acids, which show a high degree of amino acid sequence homology among the proteins, i.e., sequence matching in at least about 4 of the 7 amino acid positions. The relatively greater ,, '': ~' . . ' ' , . .
. .

2~3~J ~3 .. _ homology among TCS, ricin A chain and abrin A chain, as compared with barley protein synthesis inhibitor, presumably reflects evolutionary divergence since TCS, ricin A and abrin A chain are all derived from - 5 dicotyledons, and barley inhibitor is obtained from a monocotyledon.
Considering the sequences from amino acids 63-70 in the upper line in Figure 9, it is seen that the amino acid sequence for abrin A chain --GIDVTNAY-- differs from the corresponding TCS and ricin A chain sequences by only 2 amino acids each, and therefore is a likely choice for design of the primer set. The disadvantaye of this sequence region is that the presence of a leucine residue (L) in ricin A chaln introduces a six-fold degeneracy at that point in the sequence. However, this problem is not prohibiti~e if inosine (I) is used in the third and/or first position in each codon to reduce degeneracy.
The abrin sequence in the same 63-70 amino acid region differs from the corresponding TCS by an alanine-to-glycine substitution in the first position. Since aG(G,C)N sequence, where N is all four nucleotides, will hybridize with both the alanine and glycine codons, the primers can be made an additional two-fold degenerate at this position to encompass both TCS and abrin coding sequences- The abrin sequence alsD differs from the TCS
sequence by a valine-to-alanine substitution in the seventh position. An additional two-fold degeneracy at this position can be made that accounts for all possible ~aline and alanine codons, i.e., G(C,T)N.
Likewise, the abrin sequence in this region differs from the corresponding ricin A chain sequence by the same alanine-to-glycine substitution in the first amino acid position. Additionally, the abrin sequence differs from the ricin sequence by an isoleucine-to-leucine . ', "; ~ ',, , . ..
.

2 ~ 'g ~

substitution in the second position. Since an ITI
sequence will hybridize with all the isoleucine and leucine codons, the primer degeneracy can be normalized at this position. The other five amino acid positions - 5 are preferably made degenera~e, to optimize the specificlty of primer binding to corresponding genomic coding regions. The total number of primers in the final primer set is preferably between about 16-128 although more complex mixtures can be used. The primers are syn-thesized conventionally using commercially availableinstruments.
A second set of degenerate primers from another region of TCS which is homologous in amino acid sequence to RIPs is similarly constructed.
The two primer sets are useful in a method for selec-tively amplifying RIP coding sequences present in genomic DNA from selected plant sources, employing repeated primer-initated nucleic acid amplification. As an example, to ampli~y coding sequences for abrin A chain protein, genomic DNA from Ab_us precatorius is isolated, and mixed with the primer sets, all four deoxynucleosides triphosphates, and polymerase, as outlined in Example 2.
After repeated cycles of primer binding and strand extension, the material is ~ractionated by gel electrophoresiS and amplified fragments are identified, for example, by ethidium bromide staining or by autoradiography, according to procedures described in Example 2. Fragments amplified from an RIP gene can be identified by size, as the selection of specific primer sets would predict the size range of the fragment that is amplified. Genes for RIPs are not believed to contain any introns (Halling and the present application).
The amplified material is then used as a (radiolabeled) probe for detecting genomic library clones prepared from , - ~
~.

genomic DNA from the plant source, e.g., Abrus prec_to-rius. The identified library clones are analysed, as above, for fragments containing a complete RIP coding se quence. Alternatively, overlapping genomic library frag-- 5 ments containing amino and carboxy portions of the coding sequence can be combined to produce a complete coding se-quence. The properties of the coding sequence are then tested as outlined abo~e to determine ribosome-inhibitory properties and/or anti-viral properties. Further, these nucleic acid coding sequences can be used as probes to identify additional RIP sequences.
More generally, this aspect of the invention includes a primer mixture and method of using the mixture for selec-tively amplifying RIP coding sequences. The primer mixture includes a first set of sense-strand degenerate primers, and a second set of anti-sense primers, where each set contains at least one primer sequence which is effecti~e to hybridize with the corresponding coding sequence in TCS which encodes the region of amino acid homology with RIPs, particularly RIPs from dicotyledon plants.
Once the amplified genomic sequence of a ribosome-inactivating protein is obtained the sequence can be used as a probe for isolating genomic library fragments containing the desired RIP coding sequence. The protein products expressed from these genomic fragments can then be tested for their ribosome-inhibitory activity and/or anti-viral activities as described above for TCS.
It will be appreciated that the method can ~e used to obtain the codir.g se~uence from plants which produce known RIPs, and also to screen other plants for the presence of genes encoding as-yet-unknown RIP or RIP-like proteins.
. ~

., ~.

;

2 ~

The following examples illustrate various methods used to obtain and verify the nature of the coding sequence and recombinant proteins described above. The examples are intended to illustrate, but in no way to limit, the scope of the invention.
Materials and Methods T. kirilowii root tubers were obtained from the Canton region of the People's Republic of China. Leaves of T.
kirilowii were obtained from Korea and were collected and immediately frozen on dry ice for shipment. Samples were than stored at -70C.
QAE Zetaprep~ anion exchange cartridges and SP Zeta-prep~ cation exchange cartridges were supplied by AMF
Cuno Corp. (Meridan, CT); and Pellicon ultrafiltration membranes (10,000 MW cuto~f), from Millipore Corp.
(Bedford, NA).
M13~MP18 and M13/MP19 were obtained from New England Biolabs (BeYerly, MA). Lambda-Zap II~M cloning ~ector system was supplied by Stratagene (La Jolla, CA).
Expression vector PKK233-2 and its IPTG-inducible E. col host strain, XLI-blue, were obtained fxom Pharmacia (piscataway~ NJ) and Stratagene (La Jolla, CA), respectively. Restriction enzymes were obtained ~rom New England Biolabs (Beverly, M~ or Promega tMadison, WI).
DNA primer-initiated amplificatio~ reagents were obtained from Perkin-Elmer/Cetus (Norwalk, CT).
SynthetiC olisonucleotide primers were prepared by conventional, automated phosphoramidite methods using either a Biosearch Cyclone or an Applied Biosystems Model 380B instrument.
The methods for preparation and manipulation of nucleic acids, and t~e recombinant DNA techniques emloyed herein are broadly accepted and applied and are generally refer-enced by Ausubel, F.M. et al. .(eds) "Current Protocols in ::

. .

Molecular Biology" Vols. 1 and 2, John Wiley & Sons, New York (1988) and Maniatis, T., et al., "Molecular Cloning:
A Laboratory Manual," Cold Spring Harbor Laboratory, 1982.

Example Purificatlon of TCS
The purification of TCS has been described in co-owned, co-pending U. S. Application No. 07/333,181. A clarified extract of the roots of T. kirolowii was obtained by overnight extraction of homogenized tubers of T. kiri-lowii in normal saline (0.9% NaCl) adjusted to pH=8.0 with NaOH. The extract was clarified by centrifugation, and the clarified material was passed through a QAE
Zetaprep~ anion exchange resin, which is supplied commercially in cartridge form. The ion exchange step was carried out at low ionic strength, i.e., low conductivity, which has been found effectlve to enhance TCS purification, and in particular, to remove hemagglutinin contaminants. The low-conductivity buffer was 20 mM phosphate, pH 8Ø
The flowthrough ~rom the anion exchange resin was adjusted in pH and ionic strength, and preferably concentrated, preparatory to further protein purification by chromatograph~ on a cation exchange resin. The concentration step was carried out by ultra~iltration using a 10,000 molecular weight filtration membrane, yielding a solution which is largely free of low-molecular weight contaminants.
The treated flowthrough material equilibrated with 50 mM phosphate, pH 6.0 buffer was applied to an equilibrated SP Zetaprep~ cation exchange resin, and the column was washed extensively with buffer tl5-20 volumes) until the elution profile reached a baseline value. The . . . ~
- ~ ~

,, : ., . .

extensive washing removed loosely bound material, including, particularly, endotoxins and high molecular weight lipopolysaccharides (LPS), and is necessary for achieving high purity TCS.
TCS was now eluted from the column in highly purified form by elution with 50 mM phosphate buffer, pH 6.0 containing 60 mM NaCl, to release bound TCS ~rom the resin. The purified TCS protein was at least about 98~
pure, as evidenced by HPLC profile and staining patterns on SDS gel elect.rophoresis.

Example 2 Pre~ ing Cloned Genomic Fra~ment Containlng TCS Codin~ Sequence A. Amplified TCS Codiny Sequence Genomic DNA was isolated from frozen T. kirilowii leaves by a modification of published methods (Taylor).
~riefly, frozen tissue was ground to a fine powd~r using a mortar and pestle kept on dry ice. ~-mercaptoethanol was then added to 2% of the initial volume followed by an equal volume of hot 2x extraction buffer ~2~
cetyltrimethyl~ammonium bromide (CTAB), 100 mM Tris-Cl, pH 8.0, 20 mM EDTA, 1.4 M NaCl).
This slurry was gently stirred in a 55C water bath until the temperature reached 50C. The slurry was then transferred to appropriate centrifuge bottles and extracted twice with an equal volume of chloroform:isoamYl alcohol (24:1). Phase separation was achieved by centrifugation. A 1/10 volumP of 10% CTAB
was added and the extraction repeated.
The upper aqueous phase was removed to another con-tainer and the DNA precipitated by adding an equal volume of precipita~ion buffer (1~ CTAB, 50 mM Tris-Cl, pH 8.0, 10 mM EDTA) to lower the sodium concentration to 0.35 M.

, ~ -3 7 ~ ~ r, ~) ~ ",~
The DNA was collected and washed with cold 70% ethanol, O.1 M sodium acetate to convert the DNA to a sodium salt, followed by a wash by 95% cold ethanol. The DN~ could then be dried and redissolved in 10 mM Tris-Cl, pH 7.5, lmM EDTA. To further eliminate contaminants, the DNA was re-precipitated from CTAB by adding an equal volume ~original) of 2x extraction buffer, followed by two volumes of ~original) TE buffer (10 mM Tris-HCl, 1 mM
EDTA), pH 8Ø The DNA was once again converted to the sodium salt, washed with ethanol as above, dried, and dissolved in TE buffer, pH 8Ø Greater than 5 mg of high molecular weight DNA was obtained from approximately 35 g of tissue.
Three degenerate sets of probe sequences were syn-thesized, corresponding to two separate coding regions.
The first DNA sequence is a 35-mer and encompasses the protein sPquence overlined and denoted A in Figure 1, and the second sequence is a 32-mer and encompasses the prot~in sequence overlined and denoted B in the figure.
The probe sets were prepared by conventional automated methods using instruments commercially available and followiny the manu~acturers~ instructions. (Biosearch, San Rafael, CA, and Applied Biosystems, Foster City, CA).
Deoxyinosine nucleotides were incorporated in order to generate probes longer than 20 nucleotides of manageable complexity ~Ohtsuka; Takahashi). The sense-strand probe set corresponding to the 35-mer, designated MPQP-1, included 128 lsomers. The anti-sense-strand second and third sets, designated MPQP-2 and MPQP-3, each included 128 isomers and were 32-mers.
.~ DNA amplification reaction was carried out by repea-ted primer initiated strand extension, in a reaction mixture containing ~a) 1-2 micrograms of the T. kirilowii DNA isolated as above, (b) 32P-labeled MPQP-1 and an . : . , 2 ~ 7' ~9 ~

equimolar mix of unlabeled MPQP-2 and -3, as primers, (c) all four deoxynucleoside triphosphates, and (d) Taq polymerase. About 20 rounds of thermal cycling were performed, employing convent:ional DNA amplification reaction conditions, as outlined in instructions from the manufacturer (Perkin Elmer-Cetus, Norwalk, CT). A
similar DNA-amplificatiOn reaction was carried out using unlabeled primer setsO
The product of the DNA amplification step was frac-tionated on 3% Nusieve, 1% ~E agarose (Seakem~, FMCBioproductS, Rockville, MD) and stained with ethidium bromide. A major product of about 255 base pairs was detected. The material was also fractionated on 5%
polyacrylamide gel electrophoresis and the bands detected by autoradiograp~, with similar results. In both cases, very little DNA other than the amplified material was detected.
Amplified DNA was recovered from polyacrylamide gels by elution followed by ethanol precipitation. A portion of one such preparation, approximately 100 nanograms, was taken for DNA sequence analysis. The DNA sample plus 30 ng of unlabeled MPQP-l were taken up in 10 ~1 of TE ~lOmM
Tris-HCl, pH 7.5, lmM EDTA) and heated to 100C for 5 minutes to denature the double-stranded fragment. The mixture was quick-frozen on dry ice to prevent the template from annealing. Two ~1 of 5X Sequenase sequencing buffer (USB Biochemicals, Cleveland, OH) was added and the primer allowed to anneal to the template for 5 minutes at 37C. The standard sequencing protocol supplied by the manufacturer was then followed.
The DNA sequence obtained and its translation into all three reading frames is shown in Figures 3A (for the sense strand) and in Figure 3B (for the complementary strand). ~

.
, " .

: , : ~

~ ~ r~

B. Cloned Library Fragment with the Complete TCS
Coding Seq~
Genomic DNA obtained as above was digested to comple-- 5 tion with EcoRI and cloned into a standard library cloning vector, in this case, the Lambda-Zap II$~ system of Stratagene (La Jolla, CA). For use as a probe, the amplified 255-bp fragment from above was radiolabeled by random priming (Boehringer-Mannheim kit, Indianapolis, IN).
Approximately 0.5-1.0 x Io6 plaques were probed with the 32P-radiolabeled 255-bp probe. Two clearly positive plaques were picked, amplified and converted to plasmid, according to protocols supplied by Stratagene. One clone, designated pQ21D, contained an approximate 4kb insert which included the complete TCS coding sequence;
the other, designated pQ30E, contained an approximate 0.6 kb insert.
The region of pQ21D containing the TCS coding region was sequenced by standard double-strand sequence methods, using universal sequence primers as well as unique - synthetic oligonucleotide primers as needed. A smaller subclone containing only the TCS coding region was generated by subcloning the 1.2 kb EcoRI to NcoI fragment 25 (Figure 4) from pQD21D into pKK233-2. The resulting recombinant plasmid was designated pQDl2D/pXK233-2.

Example 3 ExPressinq Recombinant TCS (rTCS) The pQ21D~pKK233-2 cloning vector from Example 2 was divided into two samples. One sample was digested with EcoRI and SalI, to release an EcoRI to SalI fragment con-__ _ _ taining the amino portion of the TCS gene. A second portion of the DNA was digested first with NcoI, and 2 ~ 3 ~ ~ 9 ~ ~

treated with Klenow to generate a blunt end. The DNA was then digested with SalI to release a SalI to NcoI ~blunt) fragment containing the carboxy portion of the gene.
After isolating the two fragments by gel electrophoresis, - 5 the EcoRI to SalI fragment was cloned into M13MP19 (EcoRI
to SalI), and the SalI to NcoI tKlenow repaired) fragment was cloned into M13MP18 tSa~I to SmaI). Fragment insertion and production of single-strand phage DNA was performed according to ~nown methods.
The phage single-strand DNA's were subjected to primer mutagenesi~ using standard methods. The amino portion of the gene (in the M13MP19 vector) was modified with primer QNcoN (5'- CcTGcTGTGGccATGGATGTTAGc -3'), and the carboxy portion of the gene was modified with primer QTerl (5'-CGAAACAATATGGCATAATAAAGCTTCCGAGCTCG -3'). These modificationS placed an NcoI site containing an ATG start codon at the beginning of the mature TCS protein sequence and a double TAA translation stop sequence plus a HindIII
cloning site after the carboxy end of the mature sequence, as illustrated in Figure 5.
The fragments containing the modified sequences were excised from purified phage DNA as an NcoI-SalI and an SalI-HindIII fragment, respectively, and cloned together into NcoI-HindIII digested pKK233-2. pK~233-2 is a plasmid containing a synthetic tr~ lac promoter positioned appropriately ahead of a ribosome binding site that is also positioned appropriately ahead of an ATG
start codon contained within an NcoI site. It is supplied commercially (Pharmacia).
Several clones were characterized and verified to contain the modified insert. The DNA sequences of the modified regions were directly verified for one, designated pQR19.

~;
, , ~ ~3 r~

The plasmid pQR19 and similar clones were propagated in the E. coli host strain, XL1-blue. The significant fea-ture of the strain is that it carries the lacI9 represSOr gene on a F' episome (discussed above). LacIq protein controls expression from the lac operator and is blocked from repression by the addition of IPTG to 5mM.
Plasmid pQR19 and another isolate were analyzed for expression of TCS. Culturas were first grown in Luria broth medium supplemented with 100 ~g/ml ampicillin, to select for maintenance of the plasmid, to an A600 f 0.7 measured at coo nm before adding IPTG, then allowed to grow for 4 hours. These conditions did not result in high levels of expression.
Cultures were then inoculated in Luria broth plus 100 ~g/ml ampicillin containing 5 mM IPTG, and allowed to grow to saturation density overnight (pQRl9/XLl-blue induced cells). The induced cells were collected by centrifugation, resuspended in 100 mM Tris-HCL, pH 8.5, 5 mM EDTA at a concentration of about 10 A600 units/ml and disrupted by sonication. Aliquots were taken and centrifuged at 15,000 x g for 5 minutes to separate - soluble from insoluble components.
The insoluble, pelleted material was resuspended in sonication buffer to the same volume as the original ali-quot. Samples of each fraction were run on 10% SDS-PAGE.
One set of samples was stained for total protein with Coomassie Blue; another set of samples was blotted for Western analysis, with the results discussed in Section II.
30Example 4 Biolo~ical Activity of rTCS
A. Inhibition of HIV Replication The ability of rTCS to mediate selective inhibition of HIV replication in infected T-oells was evaluated in , 42 ~ ~ 3 parallel with purified plant-derived material. Cells o~
the CD4+ T-cell line VB (Lifson, 1986) were inoculated with HIV-1 by incubation at 37C for one hour with an aliquot of a titered cryopreserved HIV-1 virus stock (virus isolate HIV-lDV (Crowe, 1987)). After washing, the cells were resuspended to 1.11 x 105 per ml, and 0.9 ml of thls suspension platecl in replicate wells of 24 well culture plates. 0.1 ml. volumes of serial dilutions of purified plant-derived TCS and rTCS were then added at lOX the desired final concentrations to yield 1.0 ml cultures containing 1 x 105 cells in 1.0 ml of culture medium containing the desired concentration of TCS.
After culturing for 4 days at 37C in a humidified 5%
CO2/air atmosphere, culture supernatants were harvested and viral replication in treated and control cultures was assessed by measuring HIV p24 antigen content using a commercially available capture immunoassay kit according to manufacturer's instructions (Coulter, Hialeah, FL).
As shown in Figure 6 (closed boxes) t in accord with observations reported elsewhere (U.S. Patent No.
4,795,739), plant-derived TCS purified to apparent homogeneity from the root tubers of T. kixilowii inhibited HIV replication in a concentration-dependent fashion in this acute infection assay system. The biological activity of rTCS produced in E. coli and purified to apparent homogeneity (open boxes), was essentially indistinguishable from ~hat of the native product when tested in parallel in an assay system for inhibition of HIV replication at TCS concentrations above 0.005 ~g/ml (Figure 6). At lower concentrations, rTCS
appears to show slightly less specific activity than the plant-derived protein.

B. Inhibition of Cell Free ~ranslation In Vitro :~

, ,~ ' ' 2 ~ 3 The abilLty of TCS to irrever~sibly inactivate ribo-somes, thereby inhibiting protein synthesis, is conveniently measured in standardized assays of in vitro translation utilizing partially defined cell free systems composed, for instance, of a rabbit reticulocyte lysate preparatiOn as a source of ribosomes and various essential cofactors, mRNA template~s) and amino acids.
Us0 of radiolabelled amino acids in the reaction mixture allcws quantitation of incorporation of free amino acid precursors into trichloroacetic acid precipitable proteins.
As shown in Figure 7, the protein synthesis-inhibitory activity of rTCS produced in E. coli and purified to ap-parent homogeneity, is indistinguishable from that of plant-derived TCS.

Example 5 Hydropathy_Index Computation for the Trichosanthin Coding Sequences The SOAP program from IntelliGenetics PC/GENETM software package was used to generate the hydropathicity plot of Figure 10. The SOAP progra~ uses the method of Kyte et al. to plot the hydropathicity of the protein along its sequence. Th~ interval used for the computation was 11 amino acids. In Figure 10, the hydrophobic side of the plot corresponds to the positive values range and the hydrophiliC side to the negative values range.
The first 23 amino acids of the trichosanthin sequence are as follows:
Met Ile Arg Phe I-eu Val Leu Ser Leu Leu Ile Leu Thr Leu Phe Leu Thr Thr Pro Ala Val Glu Gly.

,, . ' ' 44 ~ 3~
As can be seen from Figure 10, this sequence has a high degree of hydrophobicity. The length of the sequence and the degree of hydrophobicity make the above sequence an ideal candidate for a secretory signal sequence.

Example 6 Structure-Function Studies of ~-Trichosanthin The protein obtained from the above described alpha-trichosanthin encoding nucleic acid sequence isolated from the genome of T. kirilowii contained two conservative amino acid changes in the alpha-TCS protein coding sequence relative to the primary sequence which was determined for alpha-TCS. This example describes the construction of a nucleic acid coding sequence for a synthetic gene alpha-trichosanthin gene based on codons selected to represent the primary amino acid sequence of the alpha-trich~santhin protein isolated from plant material. The example further describes the creation of mutant alleles of the gene encoding ~-trichosanthin.

(i) Sources of Plasmids, Bacterial Stralns, Reagents and EnzymeS
Bluescripts~ was purchased from Stratagene ~La Jolla, CA); pACYC184 ~ATCC no. 37033) was obtained from the American Type Culture Collection, ~ockville, Maryland.
The expression plasmid, pKK233-2, was obtained from Pharmacia ~Piscataway, NJ). The Escherichia coli strains used for transformation include DH5 ~F-, recAl, endAl, hsdR1 7 (r,m,~ , sup~344, ~-, thi-l, gyrA, relAl), from BRL
(Gaithersburg, MD) and XLl-Blue (recA1, endA1, gyrA96, thi, hsdR17 (r~m~), supE44, relA1, ~~, lac~, (F', proAB, lacrZ~M15, l'nlO ~tetrJ) from Stratagene. JM105 (thi, rpsL, endA, sbcBl5, hsdR4, ~ c-proAB), (F', traD36, ,, ," ~ '", ' ~

.
,, . - ~ .

..

2~i`5 ~ 3 ~ 3 proAB, lacIqZ~M15)) was used for expression of the synthetic gene. LB medium (Maniatis et al.) was used for the propagatiOn of bacteria. Expression was performed in M9 medium ~Maniatis et al.) containing supplements of glucose or glycerol (0.2%) thiamine (2~g/ml), casamino acids (CAA) (0.3~), IPTG (5 mM), and ampicillin (100~g/ml).
Restriction endonucleases were purchased from either Bethesda Research Laboratories (Gaithersburg, MD) or Promega Biotec tMadison, WI); the enzymes were used according to the manufacturers' recommendationS. T4 DNA
ligase was purchased from International 3iotechnologies, Inc. (IBI, New Haven, CT).
(ii) Vector Construction An 80 base pair lbp) polylinker was constructed which contained the restriction sites KpnI, HindIII, NsiI, SpeI, PstI, BssHII, MluI, SalI, AatII, and SacI, in the order found in the synthetic gene (Figure 11). These sites were used to clone and assemble the individual synthetic nucleic acid fragments into a contiguous sequence. One end of the polylinker contained a BanI
overhang and the other end contained an XbaI overhang to facilitate cloning into pACYC184.
A chloramphenicol-resistant (CmR) vector, pPS200, was constructed from ligation of the synthetic polylinker to a 1900 bp BanI-XbaI fragment of pACYC184 containing the chlorampheniCol acetyltransferase gene. An ampicillin-resistant (Ap) vector, pPS300, was constructed by replaGing the polylinker present in pBluescript~ with the above-deS~ribed synthetic polylinker. The relevance of using these particular vectors is that they contain compatible replicons; thus, both vectors can be stably maintained in the same cell. The compatible replicons used in construction of the pl~mids were taken advantage :
.

~ . . ;
.

' 2 ~ 9 ~J ~ b ~ ~

of in constructing pPS300 and in condensing the separate, synthetic DNA fragments into the complete gene as described below.
Construction of pPS300 involved digesting both pBluescript~ and pPS200 with KpnI, ligating the linear plasmids, transforming competent E. coli DH5 cells with the ligation mixture, and selecting doubly resistant clones on LB plates containing both ampicillin and chloramphenicol. The plasmids were isolated from positive clones and digested with SacI to determine the orientation of the ligated plasmids and to generate two new plasmids in which the polylinker sequences were exchanged. Those in which the two vector components were ligated in the correct orientation (Figure 12A), i.e., with the two polylinker sequences essentially opposite each other, not juxtaposed (Figure 12B), were selected.
The SacI digest of one of the selected plasmids was diluted and the individual components circularized by ligation. Upon transformation into competent DH5 cells, cells carrying individual plasmids were selected as being Ap~Cms or CmRAps (ampicillin-resistant/chlorampheni sensitive or chloramphenicol-resistant/ampicillin sensltive). The resulting plasmids from these manipulations are a pBluescript~ vector carrying the synthetic polylinker (pPS300) and a CmR plasmid carrying the pBluescript~ polylinker.

(iii~ Fragment Cloning Synthetic gene fragments corresponding to the sequences shown in Figure 11 were synthesized on an ABI 380B DNA
synthesizer using cyanoethyl phosphoramidite chemistry and were purified tritylated on a PRP-3 column purchased from Hamilton Company.

, ComplementarY synthetic oligonucleotides were mixed at 10 picomoles per ~1 in 10mM Tris, pH 8; 10 mM MgCl2; 10 mM NaCl. These mixtures were heated to 100~C for 1 minute, and then allowed to cool slowly to roo~
temperature for annealing.
Successful ligation of the synthetic ollgonucleotide fragments into the polylinker required a 20-40X molar excess of the annealed oligonucleotides, in a final DNA
concentration of 100 ng per ~1~ Ligation reactions were carried out at room temperature for two hours or at 16 degrees overnight ~Maniatis et al.). The ligation mixes were used to directly transform competent DH5 or XLl-Blue cells. Transformants were selected on LB containing either ampicillin or chloramphenicol. Plasmids obtained from the selected transformants were screened for presence of insert by one of two methods: digestion with a restrictiOn enzyme having a cleavage site contained in the insert fragment; or, comparison of the fragment sizes generated from digestion with restriction enzymes having sites which lie outside of the insert. Plasmids containing candidate inserts were checked further for mutations or deletions by nucleic acid sequencing of the insert DNA.
Sequencing of each cloned synthetic fragment was performed on double-strandad plasmid DNA using Pharmacia T7 (Pharmacia) and USB (U S Biochemicals, Cleveland, OH) Sequenase~ Kits. Reactions were performed essentially according to the manufacturers' recommendations, with the following cha~gesO plasmid denaturation was performed at 85C instead of at room temperature, and the samples contained in the microtiter plates were chilled on ice prior to the addition of the labeling reactions, instead of pre-heatin~ to 37 C.
. _ ..
~; .

Most fragments had few or no deletions. However, the BssHII-MluI segment, when cloned into pPS300, was either in the orientation opposite to what was desired for condensation (~ee below) or the sequence had modifications and/or deletions. To get around this problem the fragment was cloned into pPS200. Plasmids containing the insert were identified by screenin~, and the inserts were sequenced. Four plasmids havlng the correct sequence were isolated.
(iv) Condensatlon of the cloned gene fragments To facilitate the combining of the oligonucleotides into a synthetic sequence corresponding to the trichosanthin protein coding sequence, the synthetic oligonucleotide fragments were cloned separately into each type of vector (pPS200 or pPS300). The plasmid/fragment combinations were chosen such that oligonucleotide fragments which should be adjacent each other in the synthetic gene could be combined by ligation into a hybrid plasmid which could be selected by transformation and replica plating from LB-ampicillin plates to LB-chloramphenicol plates or vice versa. This selection follows the method used to transfer the synthetic polylinker from pPS200 to pPS300 described above.
Figure 13 shows a generalized schematic diagram for condensation of cloned gene fragments. The numbers at the ends of-synthetic fragments correspond to restriction enzyme sites. The figure illustrates the joining of fragment 3-4 to fragment 4-5 using the above described plasmids and cutting at site 4 in each plasmid. These single cut plasmids are then ligated and a plasmid having the two sequences joined in the correct orientation is selected. ThiS plasmid is the~cleaved at a restriction 2 ~

site which will generate a plasmid, carrying the joined fragments, which has the desired drug resistance marker (either Cmr or Ap'). The restriction enzyme digest is diluted and the fragments self-ligated to generate the desired plasmid.
The following description of the construction of the TCS synthetic gene can also be read with reference to Figure 13. To join two synthetic gene frayments, each clone was cut with a restriction enzyme which had a cleavage site common to an end of each synthetic fragment, thus producing linear plasmids. These linear plasmids were ligated together, transformed into competent DH5, and selected by their resistance to both chloramphenicol and ampicillin. Plasmids were isolated from several candidates which were both chloramphenicol (CmR) and ampicillin (ApR~ resistant. These plasmids were analyzed using restriction enzyme digestion to determine the orientation of the two synthetic fragments relative to each other. Plasmids having the correct fragment orientations (in which the two gene fragments were joined together) were identified.
Next the joined fragments were segregated to one vector by digesting the plasmid with a restriction enzyme which introduced a single cut at one end of the joined fragmentS and in the fused polylinker sequence present in the combination plasmid (for example cutting at site 3 in the combination plasmid presented in Figure 13). The resulting two plasmid pieces were self-ligated, thus preserving intact polylinker regions, and transformed into bacteria. The plasmid containing the linked fragments was selected on the basis of its drug resist-ance. The choice of whether to segregate the linked fragmentS into a CmR or ApR plasmid depended on the nature of the selection marker ln the plasmid with which ~ :.

, i3 J ~ ~

it was to be next combined. For example, if two fragments are linked and propagated in a CmR plasmid they are joined in the next round of combining with two fragments which were linked cmd propagated ln an ApR
plasmid. By cloning adjacent fragments and pairs of fragments into vectors with clifferent antibiotic resistance, the trichosanthin gene was readily condensed step-wise from eight clones clown to one clone.
Figure 14 shows a schematic of the steps taken to clone and condense the synthetic gene for a-TCS. A linear depiction of the synthetic gene and the restriction sites flanking the individual fragments is shown at the top center. These flanking restriction sites, shown in Figure 14, correspond to the restriction sltes shown in bold print in Figure ll. Throughout the remainder of the Figure the relevant restriction sites are also referred to in one letter abbr~viation where appropriate and con-venient. Also shown are unique restriCtiOn sites contained within the synthetic fragmentS that were used in screening for positive subclones. The two vectors employed, pPS200 ~CmR) and pPS300 (Ap~) are depicted at the top left and the top right of the figure, respectively. Figure 14 shows a flow chart that depicts the cloning of the individual synthetic fragments into either pPS200 or pPS300 followed by the ligation of two subclones into a single plasmid in which the two gene pieces are properly joined. In the third row from the top, examples of the two alternative orientations in which ~he subclones could be ligated are shown along with the selection of the correct orientation for further ligation reactions. For brevity, the self ligations to eliminate one vector component from the doubly-resistant plasmids, leaving the fused gene sequences as part of the second component, are not show~ The final gene fusion .. . : ' ~ .
, . . .
: .

~ .

Sl ~9 ~ 3 is shown as being placed in pPS200 (C~R) although it was also placed in pPS300 (ApR) using alternative restriction digestion and ligation to eliminate the pPS200 component.

(v) Expression of synthetic trichosanthin in E. coli After construction of the synthetic trichosanthin gene was completed, the coding region of the gene was cloned into the expression vector pKK233-2 to give a construct similar to that described previously for the genomic clone. An NcoI to SacI (Klenow repaired to blunt) gene fragment from the pPS200 subclone was placed into pKK233-2 from NcoI to HindIII (Klenow repaired to blunt) employing the two plasmid cloning approach described above. It should also be appreciated that other, more standard methods might also be used to subclone the fragment into pKX233-2 (Maniatis et al.). The resulting plasmid, designated pKQS, was initially transformed into the E. coli strain XL-1 Blue, described above. It was also transformed into another strain, JM105, which also carries lacIq for the regulation of the promoter in pRK233-2 from which the synthetic TCS coding sequences were expressed. The latter transformantS exhibited better growth characteristics than the XL-l Blue transformants and were selected for expression and production of protein.
To express the synthetic gene, pKQS/JM105 transformants were grown in M9 medium (Maniatis et al.) supplemented with 0.2% glucose, (0.2%) thiamine (2~g/ml), casamino acids (CAA~ (0.3%), IPTG (5 mM), and ampicillin (lOO~g/ml) at 37C to saturation density.

(vi) Construction of trichosanthin mutants In order to investigate the function of a region of trichosanthin suspected to have~catalytic activity : ~

-~\

against eukaryotic ribosomes, mutants of the trichosanthin coding sequence were constructed ln vitro.
First, a double mutant was made which changed the glutamic acid at position l.60 and the arginine at ~ 5 position 163 to an aspartic acid and a lysine, respectively. These mutations were made by replacing the BssHII-MluI fragment of the synthetic gene (Figures 11 and 14), in. pKQS, with a fragment containing the two codon changes and a diagnostic restriction site (SpeI).
A SpeI site was included to facilitate screening of insertion of the new fragment. The new sequence was verified by DNA sequencing.

Glu160 Argl63 GAA CGC
BssH I S~e I
CGCGCTCATGGTTTTGATTCAAAGTACTAGTGACGCTGCAAAATACAAATTCATCG~ACA
GC
GAGTACCAAAACTAAGTTTCATGATCACTGCGACGTTTTATGTTTAAGTAGCTTGTCG
Asp Lys AAATTGGCAAA
TTTAACCGTTTGCGC.
Mlu I -Plasmids containing the mutant coding sequences were screened for the presence, orientation, and number of mutant fragment inserts. A clone which contained the correct mutations, DK12, was expressed as described above for pKQS.
In order to determine the contribution of each..amlno acid mutation to the overall effec~ of the double mutant, single mutants were constructed in which either the glutamic acid to aspartic acid substitution or the arginine to lysine substitution was made; these mutant coding se~lences were named D1 and K10, respectively.

(viiJ Purificatio of mutant proteins , ~

. ~ . . , . . : .
. . - :

Overnight cultures were harvested by centrifugation and cells were resuspended in ice-cold 100 mM Tris, pH 8.5, 5 mM EDTA). The cells were disrupted by sonication:
insoluble materials and unbroken cells were then removed by centrifugation at 10,000 X g for 10 minutes. The supernatant was diluted with three volumes of water, and the conductivit~ was measured to ensure that it was less than 2 mmho. The pH of the sample was readjusted to 8.5 with dilute NaOH, and the solution was recentrifuged at 10,000 x g to remo~e any insoluble proteins, cell debris, and additional bacteria.
A column consisting of A50 QAE Sephadex ion exchange resin ~Pharmacia) was activated with 0.1 N HC1 followed by l M NaCl, and was washed extensively with 10 mM Tris HCl, pH 8.5, to equilibrate The diluted sample was loaded onto the column at a rate of 2~3 ml per minute and the flow-through material was collected. The pH of the flow-through was then adjusted to 6.0 with dilute HCl, and the solution was centrifuged at 10,000 x g to pellet any insoluble proteins. This flow-through material was then loaded onto a C25 or C50 SP Sephadex column which had been activated with 0.1 N NaOH and 1 M NaCl, and equilibrated with 20 mM phosphate buffer at pH 6Ø
After addition of the sample, the column was washed extensively with 20 mM phosphate buffer at pH 6Ø
The mut~nt trichosanthin pro~ein was eluted with 20 mM
pho~phate buffer containing 200 mM NaC1, or with standard phosphate buffered saline (Gibco) (140 mM NaCl). Eluted samples were sterilized ~y filtration through a 0.45 low-protein-binding filter (Millex HV, Millipore, Bedford, MA) in preparatiOn for in vitro translation and HIV inhibition assays. Based on SDS gel electrophoresis the homogeneity of the mutant trichosanthin proteins was - ' ' :

typically about 95%. Protei~ yield was typlcally 0.5-1.0 mg/L of cell culture, ~viii) In vitro translation and HIV inhibition assays ~ 5 The isolated mutant prote:ins were compared to isolated unmodified trichosanthin protein (KQS) (also expressed in E. coli), for their ability to inhibit in vitro translation.(IVT) in a rabbit reticulocyte system and to reduce production of p24 antigen in HIV-1 infected T-cells ~assays described in Example 4 above and ir, U.S~
Patent No. 4,869,903). The double-mutant protein ~DK12) was found to be almost 3 logs less active at inhibiting in vitro translation (Figure 15) and more than one log less active at inhibiting p24 production, compared to the unmodified protein. The singly-modified variants showed intermediate activities relative to DK12 and KQS for both inhibition of IVT (Figure 15) and production of p24 antigen.

20Ex~ e 7 Isolation of Additional Sequences Homolo~ous to_al~
.
Trichosanthln This example describes the cloning of additional coding sequences from Trichosanthes kirilowii which have homology to the alpha-trichosanthin nucleic acid coding sequence.
A second genomic DNA library was generated in the lambda ZAPII vector system (Stratagene) as a SpeI
restricted bank; genomic DNA was prepared as described abo~e in Example 2. Spe~ digested genomic DNA ~rom T
kirilowii was ligated to commercially available arms of lambda ZAPII (Stratagene, La Jolla, CA). This ligated mixture was packaged using Gigapack gold~ (mcr-;
Stratagene) and plated using NM554 (mcr-; available from , 2 ~ ~ 3 Stratagene) host. The entire packaging reaction was plated. Following plaque formation, DNA was transferred directly to nitrocellulose filters for probing; the bank was not amplified prior to filter lifts to avoid altering - 5 the representation of phage clones resulting from non-uniform plaque growth. The genomic DNA bank size was about 1.5 x 10~. The nitrocellulose filters were probed with the EcoRI-EcoRI insert of pQ30E (ie. sequence shown in Figure 16) which was radiolabelled by random-priming (Boehringer Mannheim). The filters were washed using 2 x SSC at 50C (i.e., under low stringency conditions).
Twenty-two clones demonstrated a positive hybridization reaction through tertiary plaques and reprobing. Of these, ~1 phage clones were rescued as pBluescript~M
~Stratagene, per manufacturer's instructions) plasmid clones. The rescued plasmids were isolated and their restriction enzyme digestion patterns analyzed using EcoRI, SpeI, SacI~ SalI, NcoI, ClaI and several double-digestion combinations of these enzymes. Out of the 21 plasmids, 11 different restriction enzyme digestion patterns were obser~ed.
- Preliminary double-stranded nucleic acid sequencing reactions were performed on the inserts of representative plasmids from each of the 11 groups, using the following primers: #477 - CGATACATCCTATTTTTTCAACG, derived from the insert of pQ21D;
#624 - CATCTCTGAGGAACAATC and #1546 - CTTATATcATTATGAcTccAAAG~ derived from the insert of pQ30E; and #3AS - TTGTATCTCCATGACTCCAAAGC, #3BAS - CAAGGTGG~AATGGCACTGCTC, and #3CAS - CCTCGAAGCCTCAGCAGTAGT, derived from a new clone, clone #3. The last three primers were obtained from sequences determined for clone ~3 using primer #477.

: : .

:
:

Ten of the 11 plasmid inserts yielded sequence data: 5 inserts appeared to give significantly different sequences and were selected for detailed sequence analysis.
- 5 Sequencing was achieved using the above-described primers and, further, prime!rs were generated based on the sequence data as it was obtained. The DNA sequences encoding the TCS-like protein and some flanking sequences were determined for each clone. The final results indicated that insert 24 ~Figure 17) and insert ~0 were siblings, and that insert 24 ~Figure 17) and insert 2 ~Figure 18) were likely derived ~rom alleles hecause the coding sequences were identical but the 3' flanking sequences contained a few nucleotide differences. Insert 3 (Figure 19) an~ insert 12 ~Figure 20) were unique and encode unique proteins. Figures 17 to 20 show the protein coding strand of the DNA sequence determined for each unique clone. The nucleic acids are numbered above the sequence ~or reference. Unique reStriCtiOn sites which occur in the sequences are overlined above the sequence and indicated by the restriction enzyme name.
The clones are further distinguishable and recognizable by different restriction enzyme digestion profiles in the coding regions.
Figure 16 illustrates the translation product encoded by the insert of pQ30E. Figures 17 to 20 illustrate the translation producks encoded by the sequences that are the putative pre-pro-proteins homologous to alpha-trichosanthin. The amino acids are numbered below the protein sequence for convenience. The putative secretory signal peptides are indicated by negative numbers Isee below). The secreted pro-proteins are labeled with positive n~unbers.
_ . .

. ..

g,~

An alignment of the protein sequences corresponding to the nucleic acid sequences of inserts 2, 3, 12, and of 21D (alpha-trichosanthin) and 30 (the sequence of the partial clone referred to i;n Example 2) is presented in Figure 21. The proteins have been placed into 3 major groups based on their homologies: 21 / 12; 2 / 30; and 3. With reference to Figure 21: ~1) the amino acid residues which differ within a subgroup are shaded; and, (2) the putative secretory signal peptide sequences and putative carboxy terminal extension peptides are underlined. The putative secretory and carboxy extension sequences for inserts 2, 3, 12, and 30 sequences are inferred by alignment of their transla~ed pr~tein sequence with the translated protein sequence of clone 21D (alpha-trichosanthin), and from further comparison to the known mature protein sequence for TCS. While the putative signal peptides are very homologous for the proteins being compared, the carboxy extensions differ significantly in their sequences.
Figure 22 also illustrates a comparison of the proteins encoded by the above-described RIP gene family of T.
kirilowii. This figure uses as a reference the protein coding sequence of the insert from clone pQ2. The other 4 known RIP peptide sequences are aligned as described above for Figure 21: the vertical columns represent possible amino acid substitutions at each site based on this sequence alignment. Sites where amino acid omissions occur are indicated by an asterisk (*). The hyphens are used merely to mark spaces and to allow for easier column alignment.
Although the inYention has been described with refer-ence to specific methods and compositions, it will be ap-parent to one skilled in the art how various modificationS and applications_~f the methods may be-made .

.,1 ?J

without departing from the invention.

: . :

Claims (24)

IT IS CLAIMED:

1. A cloned nucleic acid which encodes a trichosan-thin protein having the ability to inhibit human immunodeficiency virus (HIV) antigen expression in HIV-infected T cells or monocyte/macrophage.

2. The nucleic acid of claim 1, wherein said nucleic acid molecule is included in the sequence:

where basepairs 409 to 1149 encode the mature form of tri-chosanthin isolated from Trichosanthes kirilowii.

3. The nucleic acid of claim 2, which is derived from basepairs 409 to 1149 and which encodes mature tricho-santhin from Trichosanthes kirilowii.

4. The nucleic acid of claim 3, which further in-cludes sequences selected from the group consisting of: (a) basepairs 340 to 408, which encodes an amino terminal exten-sion not present in the mature form of TCS; and (b) base-pairs 1150-1206, which encodes a carboxyl terminal extension not present in the mature form of TCS.

5. The nucleic acid of claim l, extending from base-pairs 409 to 1149 and further including an expression vector containing a promoter, a ribosome binding site, and an ATG
start codon positioned before and adjacent basepair 409, and a stop codon positioned after and adjacent basepair 1149.

6. An expression vector comprising (a) a nucleic acid which encodes a trichosanthin protein having the ability to inhibit human immunodeficiency virus (HIV) antigen expression in HIV-infected T cells or monocyte/macrophage; and (b) a promoter, a ribosome binding site, and an ATG start codon positioned before and adjacent the amino-terminal codon at pOSition 409, and a stop codon positioned after and adjacent the carboxy terminal codon at position after and adjacent the carboxy terminal codon at position 1149.

7. The expression vector of claim 6, wherein said nucleic acid is included in the sequence:

which encodes the mature form of trichosanthin isolated from Trichosanthes kirilowii.

8. The expression vector of claim 6, wherein the expression vector is an E. coli expression vector, and the coding sequence is joined to the ribosome binding site in the vector at a NcoI site which includes the ATG start codon, and is joined at its opposite end at a HindIII site in the vector.

9. For use in obtaining a selectively amplified genomic fragment coding for first and second regions of trichosanthin protein from Trichosanthes kirilowii, by repeated primer-initiated strand extension, a primer mixture comprising a sense-strand set of degenerate primers containing at least one primer species which is effective to hybridize with the anti-sense strand of a genomic fragment containing the coding sequences for said first trichosanthin region, and an anti-sense set of degenerate primers containing at least one primer species which is effective to hybridize with the sense strand of a genomic fragment containing the coding sequence for said second trichosanthin region.

10. The primer mixture of claim 9, wherein (i) said first and second trichosanthin protein regions are homolo-gous in amino acid sequence to first and second spaced regions, respectively, in plant ribosome inhibitor proteins and (ii) said primer mixture can also be used to selectively amplify genomic fragment coding for said ribosome inhibitor proteins.

11. A method of selectively amplifying a genomic fragment including a region coding for first and second regions of trichosanthin from Trichosanthes kirilowii, comprising obtaining genomic DNA from T. kirilowii tissue, providing a sense-strand set of degenerate primers containing at least one primer species which is effective to hybridize with the anti-sense strand of a genomic fragment containing the coding sequence for said first trichosanthin region, providing an anti-sense set of degenerate primers containing at least one primer species which is effective to hybridize with the sense strand of a genomic fragment containing the coding sequence for said second trichosanthin region, preparing an amplification mixture containing said genomic fragment, the two sets of primers, all four deoxynu-cleoside triphosphates, and a DNA polymerase, and selectively amplifying the genomic fragment contain-ing the coding sequences of both coding regions, by repeated primer-initiated strand extension of the amplification mixture.

12. A method for obtaining. genomic coding sequences for a plant ribosome inhibiting protein comprising obtaining genomic DNA from a plant source which pro-duces such ribosome inactivating protein, providing a sense-strand set of degenerate primers containing at least one primer species which is effective to hybridize to a DNA sequence coding for a region of tricho-santhin isolated from Trichosanthes kirilowii which is homologous in amino acid sequence to one region of such ribosome inactivating protein, providing an antisense-strand set of degenerate pri-mers containing at least one primer species which is effec-tive to hybridize to a DNA sequence coding for a second region of trichosanthin isolated from Trichosanthes kirilowii which is homologous in amino acid sequence to a second region of such ribosome inactivating protein, preparing an amplification mixture containing said genomic fragments, the two sets of primers, all four deoxy-nucleoside triphosphates, and a DNA polymerase, and selectively amplifying the genomic fragment containing the coding sequences of said ribosome inactivating protein, by repeated primer-initiated strand extension of the ampli-fication mixture.

13. A trichosanthin composition (a) having the ability to inhibit human immuno-deficiency virus (HIV) antigen expression in HIV-in-fected T cells or monocyte/macrophages;
(b) containing a recombinantly produced tricho-santhin protein corresponding in sequence with alpha-trichosanthin obtained from Trichosanthes kirilowii;
and, (c) being free of contaminants associated with trichosanthin as isolated from Trichosanthes kirilowii plant material.

14. The recombinantly produced protein of claim 13, which includes the sequence:

15. The recombinantly produced protein of claim 14, which further includes, attached to the amino terminus of the sequence, a soluble CD4 peptide, which shows specific binding to the HIV-related gp120 antigen present on the surface of HIV-infected cells.

16. The recombinantly produced protein of claim 14, which further includes, attached to the carboxy terminus of the sequence, a soluble CD4 peptide, which shows specific binding to the HIV-related gp120 antigen present on the surface of HIV-infected cells.

17. The recombinantly produced protein of claim 14, which further includes a carboxy terminal extension having the sequence Ala Met Asp Asp Asp Val Pro Met Thr Gln Ser Phe Gly Cys Gly Ser Tyr Ala Ile.

18. The recombinant protein of claim 14, which fur-ther includes an amino terminal extension having the se-quence Met Ile Arg Phe Leu Val Leu Ser Leu Leu Ile Leu Thr Leu Phe Leu Thr Thr Pro Ala Yal Glu Gly.

19. A trichosanthin composition (a) having the ability to inhibit human immuno-deficiency virus (HIV) antigen expression in HIV-in-fected T cells or monocyte/macrophages;
(b) containing a recombinantly produced tricho-santhin protein encoded by a nucleic acid included in the sequence:

; and, (c) being free of contaminants associated with tri-chosanthin as isolated from Trichosanthes kirilowii plant material.

20. A recombinant process for the production of a trichosanthin protein having the functional properties of Trichosanthes-obtained trichosanthin comprising inserting a DNA sequence encoding said protein into an expression vector, transforming a suitable host with the vector, and isolating the recombinant protein expressed by the vector.

21. The process of claim 20, where the expression vector is pQR19 and the host is E. coli.

22. The process of claim 20, wherein said DNA sequence is included in the sequence:

which encodes the mature form of trichosanthin isolated from Trichosanthes kirilowii.

23. A cloned nucleic acid having a nucleotide sequence encoding a protein sequence selected from the group defined by the following amino acid sequence:

where the amino acids in columns are alternative selections for that amino acid residue and an asterisk represents the option of omitting an amino acid at that ' residue.

24. A cloned nucleic acid having a nucleotide sequence encoding a protein sequence selected from the group defined by the following amino acid sequence:

where the amino acids in columns are alternative selections for that amino acid residue and an asterisk represents the option of omitting an amino acid at that residue.