WO1994008614A1 - Recombinant alliinase, its preparation and pharmaceutical compositions comprising it - Google Patents

Abstract

A recombinant protein in a substantially pure form having the activity of garlic alliinase, a DNA sequence encoding the protein, an expression vector containing the DNA sequence, host cells transfected with the expression vector, a process for preparing biologically active garlic alliinase, and pharmaceutical compositions and food products comprising the recombinant protein and alliin.

Description

RECOMBINANT ALLIINASE, ITS PREPARATION AND PHARMACEUTICAL COMPOSITIONS COMPRISING IT

FIELD OF THE INVENTION

The present invention relates to recombinant garlic alliinase, DNA sequences coding therefor, expression vectors containing said DNA sequences and host cells transformed by said expression vectors. The invention further concerns methods for the preparation of said recombinant garlic alliinase protein and pharmaceutical preparations and food products containing it, to be used together with alliin, the natural substrate of alliinase.

BACKGROUND OF THE INVENTION

Garlic (Allium sativu ) has been known from ancient times to have many medicinal properties and has been used in popular medicine for a variety of maladies. In the past decades there has been renewed research interest in the therapeutic uses of garlic.

Garlic preparations are commercially available in the form of garlic oil, extracts, pills or tablets. Usually, the preparation procedures are unknown. Some of these garlic preparations were shown to increase the fibrinolytic activity, to inhibit platelet aggregation and to lower the levels of cholesterol in serum (Knipschild, J.K. and Ter- Riet, G., Brit. J. Clin. Pharmacol. 1989, .28.:535-544; Bordia, A. et al. Atherosclerosis 1975, 22.-103-109; Jung, E.M. et al., Arzneim. Forsch. 1991, .41.:626-630; Bordia, A. and Verma, S.A., Artery 1980, 2*428-437; DeBoer, L.W. and Folts, J.D., Am. Heart J. 1989, 117:973-975; Bordia, A. et al. , Atherosclerosis 1977, .28.:155-159; Bordia, A., Am. J. Clin. Nutr. 1981, 3_4:2100-2103; Ernst, E. et al. , Brit. Med. J. 1985, 291:139; Harenberg, J. et al. , Atherosclerosis 1988, T4:247-249; Mader, F.H., Arzneim. Forsch. 1990, 40:1111- 1116). The effective doses which gave such results were in the range equivalent to that of 5 garlic cloves per day. Most of these studies were, however, deficient in their methodological approach and cannot give conclusive indication for the therapeutic effects of garlic or any of its components.

The active principle of garlic is allicin (thio-2- propene-1-sulfinic acid S-allyl ester), the compound responsible for its pungent smell. The precursor of allicin is alliin (S-allyl-L-cysteine sulfoxide) which is converted to allicin by a C-S-lyase present in the garlic plant termed alliin lyase or alliinase [E.C. 4.4.1.4.]. Alliin and alliinase are found in different compartments of the garlic plant. Upon mixture they yield allicin according to the following reaction :

alliin

allicin pyruvate

Allicin has been shown to have a variety of biological activities such as antifungal, antiamebic, antibacterial and for lowering cholesterol (Barone, F.E. and Tansey, M.R., Mycologia, 1977, 6j>.:793-825; Mirelman, D. et al., J. Infect. Dis. 1987, 156:243-244; Cavallito, C.J. et al. , J. Am. Chem. Soc, 1944, 6:1952-1954; Augusti, K.T. and Mathew P.T. , Experientia 1974, 3_0.:468-470) . However, it is a very labile compound which easily decomposes to give an inactive component. In the garlic plant the allicin precursor, alliin, and the enzyme alliinase are stored in different compartments and are brought into contact to give allicin only when the garlic plant is damaged, for example, by bacterial or fungal attack. It would be beneficial to generate allicin in controlled and known amounts for therapeutic purposes.

A possible way of avoiding allicin decomposition prior to use is to maintain its precursor alliin in a different compartment than the degrading enzyme alliinase, e.g., in a separate microcapsule, so that only after administration and the breakdown of the microcapsule wall, the substrate and the enzyme are brought into contact to give allicin. In U.S. Patent No. 4,849,218, garlic preparations are disclosed which comprise an active garlic powder (I) and the enzyme alliinase (II) which are kept separate from each other but become combined after the composition has been taken. Preferably (I) and (II) a e (a) separately microencapsulated or microencapsulated or separately formed into fat pellets, then the 2 types of microcapsules/pellets mixed; (b) formed into separate layers of a two-layer tablet or (c) suspended in a mixture of saturated triglycerides and vegetable oil and then filled into a soft gelatin capsule. German published patent application No. 4012884 Al describes dry garlic extracts containing active alliinase and alliin. The extract contains some of the active alliinase present in the extraction residue. Compared to garlic powder, the extract contains more alliin, based on the amount of alliinase. The dried extract also contains a powdered, water soluble acid, especially in an amount of at least 0.1 wt%. The alliin content of the extract is at least 2 wt% (based on the dry weight) . Alliin- containing garlic extracts which contain no alliinase have no therapeutic value. The alliinase in the dry extracts is activated by addition of water and only then converts the alliin to allicin which is the active antibiotic effective against intestinal disease-causing bacteria.

Alliinase has been isolated and purified by several groups. However, results concerning the purification of alliinase were quite conflicting with regard to the molecular weight of the enzyme and its physical and chemical properties (Jansen, H. et al., Planta Medica 1989, j55_:434-439; Nock, L.P. and Mazelis, M. Plant Physiol. 1987, JL5:1079-1083; Nock, L.P. and Mazelis, M. , Arch. Biochem. Biophys. 1986, 249:27- 33; Kazaryan, R.A. and Goryachenkova, E.V., Biokhimia 1978 .43.:1905-1913) . It is a pyridoxal phosphate dependent enzyme. Another type of alliinase present in onion plants, catalyzes a similar reaction. The onion and garlic alliinase, however, appear to be structurally different enzymes as was determined by the difference in their molecular weight, subunit structure (the onion alliinase is a tetramer and the garlic alliinase is a dimer), Km values, isoelectric pH, amount of glycosylation and spectra, and by the fact that there was no cross reactivity of antibodies raised against the garlic alliinase with the onion alliinase (Nock, L.P. and Mazelis, M. Plant Physiol., 1987 .8_5:1079-1083) . Comparison of the NHa-terminal (23) amino acids of the onion alliinase (Lancaster J. et al., Abstract of the Third International Congress, Molecular Biology and Development, Tucson, Arizona, 6-11 October 1991) and of the garlic alliinase according to the present invention has not revealed any homology between them.

The enzymatic reaction of alliinase has been only partially studied. In the studies related to the present invention, it was found that alliinase is rapidly and irreversibly inactivated by reaction products of the alliin enzymatic conversion. During the purification procedure, alliinase comes into contact with alliin, and the products of the enzymatic reaction inactivate the enzyme. Thus, a large proportion of the alliinase obtained after the purification procedure is not active.

There is a long felt need for means to obtain alliinase in a substantially pure and active form for use in pharmaceutical preparations and in food products. Such a product could be used first to explore its therapeutic potential, to construct a dose-response curve, etc. It is important to have a reliable source of the enzyme, which provides preparations having a well-defined amount of enzyme units in each preparation.

Thus, it is an object of the invention to provide pure and active recombinant garlic alliinase, as well as DNA sequences encoding therefor, in order to overcome .the drawbacks of the prior art.

SUMMARY OF THE INVENTION

The present invention provides, by one of its aspects, a recombinant protein having the activity of garlic alliinase in a substantially pure form.

In accordance with a preferred embodiment of said one aspect, the present invention provides a recombinant protein having the activity of garlic alliinase and having an amino acid sequence selected from the group consisting of : (i) the amino acid sequence depicted in Fig. 12; (ii) the amino acid sequence depicted in Fig. 12, wherein one or more amino acids has been deleted, added, replaced or chemically modified without substantially reducing the garlic alliinase activity; (iii) a fragment of the amino acid sequence of (i) or (ii) which essentially retains the garlic alliinase activity; and (iv) a dimer of (i), (ϋ) or (iii) having garlic alliinase activity. The present invention further provides by another of its aspects, DNA sequences coding for said garlic alliinase.

In accordance with a preferred embodiment of this aspect, the present invention provides a DNA sequence selected from the group consisting of :

(i) the DNA sequence depicted in Fig. 12;

(ii) a fragment of the DNA sequence depicted in Fig.

12, encoding a protein having garlic alliinase activity. (iii) a DNA sequence derived from (i) or (ii) in which one or more amino acid encoding triplets has been added, deleted or replaced without substantially affecting the garlic alliinase activity of the protein encoded thereby;

(iv) a DNA sequence which is a degenerate equivalent of the sequences of (i), (ii), or (iii), and encodes a protein having garlic alliinase activity; and

(v) a DNA sequence hybridizable to a sequence of (i), (ii), (iii) or (iv), and which encodes a protein having garlic alliinase activity.

By a further aspect the present invention provides expression vectors comprising said DNA sequences as well as host cells transfected by such expression vectors.

The present invention still further provides a method for the manufacture of the recombinant protein having the activity of garlic alliinase.

Still further, the present invention provides pharmaceutical compositions and food products comprising, as an active ingredient, said recombinant alliinase to be administered together with alliin, in a form which does not permit an enzymatic reaction, in which alliin is converted to allicin, until after administration.

BRIEF DESCRIPTION OF THE DRAWINGS

Fig. 1 shows alliinase activity in extracts of different parts of the garlic plant : leaves, stems and cloves.

Fig. 2 shows results of garlic alliinase purification by HPLC hydrophobic interaction chromatography : protein absorbance at 280 nm (solid line) and alliinase activity in different fractions (dotted line) . Fig. 3 shows pH dependence of garlic alliinase reaction rate.

Fig. 4 shows absorption spectrum of purified garlic alliinase. 1.6 mg/ml, specific activity 130 units/mg of protein, 0.02M Na phosphate buffer, containing 10% glycerol, pH 7.4.

Fig. 5 illustrates the determination of molecular weight of garlic alliinase by HPLC size exclusion chromatography (full circle) and comparison to known proteins (empty squares) : (l)β-amylase (200 kDa); (2) alcohol dehydrogenase (150 kDa); (3) bovine serum albumin (67 kDa); (4) ovalbumin (45 kDa); (5) ribonuclease (13.7 kDa).

Fig. 6 depicts SDS-PAGE of garlic alliinase. Lane A- crude garlic extract; Lane B-purified alliinase after hydroxylapatite step; Lane C-purified alliinase after HPLC hydrophobic interaction chromatography step; Lane D- molecular weight markers.

Fig. 7 depicts SDS-PAGE of garlic alliinase before (Lane 1) and after cleavage with CNBr (Lane 2). Molecular weight markers are shown in lane 3 and before lane 1.

Fig. 8 shows HPLC analysis of a sample (20 μl) of a mixture of alliin (25 mg/ml) and purified garlic alliinase (50 units/ml) after incubation for 20 min and dilution in 60% methanol (1:200).

Fig. 9 shows the degenerate mixed oligonucleotides (Mix 1 and 2) corresponding to the N- erminal amino acid sequence 10 to 15 of garlic alliinase (Sequence I), used for the MOPAC reaction. The inclusion of the next two nucleotides from the seventh amino acid is indicated by asterisks. ("I" indicates the nucleotide derivative Inosine) . An Xbal restriction site (underlined) was also added to facilitate further manipulation of the amplified cDNA.

Fig. 10 shows the nucleotide sequence and the open reading frame (ORF) of the amplified garlic alliinase cDNA fragment termed pAli-1. The underlined nucleotides indicate the amino acid sequence 10 to 15 used for the synthesis of the mixed oligonucleotides (Mix 1 and 2 of Fig. 9) and the dots indicate a 9 amino acid sequence of the ORF product corresponding to the NH₌-terminal amino acid sequence of garlic alliinase.

Fig. 11 is a schematic representation of the construc¬ tion of the nearly full length alliinase cDNA, pSPAli.

Fig. 12 shows the full length cDNA nucleotide sequence and the deduced amino acid sequence of garlic alliinase, as is described in Examples 3.5, 3.6 and 3.7.

Fig. 13 shows the computer predictions from the nucleotide and amino acid sequences (Fig. 12) of alliinase, as described in Examples 3.5, 3.6 and 3.7.

Fig. 14 depicts the Northern blot analysis of alliinase mRNA in different tissues of the garlic plant.

Fig. 15 shows the results of SDS-PAGE followed by Western blot analysis of alliinase mRNA transcribed in vitro from pSPAli and subsequently translated in vivo in Xenopus oocytes, as described in Example 5.

Fig. 16 shows serum cholesterol levels in cholesterol- fed rabbits treated with allicin (filled squares). Control (empty squares) .

DETAILED DESCRIPTION OF THE INVENTION

As used herein, "a protein having garlic alliinase activity" refers to a protein which is capable of converting alliin and derivatives thereof to allicin and pyruvic acid. It refers to a protein having the amino acid sequence depicted in Fig. 12, which corresponds to the natural garlic alliinase subunit, to modifications and fragments thereof, and to dimers of said protein and of said modifications and fragments. The dimer of the protein of Fig. 12 corresponds to the holoenzyme, i.e., the natural garlic alliinase protein. Examples of such modifications which do not essentially alter the garlic alliinase activity of the proteins include the following : deleting one or more amino acids from either or both terminals, or from a region in the protein which does not form part of its active site; adding one or more amino acids to either terminals; replacing one or more amino acids •by other amino acids of similar polarity which are functional equivalents, resulting in a silent alteration; chemically modifying amino acids of the protein, e.g., by glycosylation, methylation, esterification, addition of fatty acids, etc. Substitutes for an amino acid within the sequence may be selected from a member of the class to which the amino acid belongs. For example, the nonpolar (hydrophobic) amino acids include alanine, leucine, isoleucine, valine, proline, phenylalanine, tryptophan and methionine. The polar neutral amino acids include glycine, serine, threonine, cysteine, tyrosine, asparagine and glutamine. The positively charged (basic) amino acids include arginine, lysine and histidine. The negatively charged (acidic) amino acids include aspartic acid and glutamic acid. The invention also concerns di ers and fragments of the above mentioned proteins and modifications thereof which retain the garlic alliinase activity and which can easily be prepared by the artisan without the need for undue experimentation.

The DNA sequences of the invention encode the protein of Fig. 12, modifications and fragments thereof. The DNA sequences of the invention may comprise additional sequences. For example, promoter sequences suitable for the host in which the DNA is to be expressed, initiation signals (which are important in cases where only a portion of the alliinase gene is used), enhancer sequences, sequences which can confer expression of the alliinase gene only upon induction, sequences which code for secretory peptides , etc. The additional sequences may be of natural or synthetic origins. An example of such a DNA sequence is the sequence depicted in Fig. 12 or degenerate equivalents thereof.

Any species of the genus Allium as well as a subspecies or a variety thereof may serve as source for the molecular cloning of the alliinase gene. The DNA may be obtained by standard procedures from cloned DNA, e.g., a DNA library, by chemical synthesis, by cDNA cloning or by the cloning of genomic DNA, or DNA fragments thereof, purified from the desired Allium plant tissues (see for example Sambrook et al., 1989, Molecular Cloning : A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Spring Harbor, New York; Ausubel et al., 1987, Current Protocols in Molecular Biology, John Wiley & Sons). Clones derived from genomic DNA may contain regulatory sequences and introns, in addition to coding sequences. Clones derived from cDNA will contain only exon sequences and a poly(A) tail.

Identification of the specific DNA fragment containing the alliinase gene can be performed in a number of ways. For example, an amount of the alliinase gene, or its corresponding RNA or a fragment thereof obtained by chemical synthesis or by PCR (polymerase chain reaction) is purified and labeled, and the generated DNA fragments are screened by nucleic acid hybridization to the labeled probe (Benton, W.D. and R.W. David, 1977, Science, 196:180: Grunstein, M. and D. Hogness, 1975, Proc. Natl. Acad. Sci. USA 72:3961) , or a cDNA expression library is immunoscreened with specific antibodies against alliinase (Mierendorf, et al., 1987, Metho. Enzymol. 152:458) . In another method, mRNA fractions enriched in alliinase may be used as a probe as an intial selection procedure. It is also possible to identify the required DNA fragment by restriction enzyme digestion and comparison of fragment sizes with those expected according to a known restriction map. Further selection based on the properties of the gene, or the physical, chemical or immunological properties of its expressed product, can be employed after the initial selection.

The alliinase gene can also be identified by mRNA selection by nucleic acid hybridization followed by in vitro (for example, rabbit reticulocyte lysate or wheat germ extract) or in vivo (for example, Xenopus lavies oocytes) translations. Immunoprecipitation analysis or functional assays can be used for identification of the cDNA clones containing the alliinase sequences. In addition, specific mRNAs may be isolated by adsorption of polysomes isolated from specific tissues to immobilized antibodies specifically directed against the alliinase protein. A radiolabeled mRNA or cDNA can be synthesized from the selected mRNA (of the adsorbed polysomes) , and can then be used as a probe to isolate alliinase DNA fragments from either genomic or cDNA libraries.

In a particular embodiment, the alliinase gene can be obtained from a cDNA library made from poly(A)+ RNAs isolated from developed garlic bulbs (mature bulb contains large amounts of the alliinase protein) which are expected to contain high amounts of the alliinase mRNA.

In another embodiment, an alliinase genomic clone can be obtained from a genomic library after selection by hybridization to an alliinase-homologous nucleic acid sequence. In a particular embodiment, alliinase probes can be derived from the amino acid sequence of the alliinase protein, by chemical synthesis of degenerate oligonucleotides. These oligonucleotides can serve directly for screening either genomic or cDNA libraries.

In still another embodiment, the degenerate oligonucleotides may be used for the generation of a "real" probe by the strategy of MOPAC (mixed oligonucleotide primed amplification of cDNA, Lee et al., 1988, Science 239:1288) . The MOPAC product can then be used for screening either genomic or cDNA libraries. Identification of the MOPAC product can be determined by sequencing analysis by either the Maxam-Gilbert procedure (Maxam, A.M. and Gilbert, W. , 1980, Meth. Enzymol. £5:499) or by the Sanger dideoxy nucleotide chain termination procedure (Sanger et al., 1977, Proc. Natl. Acad. Sci. USA .74.:5463) .

Identification of the alliinase clones obtained from a cDNA library can be accomplished in a number of ways based on the properties of the DNA itself or, alternatively, on the physical, chemical, immunological, or functional properties of its encoded protein.

Once the structure of the alliinase gene is known, it is possible to manipulate the structure for optimal use in the present invention. For example, an omega DNA sequence derived from the coat protein gene of the tobacco mosaic virus (Gallie, et al., 1987, Nucl. Acid Res. .15:3257) and shown to enhance the translation of chimeric mRNA, may be ligated upstream to the start codon, AUG, of the alliinase gene. Many manipulations are possible and within the scope of the present invention.

The identified and isolated gene can then be inserted into an appropriate cloning vector. A large number of known vector-host systems may be used. Possible vectors include, but are not limited to, plasmids or modified viruses, but the vector system must be compatible with the host cell used. Such vectors include, but are not limited to, bacteriophages such as lambda derivatives, or plasmids such as pBR322, pUC or pBluescript^R plasmid. Recombinant molecules can be introduced into the host cells via ^■ transformation, transfection, infection, electroporation, etc.

The nucleotide sequence coding for an alliinase protein, or a portion thereof, can be inserted into an appropriate expression vector, i.e., a vector which contains the necessary elements for the transcription and translation of the inserted coding sequence and expression of the protein end product. The necessary transcriptional and translational signals can also be supplied by the native alliinase gene and/or its flanking regions. A variety of vector systems may be utilized to express the protein-coding sequence. These include, but are not limited to, viruses, bacteriophages, plasmids, and cosmids. The expression elements of these vectors vary in their strength and specificities. Depending on the host-vector system utilized, any one of a number of suitable transcription and translation elements may be used. For instance, when cloning in yeast cell, promoters isolated from the genome of yeast cells (e.g., the unregulated alcohol dehydrogenase promoter or other regulated promoters such as GAL and PGK) may be used. Specific initiation signals are also required for efficient translation of the inserted coding sequences. These signals include the ATG initiation codon and adjacent sequences. In cases where the entire alliinase gene, including its own initiation codon and adjacent sequences, are inserted into the appropriate expression vectors, no additional translational control may be needed. However, in cases where only a portion of the alliinase coding sequence is inserted, exogenous translational control signals, including the ATG initiation codon and adjacent sequences, must be provided. The intiation codon must be in phase with the reading frame of the protein coding sequences to ensure translation of the entire insert. These exogenous translational control and initiation codons can be of a variety of origins, both natural and synthetic.

The host cells to be used in the present invention are both prokaryotic and eukaryotic, e.g., bacterial, yeast, insect, mammalian and plant cells. The host cells are chosen in accordance with the desired mode of expression : for example, if induced expression is desired, bacterial cells are used; if the proteins are to be secreted to the medium, eukaryotic cells are used, etc. The host cells can also be chosen in accordance with desired modifications, such as glycosylation or processings such as cleavage of the protein products. It should be noted that the native alliinase is a glycosylated protein and therefore if a protein which resembles the native protein is desired, an eukaryotic host cell, such as yeast or mammalian cells, should be chosen. If, on the other hand, an unglycosylated protein is desired, a bacterial host cell should be chosen.

The manufacture of the protein of the present invention comprises culturing the said host cell in a suitable culture medium and then isolating the protein from the medium.

Since alliinase is a glycoprotein, it might be convenient to use an expression system which is able to correctly modify the protein. Unlike prokaryotic systems, the eukaryotic subcellular organization of yeast enables them to carry many of post-translational modifications, folding, processing and other events required to produce biologically active proteins. Hence, yeasts are a suitable system for the commercial production of heterologous proteins.

Yeast cells may be transformed by the lithium acetate transformation procedure (Roths ein, R. in DNA Cloning : A Practical Approach., 1985, D.M. Glover, ed. Oxford:IRL Press, p. 45); spheroplast transformation (Hinnen, A. et al., Proc. Natl. Acad. Sci. USA 1978, 25.:1929) or by electroporation (Becker, D. and Guarente, L. , Meth. in Enzymology, 1991, Vol. 194:182-187) .

For example, the DNA encoding alliinase can be inserted into yeast secretion vectors, such as pNH008 and pYN026, downstream from the secretion signal of yeast invertase (August et al., 1991, J. Biochem. Biophys. 1089:345) or the secretion signal of the human serum albumin (HSA) (Okabayashi et al., 1991, J. Biochem. 110:103) . Expression of the alliinase gene can be controlled by regulating promoters such as the galactose inducible GAL7 promoter (August et al., 1991, J. Biochem. Biophys. 1089:345) , or the maltose inducible GAM1 promoter (Piontek et al., 1990 Yeast, 6S:422) . The processed secreted alliinase protein can be obtained from the yeast growth medium by standard procedures of protein purification. For example, commercially available plasmids, such as pYES2 or pYESHis A, B and C (Invitrogene Corporation) comprising an alliinase cDNA can be transfected into Saccharomyces cerevisiae strains, e.g., INVScl and INVsc2 (Invitrogene Corporation), for inducible expression of recombinant alliinase.

Alliinase can be produced in mammalian cells by using the mammalian episomal expression system. The Epstein-Barr virus (EBV) can be maintained extrachromosomally and produce high levels of recombinant proteins in a wide range of mammalian cells. For example, the alliinase cDNA can be inserted into EBV episomal vectors, such as pREP or pMEP, under inducible promoters (such as heavy metal or glucocorticoids) (Groger et al., 1989 Gene 81:285-294) . In a particular example, alliinase cDNA can be inserted into a commercially available plasmid, such as pEBVHis (Invitrogene Corporation). For use in mammalian cells, the control functions on the expression vectors are often provided by viral material. For example, commonly used promoters are derived from polyoma, Adenovirus 2, and most frequently Simian Virus 40 (SV40). The early and late promoters of SV40 virus are particularly useful because both are obtained easily from the virus as a fragment which also contains the SV40 viral origin of replication (Fiers, et al., Nature 1978, 273:113) . Smaller or larger SV40 fragments may also be used, provided there is included the approximately 250 bp sequence extending from the Hind III site toward the Bgll site located in the viral origin of replication. Further, it is also possible, and often desirable, to utilize promoter or control sequences normally associated with the desired gene sequence, provided such control sequences are compatible with the host cell systems. The alliinase cDNA can also be inserted into the pCD vector (Okayama-Berg vector; Okayama and Berg, 1983. Mol. Cell Biol. .3:280-289). Examples of appropriate mammalian host cells are the COS-7 monkey kidney cell lines, the C127 murine mammary epithelial cells, the Chinese Hamster Ovary (CHO) cells, the W138, BHK and MDCK cell lines.

Recombinant alliinase can also be expressed in insect cells by means of a baculovirus expression vector (Summers, M.D. and Smith, G.E., 1987, A Manual of Methods for Baculovirus Vectors and Insect Cell Culture Procedures, Texas Agricultural Experiment Station Bulletin No. 1555). Baculovirus transfer vectors are used for the expression of foreign genes under the control of Autographa californica nuclear polyhedrosis virus (AcMNPV) , the prototype virus of the family Baculoviridae, which has a wide range and infects more than 30 species of Lepidopteran insects. Expression, glycosylation and secretion of β-phaseolin (Phaseolus vulgaris) by means of baculovirus expression vector was previously described (Bustos, M.M. et al., 1988. Plant Molecular Biology .10.:475-488) . Many other foreign genes have been expressed by baculovirus vectors (Luckow, V.A. and Summers, M.D. ,1988 Bio/Technology j5.:47-55). For example, a suitable plasmid, such as pAc373 (Luckow and Summer, cited above) containing alliinase cDNA can be cotransfected with the wild-type AcMNPV DNA into insect cells, such as cultured Spodoptera frugiperda cells (Sf9), to produce recombinant alliinase that is mostly secreted into the culture medium.

Alliinase can also be produced in plants. Transfer of foreign DNA into plant cells is most often performed using plasmid vector systems derived from the tumor-inducing (Ti) plasmid of Agrobacterium tumefaciens. For example, the DNA encoding alliinase protein can be inserted into a binary Ti plasmid under the control of a strong promoter, such as 35S promoter of the cauliflower mosaic virus (CaMV), which is commercially available, and is generally expressed in most, if not all, plant tissues (Guilley et al., 1982 Cell 30:763: Odell et al., 1985 Nature 313:810) . Other promoters that can be used are inducible promoters, such as the light inducible promoter derived from the pea rbcS gene (Coruzzi et al., 1984 EMBO J. 3_:1671), or actin promoter from rice (McElroy et al., 1990 The Plant Cell ^:163). Downstream from the promoter, a short DNA sequence which enhances translation, such as the omega sequence derived from the coat protein gene of the tobacco mosaic virus (Gallie et al., 1987, Nucleic Acids. Res. .15.:3257) may be added. Downstream from the DNA sequence encoding alliinase protein, a terminator DNA sequence containing the 3'-transcription termination and polyadenylation signal derived from the octopine synthase gene of the Ti plasmid of A. tumefaciens may be installed (Greve et al., 1983, J. Mol. Appll Genet. .1:499). The alliinase gene construct can be subcloned into the expression vector pGA492 binary Ti plasmid of A. tumefaciens (An, G. 1986 Plant Physiol. j31.:86-91). The expression vector comprising the alliinase gene is then introduced into plant cells by standard procedure. Plant transformation is based on the cocultivation method (Marton, L. et al., 1979 Nature 277:129-131) in which plant cells, for example Nicotiana tabacum cv. Samsun NN. (Shaul, 0. and Galili, G. 1992 The Plant Journal \203-209) are cocultured with Agrobacterium for about 2 days and transformed plant cells are plated on an appropriate selective medium (Horsch et al., 1985 Science 227:1229-1231; Deblaere et al. 1987 Methods in Enzymology 153:277: An, G. 1987 Methods in Enzymology 153:292).

Prokaryotes may also be used for expression of alliinase. Bacteria, e.g., E. coli, baccilli such as Bacillus subtilus, and other enterobacteriaceae such as Salmonella typhimurium or Serratia marcesans, and various Pseudo onas species may be used. For example, E. coli is typically transformed using pBR322, a plasmid derived from an E. coli species (Bolivar et al., Gene, 1977, 2.:95). pBR322 contains genes for ampicillin and tetracycline resistance and thus provides easy means for identifying transformed cells. The pBR322 plasmid, or other microbial plasmid must also contain, or be modified to contain, promoters which can be used by the microbial organism for expression of its own proteins. Those promoters most commonly used in recombinant DNA construction include the β-lactamase (penicillinase) and lactose promoter systems (Chang et al., Nature, 1978, 275:615; Itakura, et al., Science, 1977, 198:1056: Goeddel, et al.. Nature, 1979, 281:544) and a tryptophan (trp) promoter system (Goeddel et al., Nucleic Acids Res. 1980, 8.:4057).

Transformation of bacterial cells with plasmid containing alliinase DNA may be performed by the calcium chloride procedure (Sambrook et al., cited above).

Pharmaceutical compositions in accordance with the present invention comprise the protein of the invention to be administered together with alliin, and a pharmaceutically acceptable carrier. The composition may comprise the protein of the invention and instructions for administration together with alliin, which is administered in a separate composition, or the protein and alliin may be included together in one composition, in which case the protein and the alliin should be in a form which does not permit a reaction between them prior to administration to the patient. This ensures that allicin, which is the labile enzymatic reaction product, is formed only upon administration and thus has a relatively long period of time for exerting its beneficial activities before it is decomposed. In addition, this ensures that the products of the enzymatic reaction will have the minimal opportunity of inactivating the enzyme. The alliin and protein of the present invention can be in a dry, for example, lyophilized form and are thus allowed to react only after being mixed with a solvent prior to injection, when using parenteral administration, or after being mixed with saliva or gastric fluid, in oral administration. The alliin and protein of the present invention may also be contained in two different compartments, for example, in two different capsules, in different compartments of a compartmentalized capsule, in two different species of microcapsules, in two different species of liposomes, etc. The compartments are broken down by gastric fluids whereby the protein and alliin react, yielding the active compound allicin.

The present invention also provides food products comprising the protein of the present invention and alliin in one of the forms mentioned above for pharmaceutical compositions. This ensures that only upon chewing the food product or upon adding water thereto, allicin is produced to yield the garlic's characteristic aroma and taste and to produce its beneficial action.

Both the pharmaceutical compositions and the food products are for use for humans as well as for animals.

The invention will now be illustrated by the following non-limiting examples.

EXAMPLES Example 1 1.1 Purification of alliinase from garlic cloves

Garlic cloves were chosen for the extraction of alliinase because, as illustrated in Fig. 1, the enzymatic distribution in the different parts of the garlic plant shows that the alliinase activity continuously increases from the leaves to the lower part of the stem by approximately 10 fold, and activity in cloves is 100 times higher than that in the leaves. Peeled garlic cloves (60 g) were homogenized in 90 ml of 0.02M Na-phosphate buffer (pH 7.2) containing 10% ' glycerol, ImM phenylmethylsulfonyl fluoride (PMSF) and 20 μ-M pyridoxal 5'-phosphate (buffer A). The homogenate was filtered through two layers of cheesecloth and the filtrate was centrifuged at 20000 x g for 30 min at 4°C. Polyethylene glycol-8000 (PEG-8000) was added to the supernatant (to 25%), the mixture was stirred slowly for 20 min at 4°C and the slurry was then centrifuged at 20000 x g for 15 min at 4°C The pellet was resuspended in 120 ml of buffer A, centrifuged at 20000 x g for 20 min at 4°C, and the supernatant solution was placed on a hydroxylapatite column (2.2 x 50 cm). The column was washed with 0.05M Na-phosphate buffer pH 7.2 and the peak containing the enzymatic activity was eluted by 0.3M Na-phosphate buffer pH 7.2. The eluate was brought up to 50% saturation with

and stirred slowly for 30 min. The slurry was then centrifuged at 20000 x g for 15 min at 4°C. The pellet was dissolved in about 2 ml of 0.1M K- phosphate buffer pH 7.2, then diluted with the same volume of 2M NaaSO_* and placed on an HPLC hydrophobic interaction chromatography column (7.5 x 75 mm, Ultrapac TSK Phenyl 5PW, LKB, Uppsala, Sweden). Elution of the enzyme was carried out with a gradient of Na₂S0_Λ (from 1M to 0) in 0.1M K- phosphate buffer pH 7.2. Fractions of 1 ml were collected and assayed for protein content by the procedure of Lowry et al. (J. Biol. Chem. 1951, 193:265-273) with ovalbumin as a standard, and for alliinase activity by the enzyme assay described in Example 1.3 hereinafter. Active fractions were combined and used for characterization studies.

The results of purification are shown in Fig. 2. Protein is located by its absorbance (A) at 280 n in the HPLC column eluates (full line). Alliinase activity was found in fractions 10 to 24 with a peak in fraction 15 (dotted line) .

The results of the different steps of the purification procedure are summarized in Table 1. The yield of purified alliinase was approximately 3.2 mg/10 g cloves (average weight of one clove : about 5 g). Total activity is defined as production of total pyruvate (μ.mol/min). Specific activity is defined as total activity/protein content. TABLE 1

In the method of purification, the polyethylene glycol step enabled elimination of most of the balast proteins after only one precipitation step, and the hydroxylapatite step successfully eliminated protein(s) tightly associated with alliinase. High efficacy and reproducibility was achieved by the HPLC hydrophobic interaction chromatography step, producing homogeneous alliinase, shown to be stable and amenable to storage. It retained activity (>60%) upon lyophilization even when tested after several weeks. Antibodies against the purified alliinase were prepared by standard procedures in rabbits. Western blots carried out using standard procedures, revealed that the polyclonal antibodies recognized alliinase also in crude extracts of the garlic clove, and inhibited its enzymatic activity. 1.2 Chemical-physical properties of alliinase

The Km value for garlic alliinase was calculated from Lineweaver-Burk plots derived from the enzymatic assays at different concentrations of synthetic alliin [ (-t-)S-allyl- cysteine sulfoxide] as substrate (Lineweaver, H. and Burk, D., J. Am. Chem. Soc. 1934, £6:658-666), and shown to be 2.4 mM. The pH optimum of the enzyme for the cleavage of alliin was 6.5 (Fig. 3) and the isoelectric point (pi) was 6.35. The pi value was determined by analytical IEF on PhastGel IEF 3.5-10 media, and a pi-marker protein kit for pi value estimation from 3.5 to 9.3. Pure alliinase preparation has an absorbance spectrum in the visible light region (Hewlett Packard 845 2A diode array spectrophotometer, USA) with a maximum at 430 nm (Fig. 4).

1.3 Enzyme Assay

Garlic alliinase activity was determined by a standard assay modified after Schwimmer and Mazelis (Arch. Biochem. Biophys. 1963, 100:66-73) . The standard assay mixture consisted of 0.1M Na-phosphate buffer (pH 6.5), 0.04mM pyridoxal 5'-phosphate, 0.2mM NADH, 10 units lactic dehydrogenase, 6mM synthetic alliin and sample containing alliinase, in a total volume of 1 ml. Enzymatic activity was traced in reaction cuvettes of 1 cm in path length by the decrease of difference in absorbance of NADH at 340 nm, corresponding to conversion of pyruvate to lactic acid by lactic dehydrogenase.

1.4 Amino acid composition

Amino acid analysis was performed on Dionex D-500 amino acid analyzer (Durram Instrument Corp., Palo Alto, Ca., U.S.A.). Protein samples were hydrolyzed with 6M HCl at 110°C for 22h. The amino acid composition of alliinase is presented in Table 2. TABLE 2

1.5 Molecular weight determination

The molecular weight of the holoenzyme (native garlic alliinase) was determined by HPLC size exclusion chromatography on a Varian TSK 3000 SW column (LKB, Uppsala, Sweden) equilibrated with 0.1M imidazole-HCl buffer pH 6.8 containing 0.15M NaCl. For calibration, marker proteins of known molecular weight (Sigma kit MW-GF-1000) were used. The molecular weight of the subunits was estimated by SDS-PAGE 7.5-20% gradient system using a Pharmacia molecular weight calibration kit containing (1) β-amylase (200 kDa), (2) alcohol dehydrogenase (150 kDa), (3) bovine serum albumin (67 kDa), (4) ovalbumin (45 kDa) and (5) ribonuclease (13.7 kDa). The results are depicted in Fig. 5. Fractions of 0.3 ml were collected. By using a standard curve based on the elution volumes of the standard protein mixture (Sigma, as noted above), the molecular weight of garlic alliinase was shown to be about 90,000 (Fig. 5).

Applying denaturing SDS-PAGE, the molecular weight of the alliinase subunits was determined using 7.5-20% gradient gels and Electrophoresis Molecular Weight Calibration Kit (as noted above, Pharmacia, Uppsala, Sweden), and shown to be 51,400 (Fig. 6), suggesting that the enzyme consists of two subunits of identical size.

1.6 Carbohydrate analysis

The carbohydrate content of alliinase was determined by using the phenol-sulfuric acid method of Dubois et al. (Anal. Chem. 1956, ^:350-356) and glucose as a standard. It was found that the enzyme contains 6% carbohydrate, of which mannose accounts for a significant portion. The glycoprotein nature of alliinase was also confirmed by its interaction with Concanavalin A. In this analysis alliinase was observed to bind tightly to a Sepharose Con A column and was eluted by α-methyl mannoside (0.2 M) (data not shown).

1.7 N-teπninal seguencing of garlic alliinase

Fractions from the HPLC column of 1.1 above containing purified alliinase were fractionated on an SDS-PAGE gel and the enzyme band was transferred by Western blotting at 300 mA 3-5h onto an I mobilon polyvinylene difluoride (PVDF) membrane (Millipore, Bedford, U.S.A.) in 10 mM CAPS (3- cyclohexylamino-1-propane sulfonic acid) buffer pH 11.0 containing 10% methanol (Matsudaira, P., J. Biol. Chem. 1987, 262:10035-10038) . The membrane was washed with water and stained with Coomassie blue (0.25%) in water solution containing 50% methanol/10% acetic acid. The stained bands were cut off from the membrane after drying and subjected to automated Edman degradation on an Applied Biosystems Model 475A Protein Sequencer with controller 900A.

N-terminal sequencing of purified native garlic alliinase revealed a 25 amino acid sequence as follows :

KMTETMKAAEEAEAVANINCSEHGR (Sequence I)

At room temperature, purified garlic alliinase undergoes degradation after 24-48 hours, producing a truncated alliinase component of 40 kDa. N-terminal sequencing of the truncated component was performed as above and a 19 amino acid sequence was obtained :

FNPVSNFISFELEKTIKEL (Sequence II)

In order to obtain additional information on amino acid sequences of alliinase, the purified enzyme (2 mg/ml) was partially cleaved with CNBr, which cleaves at methionine residues, by standard procedures. Fig. 7 depicts separation on SDS-PAGE (Fling, S.P. and Gregerson, D.S. 1986, Anal. Biochem. 155:83-88) of native alliinase (lane 1) and of CNBr- cleaved peptides (lane 2). For calibration, a Sigma molecular weight calibration kit (MW-SDS-17 kit) containing yoglobin and fragments thereof (lane before lane 1) and a calibration kit containing β-galactosidase (135 kDa), phosphorylase b (94 kDa), bovine serum albumin (67 kDa), ovalbumin (43 kDa) and carbonic anhydrase (28 kDa) (lane 3), were used. Since at least 12 methionine residues are present in alliinase (Table 2) , a number of peptides were obtained after CNBr cleavage (Fig. 7, lane 2). Of these peptides, the peptides of molecular weight 30 and 5 kDa were blotted on Immobilon PVDF membranes and sequenced as above, revealing the following N- terminal sequences III and IV, respectively :

KAAEEAEAVANINCSEHGRAFL (Sequence III)

RPPSPSYAWVNCEWEEDKDCYQTFQNGRI (Sequence IV)

The 22 amino acid N-terminal sequence of the 30 kDa peptide (sequence III above) exactly coincides with a CNBr cleavage at the second methionine (met-6) residue close to the N-terminal end of the intact polypeptide chain (see sequence I, above, and Fig. 12, described in Example 3.6, below) . Example 2

2.1 Synthesis of alliin

Alliin was synthesized from L-cysteine by using an improvement of the procedure of Stoll and Seebeck (Helv. Chim. Acta, 1949, £2:197-205). Thus L-cysteine (400 g, 3.3 moles) in water was reacted with allyl^■bromide (335 ml, 3.87 moles) in the presence of NaOH (145 ml, 2.72 moles) at pH 7.5 and temperature 35-40°C. The resulting S-allyl-L-cysteine (α=-12°, yield:332 g, about 62%) was purified and oxidized (190 g, 1.17 moles) with H₂0₂ (50%, 79.8 ml) to produce (±)S-allyl-L-cysteine sulfoxide. Fractional crystallization of the end product resulted in (+)S-allyl-L-cysteine sulfoxide (M.P.=164°, [α]o²° in H₂0=+62.1°, overall yield : 15.7%), which was shown to be identical to natural alliin, the true substrate of garlic alliinase. Careful regulation of the pH during the various steps increased the yield of the stereospecific end product. Thin layer chromatography (TLC) identification of S-allyl-L-cysteine and alliin was carried out on precoated cellulose plates (Merck, Darmstadt, Germany) using solvent system consisting of n- butanol-acetic acid-H₂0 (4:1:1) . After drying, the plates were sprayed with 0.25% ninhydrin reagent and placed in an oven at 110°C for 10 min. HPLC analysis of alliin was carried out on an LKB HPLC system with SP 4290 integrator (Spectraphysics) . For separation, Lichrosorb RP-18 reverse phase HPLC column (Merck, Germany) was used. The eluant was methanol (60%) in water containing 0.1% formic acid.

2.2 Enzymatic preparation of allicin

To 350 mg of alliin dissolved in 300 ml of 0.1M Na- phosphate buffer pH 6.5 were added 30 mg lyophilized alliinase (from hydroxylapatite step of Example 1.1. Specific activity : 30 units/mg) . After 2h of incubation, the product was extracted first with 300 ml and then with 200 ml of ether. The combined ether extracts were dried with Na₂S0-ι and paper filtered. Silica gel (dried at 150°C) was added to the solution and the ether removed with a stream of dry air at room temperature. The allicin thus obtained (92 mg) was dried to constant weight and stored in a refrigerated desiccator over sulfuric acid (98%).

The formation of allicin was extremely rapid, but significant amounts of alliinase were required to yield a satisfactory conversion of alliin into allicin, indicating that alliinase is inactivated during the reaction. Most of the enzyme retained its activity for 2 min. When preincubated with excess alliin for 15 min, and then separated from the substrate on a Sephadex G-25 column, alliinase was found to be completely inactive. The enzyme is therefore able to convert only limited amounts of alliin.

2.3 Identification of allicin

Two methods were used for identification of the allicin produced in the enzymatic reaction of 2.2 above.

Allicin was qualitatively identified by thin layer chromatography (TLC) using silica gel plates (Merck, Germany) and a solvent system consisting of benzene:ethyl acetate (90:10), revealing, in the presence of I_2/ an Rf of 0.375, as previously reported (Barone F.E. and Tansey, M.R. , Mycologia 1977, 60:793-825) .

Allicin was quantitatively determined by HPLC analysis using LKB HPLC system with SP 4290 integrator (Spectraphysics) .

The separation of alliin and allicin was carried out by loading a mixture of alliin and allicin on a Lichrosorb RP-18 reverse . phase HPLC column and eluting with 60% methanol in water containing 0.1% formic acid. The integrator was calibrated with known amounts of alliin and allicin standards prepared as in 2.1 and 2.2 above (from 0.3 to 3 μg), enabling quantitative determination of allicin. This is a very rapid, sensitive and useful procedure which enables the detection of both alliin and allicin in body fluids.

Fig. 8 shows HPLC analysis of a mixture of alliin (25 mg/ml) with purified garlic alliinase (50 units/ml) after incubation for 20 min. Aliquots of 0.1 μl (mixture diluted 1:1000 in 60% methanol) were applied to the HPLC column (flow rate of 0.54 ml/min) . After elution, the reaction products were detected at 205 nm : alliin at the lower peak and allicin at the higher peak.

Further incubation of the enzymatic reaction did not increase the yields of alliicin, neither did the addition of substrate to the reaction mixture. Re-isolation of alliinase protein from the reaction mixture by Sephadex G-25 revealed that the enzyme was inactivated to a great extent. Rabbit serum containing anti-alliinase antibodies inhibited the enzymatic activity of alliinase at dilutions of 1:100.

Example 3

Isolation of alliinase cDNA

3.1 Isolation of mRNA for gaxlic alliinase

Total RNA was isolated from various parts, including the leaves, of garlic plants by the guanidium/CsCla method (Sambrook et al., 1989 Molecular Cloning : A Laboratory Manual, Cold Spring Harbor Laboratory, New York, USA). It should be noted that the preferred source of total RNA and hence polyA*RNA, in particular, those RNAs specific for alliinase, are the mature bulbs of garlic plants (see Example 1, above, and Example 4, below). The polyA- RNA was purified from total RNA by oligo-dT cellulose chromatography column (Boehringer Mannheim), as described by Aviv and Leder, Proc. Natl. Acad. Sci. USA 1972, 69:1408.

3.2 Preparation of cDNA library in lambda gtll expression vector

5 μg of polyA* RNA obtained from mature garlic bulbs of step 3.1 above was used as template to construct a cDNA library in lambda gtll expression vector using standard techniques (Gubler et al. , Gene 1983, £5.5 63; Ausubel et al. , eds. Current Protocols in Molecular Biology Vol. 1, 1987). The cDNA library was constructed by using cDNA synthesis and cDNA cloning kits (Promega Corporation, Madison WI, USA). The cDNA was generated by reverse transcriptase and oligo-dT primers. The single-stranded cDNA was rendered double- stranded with DNA polymerase I. After ligation to EcoRI adaptors, the cDNAs were ligated with EcoRI cut and dephosphorylated lambda gtll vector. The ligation mix was packaged in vitro using Packagene^R system (Promega) . The cDNA library contained about 75,000 primary recombinant phages.

3.3 Preparation of synthetic oligonucleotide probes

Degenerate mixed oligonucleotides (Mix 1 and 2, Fig. 9) corresponding to the N-terminal amino acid sequence 10 to 15 of the purified garlic alliinase (Sequence I in Example 1.7 above; and also corresponding to the N-terminal amino acid sequence 4 to 9 of the 30 kDa peptide, sequence III in Example 1.7 above), were synthesized by standard procedure using an Applied Biosystems 480B DNA synthesizer. The inclusion of the next two nucleotides from the seventh amino acid is indicated in Fig. 9 by asterisks. I - indicates the nucleotide derivative ionosine. An Xbal restriction site (underlined) was also added to facilitate further manipulation of the amplified cDNA.

3.4 Isolation of alliinase cDNA sequence by MOPAC

The strategy of MOPAC (Mixed Oligonucleotide Primed Amplification of cDNA) , (Lee et al., 1988 Science 239:1288- 1291) was used in order to isolate and clone a fragment of alliinase cDNA.

The first strand cDNA was generated from the purified polyA* RNA of step 3.1 above by reverse transcriptase and oligo-dT primers and used as a template for MOPAC reaction. PCR was performed with either Mix 1 or Mix 2 (see Fig. 9) as sense primers and oligo-dT₃₀ as antisense primer. The reactions were carried out in 0.25mM of each deoxynucleotide triphosphate (dATP, dCTP, dGTP or dTTP) , 120 pmoles mixed oligonucleotides (Mix 1 or Mix 2), 30 pmoles oligo-dT₃o, 2 μl of the first strand cDNA reaction, and 1 unit of a Taq polymerase in a final volume of 100 μl. After 30 cycles of amplification (40 sec. 92°C, 40 sec. 58°C and 40 sec. 72°C), a PCR product of about 400 base pairs was detected on the gel (1.5% agarose gel containing ethydium bromide). The PCR fragment was directly subcloned into plasmid pCR 1000 by using TA cloning kit (Invitrogen Corporation, USA), according to the manufacturer's protocol. Several clones containing the about 400 bp of the PCR insert were identified by plasmid purification and restriction analysis by Xbal (Sambrook et al. , cited above) .

One of these clones, designated pAli-1, was sequenced and revealed an open reading frame (ORF) sequence corresponding, with one mismatch, to the NH₂-terminal amino acid sequence of alliinase (see sequence I in Example 1.7 above) . DNA sequencing was performed by the dideoxynucleotide chain termination method (Sanger et al. , 1977, Proc. Natl. Acad. Sci. USA 2i^:*5463), by using the Sequenase kit (US Biochemicals) , according to the manufacturer's protocol. The sequence of Ali-1 is shown in Fig. 10, in which the underlined nucleotides indicate the codons for the alliinase amino acid sequence 10 to 16 (see sequence I in Example 1.7 above), used for the synthesis of the oligonucleotide probes Mix 1 and Mix 2 (with deletion of last nucleotide T of codon GCT of ala in position 16), and the dots indicate a 9-amino acid sequence of the ORF product corresponding to the NH₂-terminal amino acid sequence of the protein alliinase (NINCSEHGR in Sequence I above) . Matching was also observed with three amino acids, AFL, which corresponded to the amino acid sequence obtained from the 30 kDa CNBr-cleaved alliinase peptide (see sequence III in Example 1.7 above). These results clearly indicate that the PCR product pAli-1 corresponded to the alliinase cDNA.

3.5 Screening of garlic alliinase cDNA library

Screening of the cDNA library prepared in step 3.2 above was performed essentially as described (Sambrook et al., cited above). Duplicate plaque lift filters were incubated with ³²P-labeled Ali-1 (labelling of Ali-1 being by standard procedures), in hybridization buffer containing 5XSSC, 5X Denhardt solution and 1% SDS at 60°C for 24 hours. The filters were then washed at 65°C in 0.2XSSC, ImM EDTA and 0.1% SDS and exposed to X-ray film. Three positive plaques were purified for further analysis. Each of the isolated plaques was amplified and the phage DNA was prepared by standard procedure (Ausubel et al., eds. Current Protocols in Molecular Biology, Vol. 1, 1987). The DNAs were then digested with EcoRI to release the cDNA insert, and electrophoresed on standard non-denatured agarose gel. From one of the above positive plaques (or clones) a cDNA fragment of about 1.3 kb was eluted from the gel, subcloned into the dephosphorylated EcoRI site of pBluescript*¹ (Stratagene) to yield plasmid pAli-2 and sequenced. The arrowhead in Fig. 10 indicates, within the sequence of pAli-1, the start codon of clone pAli- 2. The complete sequence of pAli-2 is contained in the sequence depicted in Fig. 12. The Ali-2 sequence contains an open reading frame of 380 amino acids. Its sequence includes two regions (residues 358-407 and residues 96-114 of Fig. 12) which share homology with the N-terminal amino acid sequences obtained from the 5 kDa CNBr cleavage peptide of alliinase (sequence IV in Example 1.7 above), as well as with the truncated (40 kDa) product of alliinase (see sequence II in Example 1.7 above). The pAli-2 clone was found to lack 222 bp at the 5' sequence corresponding to 74 amino acids off the NH₂-terminal of the mature alliinase (see full sequence in Fig. 12, where black arrow denotes start of pAli-2 sequence). The 3' sequence showed a stop codon at 1507 bp (Fig. 12), 3 potential polyadenylation signals, and a poly(A) tail of 28 residues (Fig. 12). Sequence comparisons between the PCR amplified fragment pAli-1 and clone pAli-2 revealed an overlap of about 200 bp (see Figs. 10 and 12). Comparison of the deduced amino acid composition of pAli-1 and pAli-2 with that of the natural alliinase purified from garlic bulbs (Table 2) showed considerable similarity. Differences were found in cysteine which is known to be sensitive to acid degradation. The deduced composition confirmed that alliinase has 10 cysteines which are potentially capable of formation of five intra-molecular disulfide bonds. In addition, four N- linked glycosylation sites and a hydrophilic character were predicted by the computer program (Fig. 13). This matches well with the water soluble characteristics and the carbohydrate content of the alliinase protein. OPne subunit of the mature enzyme consists of 448 amino acid residues and has a molecular weight of 51,500. The theoretical extinction coefficient of alliinase is

.

3.6 Construction of alliinase cDNA

Fig. 11 is a schematic representation of the construction of the garlic alliinase cDNA, designated pSPAli.

The construction involved the insertion of two fragments upstream from the 5' end of the alliinase cDNA fragment pAli- 2 of step 3.5 above : the PCR fragment, pAli-1 and a signal peptide fragment derived from the wheat storage protein, α- gliadin (Rafalski et al., 1984 EMBO J. £:1409-1415) , to direct the protein into the endoplasmic reticulum (ER) compartment for correct processing and folding. The resulting construct, pSPAli, contains all the sequences encoding for the mature alliinase protein, except for the DNA encoding the first nine N-terminal amino acid residues, downstream from the T7 promoter of the above noted Bluescript^R plasmid.

At the first step, an Ncol/Hindi (210 bp) fragment from pAli-2 was replaced by an EcoRI (blunted)/Ncol fragment (384 bp) from pAli-1 (the common Ncol site of pAli-1 and pAli-2 is shown in Fig. 12), to produce pAli-3. For further manipulation, the orientation of this clone was changed by subcloning of the EcoRI/NotI fragment into the EcoRI/NotI sites of pBluescript™ KS- (Stratagene) to give pAli-4. Since alliinase is expected to be a secretory protein, a synthetic linker corresponding to amino acids 1 to 9 of the NH₂-terminal sequence of alliinase, and nucleotide sequence encoding a signal peptide SacI/PstI fragment (115 bp) derived from wheat α-gliadin, were ligated into the Sacl/Xbal sites (blunted by mung bean nuclease) at the 5' end of pAli-4 to produce pSPAli. The enzyme sites are as follows : X, Xbal; N, NotI; RI, EcoRI; RV, EcoRV; H2, Hindi; S, Sa . Parentheses indicate blunted sites, either by Klenow fragment of E. coli DNA Polymerase I or mung bean nuclease. T7 and T3 indicate RNA polymerase promoters from the pBluescript KS* plasmid (Stratagene) . All nucleic acid manipulations were performed as recommended by the suppliers or by standard methods (Sambrook et al. , cited above) .

Re-sequencing of the recombinant plasmid pSPAli verified that the signal peptide was added in frame 5' to the alliinase coding sequence (see also Fig. 12).

Moreover, as described hereinbelow (Example 5), synthetic alliinase mRNA, prepared using the linearized pSPAli clone as template, was injected into Xenopus Oocytes and resulted in the production of a protein recognized by anti-alliinase serum and having a similar molecular weight to alliinase.

Fig. 12 shows the full length cDNA and deduced amino acid sequence of garlic alliinase. The pAli-1 fragment nucleotide sequence (see Example 3.4) is from nucleotides 184-605, the triangle indicating the end of pAli-1. The arrowed line indicates the sequence of the oligonucleotide primers used in the preparation of pAli-1 (Examples 3.3 and 3.4). The pAli-2 nucleotide sequence (see Example 3.5) is from nucleotides 385-1712, the start of pAli-2 being denoted by a black arrowhead. The overlapping region of the DNA fragments pAli-1 and pAli-2 is between nucleotides 385 and 605 (Example 3.5), and the common, above noted, Ncol site is underlined. The stop codon for translation is indicated by asterisks. The three transcription termination and polyadenylation signals are in italics and bold-faced. The boxed region represents the Kozak-like structure (see Example 3.7). The two regions sharing homology with the N-terminal amino acid sequences of CNBr cleavage peptides as well as that of the truncated alliinase (see Example 1.7) are indicated in bold-face (amino acid residue 25 denoted by an open arrowhead, amino acid residue 118 and amino acid residue 382). The 5' and 3'-ends of the DNA sequence, shown in lower case letters and the deduced N-terminal amino acid sequence also shown in the lower case were obtained from clone pAli-5A after screening the lambda ZAPII cDNA library (Example 3.7).

The above noted clone pAli-3 which starts at nucleotide G at position 191 of Fig. 12, was introduced into E. Coli DM- 5α and deposited with the Collection Nationale de Cultures de Microorganisms, Institut Pasteur, Paris, France under No. I- 1271.

Fig. 13 shows computer predictions from the nucleotide and amino acid sequences of alliinase (Fig. 12). The deduced amino acid sequence of the alliinase gene was analyzed on the computer using a program called "Plotstructure™" . The hydrophilic nature and four glycosylation sites of the protein are shown.

3.7 Isolation of a full length cDNA clone from a lambda ZAPII cDNA library

Total RNA from garlic bulbs was extracted as described above (Examples 3.1 and 3.2) and a garlic cDNA library was constructed in lambda ZAPII vectors (Clontech. Calif. USA), using the manufacturers' protocol and standard procedures. The number of recombinant .phages obtained was about 1.6xl0^β/ml, of which about 7.2x10* were screened using the above noted pAli-1 as a probe. The phagemids were excised from the positive clones by the automatic excision protocol according to the above noted manufacturers' instructions.

A positive clone was obtained from the lambda ZAPII cDNA library, which was designated pAli-5A, and was sequenced. A sequence comparison of pAli-5A to the sequences of pAli-5, pAli-1 and pAli-2 (see Examples 3.4-3.6 and the sequence shown in Fig. 12), revealed that pAli-5A has an additional 183 base pairs at the 5' end of its sequence, these additional bases having been added to Fig. 12 (see lower case letters at 5' end of sequence of Fig. 12, as described in Example 3.6 above). The additional information obtained from the sequence of pAli-5A provided proof that the deduced sequence of the first nine N-terminal amino acids of pAli-5A shared homology with the N-terminal amino acid sequence of the natural garlic alliinase protein (see sequence I in Example 1.7 and Fig. 12). The open reading frame of pAli-5A started 75 nucleotides downstream of the coding region for the mature protein, beginning with a methionine and a peptide possessing a hydrophobic character, suggesting that it may serve as the signal structure which allows secretion of the protein from the endoplasmic reticulum (ER). A Kozak-like motif (Kozak, M. (1981) Nucl. Acids Res. £:5233-5253) was detected from nucleotides 73-79 (see boxed region in Fig. 12 as described in Example 3.6 above). Some sequence differences with pAli-2 were also detected at the 3' end of the pAli-5A sequence after the stop codon of the alliinase coding region (see Fig. 12) .

A sample of the full length cDNA clone of alliinase (pAli-5A, isolated from the lambda ZAPII library, as noted above) has been deposited at the Collection Nationale de Cultures de Microorganisms, Institut Pasteur, Paris, France under No. 1-1360.

Example 4

Northern blot analysis of alliinase mRNA

In order to prepare radioactive RNA probes for Northern analysis, plasmids pAli-1 to pAli-4 linearized by restriction enzyme EcoRI were treated with 50 μg/ml proteinase K (30 min at 37°C), extracted with phenol : chloroform : isoamyl alcohol (25:24:1), and precipitated by ethanol in the presence of 2M ammonium acetate. The linearized plasmids were then subjected to transcription by the appropriate RNA polymerase (T7 or T3) and radioactive mRNA probes were prepared as described by Kawata et al. (1988 Plant Molecular Biology Manual Bl_:1-22), with the exception that 100 μCi [α³²P]UTP (400 Ci/mmol) was used, instead of "cold" UTP.

Northern blot analysis using ³²P-pAli-l showed that an about 1.8 Kb alliinase mRNA is present in leaves as well as in developed bulbs, but no alliinase mRNA could be detected in roots. Fig. 14 is the Northern blot analysis of alliinase mRNA in different tissues of the garlic plant. Garlic poly(A)+ RNA from developed bulbs (5 μg, lane 1) and total RNA (30 μg) extracted from roots (lane 2), developed bulbs (lane 3), or from the lower part of the leaves (lane 4) were electrophoresed on denatured 1.5% formaldehyde-agarose gel, blotted onto nylon membrane (Hybond-N, Amersham) and hybridized with ³=P-labeled Ali-1 antisense RNA at 65°C for 12-16 hours in the presence of 1% SDS, 5 x SSC and 5 x Denhardt solution (Sambrook et al., 1989. Molecular Cloning : A Laboratory Manual, Cold Spring Harbor Laboratories, New York, USA), followed by several washings in 0.2 x SSC and 0.1% SDS at 65°C to remove nonspecific hybrids. Filters were dried and autoradiographed. In Fig. 14 : (a) methylene blue stained blot. (b) autoradiogram. The ribosomal RNA (28S and 18S) position is indicated, which served as a marker for size estimation of the RNAs separated on the above agarose gel and stained with methylene blue.

Since the above noted single band of about 1.8 Kb was detected in the Northern blots and since the molecular weight of alliinase is about 50 kDa, these results indicate that the pAli-1 probe i.e. the pAli-1 sequence hybridizes specifically to alliinase mRNA.

Example 5

Expression of alliinase in Xenopus oocytes

The viability of the chimeric alliinase cDNA construct of Fig. 11, i.e. pSPAli, is assessed by microinjection of its corresponding mRNA transcript into Xenopus laevis oocytes (Kawata et al., 1988 Plant Molecular Biology Manual B7:l-22) . Stage VI oocytes are injected (approximately 2-5 ng mRNA in a 50 nl solution per Oocyte) with the in vitro transcribed alliinase mRNA. For in vitro transcription the pAli-5 construct (see Example 3.6 and Figs. 11 and 12) containing the alliinase cDNA was linearized at its 3' end downstream to the poly(A) sequence by digestion with Hindlll. Capped poly(A)* RNA was transcribed in vitro using T7 RNA polymerase and the cap analogue G(5' )ppp(5' )G. After injection with the in vitro transcribed mRNA, the oocytes were then incubated for 24 hours in OR-2 medium at 20°C (Colman, A., 1984, in : Transcription and Translation, Hames, B.D. and Higgins, S.J., eds. Oxford:IRL Press, pp. 271-302). For radiolabeling of proteins, oocytes are transferred after 6 hours into a 96-well culture dish (5 oocytes in each well) and incubated in OR-2 medium, with or without L-[^3SS] methionine (>800 Ci/mmol, 25 μCi/well). 24 hours after injection, the oocytes and the medium are combined and stored at -80°C. For extraction of alliinase protein, frozen oocytes (15 oocytes) and medium are homogenized in 0.2 ml ice cold buffer containing 20 M Tris-Cl pH 7.4, 5 mM Na₂EDTA, 100 mM NaCl, 1% Triton X-100, 0.1 mM PMSF and 0.5 mM leupeptin. Following centrifugation (12000 x g for 10 min at 4°C), the supernatant is collected. Extracts from unlabeled oocytes are tested for alliinase activity as described in Example 1.3 above and the radiolabeled extracts are immunoprecipitated with anti-alliinase serum (Colman, A., 1984, cited above).

In Fig. 15 there is shown the results of the expression of alliinase mRNA in the oocytes as analyzed by SDS-PAGE and Western blotting using anti-alliinase antibodies. Equal amounts of protein from the culture medium and extracts of the injected oocytes, as well as extracts from control non- injected oocytes were subjected to SDS-PAGE followed by Western blot analysis with anti-alliinase antibodies. In Fig. 15, Lane 1 represents the extracts of injected oocytes; Lane 2 represents the oocyte culture medium; Lane 3 represents the negative control (non-injected oocytes); and Lane 4 represents crude protein extracts from garlic bulbs. The position of Alliinase is indicated by an arrow. Thus, in Fig. 15 there is clearly observed a band in Lane 1 which binds anti-alliinase antibodies, indicating that the pSPAli construct is capable of being transcribed into an mRNA that can be translated in the oocytes to produce a product (alliinase) recognized by anti-alliinase antibodies. A faint band is also detected in the oocyte culture medium (lane 2 Fig. 15) indicating that the above translated product is capable of being secreted by the oocytes. No signal (band) is visualized in the control, non-injected oocyte extracts (lane 3, Fig. 15).

Example 6

Allicin treatment of cholesterol-fed rabbits

In order to establish a rabbit model for atherosclerosis, cholesterol-rich food (0.5%) was administered ad libidum the first 2 weeks and 0.25% cholesterol for the remaining times was given to 2 groups of rabbits, with 4 rabbits in each group. Average weight of rabbits at the beginning was 2 kg. The levels of blood cholesterol and trigylcerides were determined in blood samples taken every two weeks. During the first two weeks, the blood cholesterol levels jumped from 100 μg/ml to 1,200 on average and the variation between the 4 rabbits was ±20% (Fig. 16).

After the initial two weeks, the first group of rabbits was left as control receiving only placebo (saline). The second group was the treatment group. Each day, immediately before administration, a dose containing 10 mg of powdered alliin was mixed with 3 mg purified garlic alliinase (lyophilized) which is approximately equivalent to the amounts present in 30 g of garlic cloves. 1 ml of water was added to the dry mixture and the suspension immediately dripped with a syringe into the throat of the rabbit. The amount of allicin generated by this mixture within two minutes, as determined by the HPLC analysis described in Example 3.2 above was equivalent to 7 mg allicin, corresponding to about 3 mg/kg rabbit.

After 3 months from beginning of the experiment, the animals were sacrificed by 5% phenobarbitol solution injection and bleeding from a femoral artery. The aorta of all the animals was cut open longitudinally and then divided into two equal halves. One complete half of each aorta was fixed flat for 24h or more in 10% formalin and was stained with Sudan IV for gross examination. The lesions were graded according to the criteria recommended by the WHO Study Group on Atherosclerosis. The percentage of intimal surface area covered by atherosclerosis lesions was also determined. After grading the atherosclerosis lesions and gross photography, these complete halved stained with Sudan IV were used for histopathological studies. The results showed that the degree of atherosclerosis lesions in the aortas of control cholesterol-fed rabbits was about 77%, while in the allicin- treated rabbits it was significantly lower (about 58%).

Claims

1. A recombinant protein in a substantially pure fo having the activity of garlic alliinase.

2. A recombinant protein according to claim 1 having amino acid sequence selected from the group consisting of :

(i) the amino acid sequence depicted in Fig. 12;

(ii) the amino acid sequence depicted in Fig. 12, where one or more amino acids has been deleted, adde replaced or chemically modified without substantial reducing the garlic alliinase activity;

(iii) a fragment of the amino acid sequence of (i) or (i which essentially retains the garlic alliina activity; and

(iv) a dimer of (i), (ϋ) or (iii) having garlic alliina activity.

_*

3. A DNA sequence coding for a recombinant prote according to claim 1 or 2.

4. A DNA sequence according to claim 3 selected from t group consisting of :

(i) the DNA sequence depicted in Fig. 12;

(ii) a fragment of the DNA sequence depicted in Fig. encoding a protein having garlic alliinase activity

(iii) a DNA sequence derived from (i) or (ii) in which o or more amino acid encoding triplets has been adde deleted or replaced without substantially affecti the garlic alliinase activity of the protein encod thereby; (iv) a DNA sequence which is a degenerate equivalent of the sequence (i), (ϋ) or (iii), and encodes a protein having garlic alliinase activity; and

(v) a DNA sequence hybridizable to a sequence (i), (ii), (iii) or (iv) and which encodes a protein having garlic alliinase activity.

5. An expression vector comprising a DNA sequence according to claim 3 or 4.

6. An expression vector according to claim 5 being a plasmid.

7. An expression vector according to claim 5 being a virus.

8. A host cell transfected with an expression vector of any one of claims 5 to 7.

9. A host cell according to claim 8 selected from the group consisting of bacterial, yeast, plant, mammalian, and insect cells.

10. A process for the preparation of biologically active garlic alliinase comprising culturing the transfected host cell of claim 8 in a suitable medium and isolating the protein having the garlic alliinase activity produced by the transfected host cells.

11. A pharmaceutical composition comprising the recombinant protein of claim 1 and instructions for the administration thereof together with alliin.

12. A pharmaceutical composition comprising the recombinant protein of claim 1 and alliin, wherein the alliin and the recombinant protein are in a form which does not permit an enzymatic reaction in which alliin is broken down by alliinase, until the composition is administered to a patient.

13. A pharmaceutical composition according to claim 12 wherein the recombinant protein and the alliin are in a dried form.

14. A pharmaceutical composition according to claim 12, wherein the recombinant protein and the alliin are contained in different compartments, and come into contact with one another only after the composition is administered to the patient.

15. A food product comprising the recombinant protein of claim 1 and alliin, wherein the alliin and recombinant protein are in a form which does not permit an enzymatic reaction in which alliin is broken down by alliinase, until ingestion of the food.

16. A food product according to claim 15, wherein the recombinant protein and the alliin are in a dried form.

17. A food product according to claim 15, wherein the recombinant protein and the alliin are contained in different compartments, and come into contact with one another only after the composition is administered to the patient.