US 20020044950 A1
A polypeptide has first and second domains which enable the polypeptide to be translocated into a target cell or which increase the solubility of the polypeptide, or both, and further enable the polypeptide to cleave one or more vesicle or plasma-membrane associated proteins essential to exocytosis. The polypeptide thus combines useful properties of a clostridial toxin, such as a botulinum or tetanus toxin, without the toxicity associated with the natural molecule. The polypeptide can also contain a third domain that targets it to a specific cell, rendering the polypeptide useful in inhibition of exocytosis in target cells. Fusion proteins comprising the polypeptide, nucleic acids encoding the polypeptide and methods of making the polypeptide are also provided. Controlled activation of the polypeptide is possible and the polypeptide can be incorporated into vaccines and toxin assays.
1. A polypeptide comprising first and second domains, wherein said first domain is adapted to cleave one or more vesicle or plasma-membrane associated proteins essential to exocytosis, and wherein said second domain is adapted (i) to translocate the polypeptide into a cell or (ii) to increase the solubility of the polypeptide compared to the solubility of the first domain on its own or (iii) both to translocate the polypeptide into a cell and to increase the solubility of the polypeptide compared to the solubility of the first domain on its own, said polypeptide being free of clostridial neurotoxin and free of clostridial neurotoxin precursor that can be converted into toxin by proteolytic action.
2. A polypeptide according to
3. A polypeptide according to
4. A polypeptide according to
5. A polypeptide according to any preceding claim wherein the first domain exhibits endopeptidase activity specific for a substrate selected from one or more of SNAP-25, synaptobrevin/VAMP and syntaxin.
6. A polypeptide according to any preceding claim wherein said second domain comprises a clostridial toxin heavy chain HN portion.
7. A polypeptide according to any of claims 1-5 wherein said second domain comprises a fragment or variant of a clostridial toxin heavy chain HN portion.
8. A polypeptide according to
9. A polypeptide according to any of claims 1-8 further comprising a third domain adapted for binding of the polypeptide to a cell, by binding of the third domain directly to a cell or by binding of the third domain to a ligand or to ligands that bind to a cell.
10. A polypeptide according to
11. A polypeptide according to
12. A polypeptide according to
13. A polypeptide according to
14. A polypeptide according to any preceding claim comprising a botulinum toxin light chain or a fragment or a variant of a botulinum toxin light chain and a portion designated HN of a botulinum toxin heavy chain.
15. A polypeptide according to
16. A polypeptide according to
17. A polypeptide according to
18. A polypeptide according to any of claims 1-13 comprising a botulinum toxin light chain or a fragment or a variant of a botulinum toxin light chain and at least 100 N-terminal amino acids of a botulinum toxin heavy chain.
19. A polypeptide according to
20. A polypeptide according to
21. A polypeptide according to
22. A polypeptide according to
23. A polypeptide according to
24. A polypeptide according to
25. A polypeptide according to any of claims 14-24 lacking a portion designated HC of a botulinum toxin heavy chain.
26. A polypeptide comprising a botulinum toxin light chain and a fragment of a botulinum toxin heavy chain, said fragment being not capable of binding to cell surface receptors.
27. A polypeptide according to
28. A polypeptide according to any preceding claim comprising a variant of a clostridial toxin and further comprising a site for cleavage by a proteolytic enzyme, which cleavage site is not present in the native toxin.
29. A polypeptide according to
30. A polypeptide according to
31. A polypeptide according to
32. A fusion protein comprising a fusion of (a) a polypeptide according to any of claims 1-31 with (b) a second polypeptide being a polypeptide or oligopeptide adapted for binding to an affinity matrix so as to enable purification of the fusion protein using said matrix.
33. A fusion protein according to
34. A fusion protein according to
35. A composition comprising a derivative of a clostridial toxin, said derivative retaining at least 10% of the endopeptidase activity of the botulinum toxin, said derivative further being non-toxic in vivo due to its inability to bind to cell surface receptors, and wherein the composition is free of any component, such as toxin or a further toxin derivative, that is toxic in vivo.
36. A composition according to
37. A composition according to
38. A composition according to
39. A pharmaceutical composition comprising a composition according to
40. A nucleic acid encoding a polypeptide or a fusion protein according to any of claims 1-34.
41. A nucleic acid encoding a polypeptide or a fusion protein according to claim 40 and comprising nucleotides encoding residues 1-448 of a botulinum toxin type A light chain.
42. A nucleic acid according to
43. A nucleic acid encoding a polypeptide or a fusion protein according to
44. A nucleic acid encoding a polypeptide or a fusion protein according to
45. A nucleic acid according to any of claims 40-44 comprising nucleotides encoding a restriction endonuclease cleavage site not present in native clostridial toxin sequence.
46. A nucleotide according to
47. A DNA according to any of claims 40-46.
48. A DNA selected from SEQ ID No:s 1, 8, 10, 12, 14, 16, 18, 23 and 24.
49. A method of manufacture of a polypeptide according to any of claims 1-31 comprising expressing in a host cell a nucleic acid according to any of claims 40-48 and recovering the polypeptide.
50. A method of manufacture of a polypeptide according to any of claims 1-31 comprising expressing in a host cell a nucleic acid encoding a fusion protein according to
51. A method of manufacture according to
52. A cell expressing a polypeptide or fusion protein according to any of claims 1-34.
 This invention relates to recombinant toxin fragments, to DNA encoding these fragments and to their uses such as in a vaccine and for in vitro and in vivo purposes.
 The clostridial neurotoxins are potent inhibitors of calcium-dependent neurotransmitter secretion in neuronal cells. They are currently considered to mediate this activity through a specific endoproteolytic cleavage of at least one of three vesicle or pre-synaptic membrane associated proteins VAMP, syntaxin or SNAP-25 which are central to the vesicle docking and membrane fusion events of neurotransmitter secretion. The neuronal cell targeting of tetanus and botulinum neurotoxins is considered to be a receptor mediated event following which the toxins become internalised and subsequently traffic to the appropriate intracellular compartment where they effect their endopeptidase activity.
 The clostridiai neurotoxins share a common architecture of a catalytic L-chain (LC, ca 50 kDa) disulphide linked to a receptor binding and translocating H-chain (HC, ca 100 kDa). The HC polypeptide is considered to comprise all or part of two distinct functional domains. The carboxy-terminal half of the HC (ca 50 kDa), termed the HC domain, is involved in the high affinity, neurospecific binding of the neurotoxin to cell surface receptors on the target neuron, whilst the amino-terminal half, termed the HN domain (ca 50 kDa), is considered to mediate the translocation of at least some portion of the neurotoxin across cellular membranes such that the functional activity of the LC is expressed within the target cell. The HN domain also has the property, under conditions of iow pH, of forming ion-permeable channels in lipid membranes, this may in some manner relate to its translocation function.
 For botulinum neurotoxin type A (BoNT/A) these domains are considered to reside within amino acid residues 872-1296 for the HC, amino acid residues 449-871 for the HN and residues 1-448 for the LC. Digestion with trypsin effectively degrades the HC domain of the BoNT/A to generate a non-toxic fragment designated LHN, which is no longer able to bind to and enter neurons (FIG. 1). The LHN fragment so produced also has the property of enhanced solubility compared to both the parent holotoxin and the isolated LC.
 It is therefore possible to provide functional definitions of the domains within the neurotoxin molecule, as follows:
 (A) Clostridial Neurotoxin Light Chain:
 a metalloprotease exhibiting high substrate specificity for vesicle and/or plasma-membrane associated proteins involved in the exocytotic process. In particular, it cleaves one or more of SNAP-25, VAMP (synaptobrevin/cellubrevin) and syntaxin.
 (B) Clostridial Neurotoxin Heavy Chain HN Domain:
 a portion of the heavy chain which enables translocation of that portion of the neurotoxin molecule such that a functional expression of light chain activity occurs within a target cell.
 the domain responsible for translocation of the endopeptidase activity, following binding of neurotoxin to its specific cell surface receptor via the binding domain, into the target cell.
 the domain responsible for formation of ion-permeable pores in lipid membranes under conditions of low pH.
 the domain responsible for increasing the solubility of the entire polypeptide compared to the solubility of light chain alone.
 (C) Clostridial Neurotoxin Heavy Chain HC Domain.
 a portion of the heavy chain which is responsible for binding of the native holotoxin to cell surface receptor(s) involved in the intoxicating action of clostridial toxin prior to internalisation of the toxin into the cell.
 The identity of the cellular recognition markers for these toxins is currently not understood and no specific receptor species have yet been identified although Kozaki et al. have reported that synaptotagmin may be the receptor for botulinum neurotoxin type B. It is probable that each of the neurotoxins has a different receptor.
 It is desirable to have positive controls for toxin assays, to develop clostridial toxin vaccines and to develop therapeutic agents incorporating desirable properties of clostridial toxin.
 However, due to its extreme toxicity, the handling of native toxin is hazardous.
 The present invention seeks to overcome or at least ameliorate problems associated with production and handling of clostridial toxin.
 Accordingly, the invention provides a polypeptide comprising first and second domains, wherein said first domain is adapted to cleave one or more vesicle or plasma-membrane associated proteins essential to neuronal exocytosis and wherein said second domain is adapted (i) to translocate the polypeptide into the cell or (ii) to increase the solubility of the polypeptide compared to the solubility of the first domain on its own or (iii) both to translocate the polypeptide into the cell and to increase the solubility of the polypeptide compared to the solubility of the first domain on its own, said polypeptide being free of clostridial neurotoxin and free of any clostridial neurotoxin precursor that can be converted into toxin by proteolytic action. Accordingly, the invention may thus provide a single polypeptide chain containing a domain equivalent to a clostridial toxin light chain and a domain providing the functional aspects of the HN of a clostridial toxin heavy chain, whilst lacking the functional aspects of a clostridial toxin HC domain.
 For the purposes of the invention, the functional property or properties of the HN of a clostridial toxin heavy chain that are required to be exhibited by the second domain of the polypeptide of the invention are either (i) translocation of the polypeptide into a cell, or (ii) increasing solubility of the polypeptide compared to solubility of the first domain on its own or (iii) both (i) and (ii). References hereafter to a HN domain or to the functions of a HN domain are references to this property or properties. The second domain is not required to exhibit other properties of the HN domain of a clostridial toxin heavy chain.
 A polypeptide of the invention can thus be soluble but lack the translocation function of a native toxin-this is of use in providing an immunogen for vaccinating or assisting to vaccinate an individual against challenge by toxin. In a specific embodiment of the invention described in an example below a polypeptide designated LH423/A elicited neutralising antibodies against type A neurotoxin. A polypeptide of the invention can likewise thus be relatively insoluble but retain the translocation function of a native toxin—this is of use if solubility is imparted to a composition made up of that polypeptide and one or more other components by one or more of said other components.
 The first domain of the polypeptide of the invention cleaves one or more vesicle or plasma-membrane associated proteins essential to the specific cellular process of exocytosis, and cleavage of these proteins results in inhibition of exocytosis, typically in a non-cytotoxic manner. The cell or cells affected are not restricted to a particular type or subgroup but can include both neuronal and non-neuronal cells. The activity of clostridial neurotoxins in inhibiting exocytosis has, indeed, been observed almost universally in eukaryotic cells expressing a relevant cell surface receptor, including such diverse cells as from Aplysia (sea slug), Drosophila (fruit fly) and mammalian nerve cells, and the activity of the first domain is to be understood as including a corresponding range of cells.
 The polypeptide of the invention may be obtained by expression of a recombinant nucleic acid, preferably a DNA, and is a single polypeptide, that is to say not cleaved into separate light and heavy chain domains. The polypeptide is thus available in convenient and large quantities using recombinant techniques.
 In a polypeptide according to the invention, said first domain preferably comprises a clostridial toxin light chain or a fragment or variant of a clostridial toxin light chain. The fragment is optionally an N-terminal, or C-terminal fragment of the light chain, or is an internal fragment, so long as it substantially retains the ability to cleave the vesicle or plasma-membrane associated protein essential to exocytosis. The minimal domains necessary for the activity of the light chain of clostridial toxins are described in J. Biol. Chem., Vol.267, No. 21, July 1992, pages 14721-14729. The variant has a different peptide sequence from the light chain or from the fragment, though it too is capable of cleaving the vesicle or plasma-membrane associated protein. It is conveniently obtained by insertion, deletion and/or substitution of a light chain or fragment thereof. In embodiments of the invention described below a variant sequence comprises (i) an N-terminal extension to a clostridial toxin light chain or fragment (ii) a clostridial toxin light chain or fragment modified by alteration of at least one amino acid (iii) a C-terminal extension to a clostridial toxin light chain or fragment, or (iv) combinations of 2 or more of (i)-(iii).
 In further embodiments of the invention, the variant contains an amino acid sequence modified so that (a) there is no protease sensitive region between the LC and HN components of the polypeptide, or (b) the protease sensitive region is specific for a particular protease. This latter embodiment is of use if it is desired to activate the endopeptidase activity of the light chain in a particular environment or cell. Though, in general, the polypeptides of the invention are activated prior to administration.
 The first domain preferably exhibits endopeptidase activity specific for a substrate selected from one or more of SNAP-25, synaptobrevin/VAMP and syntaxin. The clostridial toxin is preferably botulinum toxin or tetanus toxin.
 In an embodiment of the invention described in an example below, the toxin light chain and the portion of the toxin heavy chain are of botulinum toxin type A. In a further embodiment of the invention described in an example below, the toxin light chain and the portion of the toxin heavy chain are of botulinum toxin type B. The polypeptide optionally comprises a light chain or fragment or variant of one toxin type and a heavy chain or fragment or variant of another toxin type.
 In a polypeptide according to the invention said second domain preferably comprises a clostridial toxin heavy chain HN portion or a fragment or variant of a clostridial toxin heavy chain HN portion. The fragment is optionally an N-terminal or C-terminal or internal fragment, so long as it retains the function of the HN domain. Teachings of regions within the HN responsible for its function are provided for example in Biochemistry 1995, 34, pages 15175-15181 and Eur. J. Biochem, 1989, 185, pages 197-203. The variant has a different sequence from the HN domain or fragment, though it too retains the function of the HN domain. It is conveniently obtained by insertion, deletion and/or substitution of a HN domain or fragment thereof. In embodiments of the invention, described below, it comprises (i) an N-terminal extension to a HN domain or fragment, (ii) a C-terminal extension to a HN domain or fragment, (iii) a modification to a HN domain or fragment by alteration of at least one amino acid, or (iv) combinations of 2 or more of (i)-(iii). The clostridial toxin is preferably botulinum toxin or tetanus toxin.
 The invention also provides a polypeptide comprising a clostridial neurotoxin light chain and a N-terminal fragment of a clostridial neurotoxin heavy chain, the fragment preferably comprising at least 423 of the N-terminal amino acids of the heavy chain of botulinum toxin type A, 417 of the N-terminal amino acids of the heavy chain of botulinum toxin type B or the equivalent number of N-terminal amino acids of the heavy chain of other types of clostridial toxin such that the fragment possesses an equivalent alignment of homologous amino acid residues.
 These polypeptides of the invention are thus not composed of two or more polypeptides, linked for example by di-sulphide bridges into composite molecules. Instead, these polypeptides are single chains and are not active or their activity is significantly reduced in an in vitro assay of neurotoxin endopeptidase activity.
 Further, the polypeptides may be susceptible to be converted into a form exhibiting endopeptidase activity by the action of a proteolytic agent, such as trypsin. In this way it is possible to control the endopeptidase activity of the toxin light chain.
 In a specific embodiment of the invention described in an example below, there is provided a polypeptide lacking a portion designated HC of a clostridial toxin heavy chain. This portion, seen in the naturally produced toxin, is responsible for binding of toxin to cell surface receptors prior to internalisation of the toxin. This specific embodiment is therefore adapted so that it can not be converted into active toxin, for example by the action of a proteolytic enzyme. The invention thus also provides a polypeptide comprising a clostridial toxin light chain and a fragment of a clostridial toxin heavy chain, said fragment being not capable of binding to those cell surface receptors involved in the intoxicating action of clostridial toxin, and it is preferred that such a polypeptide lacks an intact portion designated HC of a clostridial toxin heavy chain.
 In further embodiments of the invention there are provided compositions containing a potypeptide comprising a clostridial toxin light chain and a portion designated HN of a clostridial toxin heavy chain, and wherein the composition is free of clostridial toxin and free of any clostridial toxin precursor that may be converted into clostridial toxin by the action of a proteolytic enzyme. Examples of these compositions include those containing toxin light chain and HN sequences of botulinum toxin types A, B, C1, D, E, F and G.
 The polypeptides of the invention are conveniently adapted to bind to, or include, a ligand for targeting to desired cells. The polypeptide optionally comprises a sequence that binds to, for example, an immunoglobulin. A suitable sequence is a tandem repeat synthetic IgG binding domain derived from domain B of Staphylococcal protein A. Choice of immunoglobulin specificity then determines the target for a polypeptide-immunoglobulin complex. Alternatively, the polypeptide comprises a non-clostridial sequence that binds to a cell surface receptor, suitable sequences including insulin-like growth factor-1 (IGF-1) which binds to its specific receptor on particular cell types and the 14 amino acid residue sequence from the carboxy-terminus of cholera toxin A subunit which is able to bind the cholera toxin B subunit and thence to GM1 gangliosides. A polypeptide according to the invention thus, optionally, further comprises a third domain adapted for binding of the polypeptide to a cell.
 In a second aspect the invention provides a fusion protein comprising a fusion of (a) a polypeptide of the invention as described above with (b) a second polypeptide adapted for binding to a chromatography matrix so as to enable purification of the fusion protein using said chromatography matrix. It is convenient for the second polypeptide to be adapted to bind to an affinity matrix, such as a glutathione Sepharose, enabling rapid separation and purification of the fusion protein from an impure source, such as a cell extract or supernatant.
 One possible second purification polypeptide is glutathione-S-transferase (GST), and others will be apparent to a person of skill in the art, being chosen so as to enable purification on a chromatography column according to conventional techniques.
 As noted above, by proteolytic treatment, for example using trypsin, of a polypeptide of the invention it is possible to induce endopeptidase activity in the treated polypeptide. A third aspect of the invention provides a composition comprising a derivative of a clostridial toxin, said derivative retaining at least 10% of the endopeptidase activity of the clostridial toxin, said derivative further being non-toxic in vivo due to its inability to bind to cell surface receptors, and wherein the composition is free of any component, such as toxin or a further toxin derivative, that is toxic in vivo. The activity of the derivative preferably approaches that of natural toxin,-and is thus preferably at least 30% and most preferably at least 60% of natural toxin. The overall endopeptidase activity of the composition will, of course, also be determined by the amount of the derivative that is present.
 While it is known to treat naturally produced clostridial toxin to remove the Hc domain, this treatment does not totally remove toxicity of the preparation, instead some residual toxin activity remains. Natural toxin treated in this way is therefore still not entirely safe. The composition of the invention, derived by treatment of a pure source of polypeptide advantageously is free of toxicity, and can conveniently be used as a positive control in a toxin assay, as a vaccine against clostridial toxin or for other purposes where it is essential that there is no residual toxicity in the composition.
 The invention enables production of the polypeptides and fusion proteins of the invention by recombinant means.
 A fourth aspect of the invention provides a nucleic acid encoding a polypeptide or a fusion protein according to any of the aspects of the invention described above. in one embodiment of this aspect of the invention, a DNA sequence provided to code for the polypeptide or fusion protein is not derived from native clostridial sequences, but is an artificially derived sequence not preexisting in nature.
 A specific DNA (SEQ ID NO: 1) described in more detail below encodes a polypeptide or a fusion protein comprising nucleotides encoding residues 1-871 of a botulinum toxin type A. Said polypeptide comprises the light chain domain and the first 423 amino acid residues of the amino terminal portion of a botulinum toxin type A heavy chain. This recombinant product is designated LH423/A (SEQ ID NO: 2).
 In a second embodiment of this aspect of the invention a DNA sequence which codes for the polypeptide or fusion protein is derived from native clostridial sequences but codes for a polypeptide or fusion protein not found in nature.
 A specific DNA (SEQ ID NO: 19) described in more detail below encodes a polypeptide or a fusion protein and comprises nucleotides encoding residues 1-1171 of a botulinum toxin type B. Said polypeptide comprises the light chain domain and the first 728 amino acid residues of the amino terminal protein of a botulinum type B heavy chain. This recombinant product is designated LH728/B (SEQ ID NO: 20).
 The invention thus also provides a method of manufacture of a polypeptide comprising expressing in a host cell a DNA according to the third aspect of the invention. The host cell is suitably not able to cleave a polypeptide or fusion protein of the invention so as to separate light and heavy toxin chains; for example, a non-clostridial host.
 The invention further provides a method of manufacture of a polypeptide comprising expressing in a host cell a DNA encoding a fusion protein as described above, purifying the fusion protein by elution through a chromatography column adapted to retain the fusion protein, eluting through said chromatography column a ligand adapted to displace the fusion protein and recovering the fusion protein. Production of substantially pure fusion protein is thus made possible. Likewise, the fusion protein is readily cleaved to yield a polypeptide of the invention, again in substantially pure form, as the second polypeptide may conveniently be removed using the same type of chromatography column.
 The LHN/A derived from dichain native toxin requires extended digestion with trypsin to remove the C-terminal 1/2 of the heavy chain, the HC domain. The loss of this domain effectively renders the toxin inactive in vivo by preventing its interaction with host target cells. There is, however, a residual toxic activity which may indicate a contaminating, trypsin insensitive, form of the whole type A neurotoxin.
 In contrast, the recombinant preparations of the invention are the product of a discreet, defined gene coding sequence and can not be contaminated by full length toxin protein. Furthermore, the product as recovered from E. coli, and from other recombinant expression hosts, is an inactive single chain peptide or if expression hosts produce a processed, active polypeptide it is not a toxin. Endopeptidase activity of LH423/A, as assessed by the current in vitro peptide cleavage assay, is wholly dependent on activation of the recombinant molecule between residues 430 and 454 by trypsin. Other proteolytic enzymes that cleave between these two residues are generally also suitable for activation of the recombinant molecule. Trypsin cleaves the peptide bond C-terminal to Arginine or C-terminal to Lysine and is suitable as these residues are found in the 430-454 region and are exposed (see FIG. 12).
 The recombinant poiypeptides of the invention are potential therapeutic agents for targeting to cells expressing the relevant substrate but which are not implicated in effecting botulism. An example might be where secretion of neurotransmitter is inappropriate or undesirable or alternatively where a neuronal cell is hyperactive in terms of regulated secretion of substances other than neurotransmitter. In such an example the function of the Hc domain of the native toxin could be replaced by an alternative targeting sequence providing, for example, a cell receptor ligand and/or translocation domain.
 One application of the recombinant polypeptides of the invention will be as a reagent component for synthesis of therapeutic molecules, such as disclosed in WO-A-94/21300. The recombinant product will also find application as a non-toxic standard for the assessment and development of in vitro assays for detection of functional botulinum or tetanus neurotoxins either in foodstuffs or in environmental samples, for example as disclosed in EP-A-0763131.
 A further option is addition, to the C-terminal end of a polypeptide of the invention, of a peptide sequence which allows specific chemical conjugation to targeting ligands of both protein and non-protein origin.
 In yet a further embodiment an alternative targeting ligand is added to the N-terminus of polypeptides of the invention. Recombinant LHN derivatives have been designated that have specific protease cleavage sites engineered at the C-terminus of the LC at the putative trypsin sensitive region and also at the extreme C-terminus of the complete protein product. These sites will enhance the activational specificity of the recombinant product such that the dichain species can only be activated by proteolytic cleavage of a more predictable nature than use of trypsin.
 The LHN enzymatically produced from native BoNT/A is an efficient immunogen and thus the recombinant form with its total divorce from any full length neurotoxin represents a vaccine component. The recombinant product may serve as a basal reagent for creating defined protein modifications in support of any of the above areas.
 Recombinant constructs are assigned distinguishing names on the basis of their amino acid sequence length and their Light Chain (L-chain, L) and Heavy Chain (H-chain, H) content as these relate to translated DNA sequences in the public domain or specifically to SEQ ID NO: 2 and SEQ ID NO: 20. The ‘LH’ designation is followed by ‘/X’ where ‘X’ denotes the corresponding clostridial toxin serotype or class, e.g. ‘A’ for botulinum neurotoxin type A or ‘TeTx’ for tetanus toxin. Sequence variants from that of the native toxin polypeptide are given in parenthesis in standard format, namely the residue position number prefixed by the residue of the native sequence and suffixed by the residue of the variant.
 Subscript number prefixes indicate an amino-terminal (N-terminal) extension, or where negative a deletion, to the translated sequence. Similarly, subscript number suffixes indicate a carboxy terminal (C-terminal) extension or where negative numbers are used, a deletion. Specific sequence inserts such as protease cleavage sites are indicated using abbreviations, e.g. Factor Xa is abbreviated to FXa. L-chain C-terminal suffixes and H-chain N-terminal prefixes are separated by a ‘/’ to indicate the predicted junction between the L and H-chains. Abbreviations for engineered ligand sequences are prefixed or suffixed to the clostridial L-chain or H-chain corresponding to their position in the translation product.
 Following this nomenclature,
 LH423/A=SEQ ID NO: 2, containing the entire L-chain and 423 amino acids of the H-chain of botulinum neurotoxin type A;
2LH423/A=a variant of this molecule, containing a two amino acid extension to the N-terminus of the L-chain;
2L/2H423/A=a further variant in which the molecule contains a two amino acid extension on the N-terminus of both the L-chain and the H-chain;
2LFXa/2H423/A=a further variant containing a two amino acid extension to the N-terminus of the L-chain, and a Factor Xa cleavage sequence at the C-terminus of the L-chain which, after cleavage of the molecule with Factor Xa leaves a two amino acid N-terminal extension to the H-chain component; and
2LFXa/2H423/A-IGF-1=a variant of this molecule which has a further C-terminal extension to the H-chain, in this example the insulin-like growth factor 1 (IGF-1) sequence.
 There now follows description of specific embodiments of the invention, illustrated by drawings in which:
FIG. 1 shows a schematic representation of the domain structure of botulinum neurotoxin type A (BoNT/A);
FIG. 2 shows a schematic representation of assembly of the gene for an embodiment of the invention designated LH423/A;
FIG. 3 is a graph comparing activity of native toxin, trypsin generated “native” LHN/A and an embodiment of the invention designated 2LH423/A (Q2E,N26K,A27Y) in an in vitro peptide cleavage assay;
FIG. 4 is a comparison of the first 33 amino acids in published sequences of native toxin and embodiments of the invention;
FIG. 5 shows the transition region of an embodiment of the invention designated L/4H423/A illustrating insertion of four amino acids at the N-terminus of the HN sequence;
 amino acids coded for by the Eco 47 III restriction endonuclease cleavage site are marked and the HN sequence then begins ALN . . . ;
FIG. 6 shows the transition region of an embodiment of the invention designated LFXa/3H423/A illustrating insertion of a Factor Xa cleavage site at the C-terminus of the L-chain, and three additional amino acids coded for at the N-terminus of the H-sequence; the N-terminal amino acid of the cleavage-activated HN will be cysteine;
FIG. 7 shows the C-terminal portion of the amino acid sequence of an embodiment of the invention designated LFXa/3H423/A-IGF-1, a fusion protein; the IGF-1 sequence begins at position G882;
FIG. 8 shows the C-terminal portion of the amino acid sequence of an embodiment of the invention designated LFXa/3H423/A-CtxA14, a fusion protein; the C-terminal CtxA sequence begins at position Q882;
 FIG. 9 shows the C-terminal portion of the amino acid sequence of an embodiment of the invention designated LFXa/3H423/A-ZZ, a fusion protein; the C-terminal ZZ sequence begins at position A890 immediately after a genenase recognition site (underlined);
 show schematic representations of manipulations of
FIGS. 10 & 11 polypeptides of the invention; FIG. 10 shows LH423/A with N-terminal addition of an affinity purification peptide (in this case GST) and C-terminal addition of an lg binding domain; protease cleavage sites R1, R2 and R3 enable selective enzymatic separation of domains; FIG. 11 shows specific examples of protease cleavage sites R1, R2 and R3 and a C-terminal fusion peptide sequence;
FIG. 12 shows the trypsin sensitive activation region of a polypeptide of the invention;
FIG. 13 shows Western blot analysis of recombinant LH107/B expressed from E.coli; panel A was probed with anti-BoNT/B antiserum; Lane 1, molecular weight standards; lanes 2 & 3, native BoNT/B; lane 4, immunopurified LH107/B; panel B was probed with anti-T7 peptide tag antiserum; lane 1, molecular weight standards; lanes 2 & 3, positive control E.coli T7 expression; lane 4 immunopurified LH107/B.
 The sequence listing that accompanies this application contains the following sequences:
 A 2616 base pair, double stranded gene sequence (SEQ ID NO: 1) has been assembled from a combination of synthetic, chromosomal and polymerase-chain-reaction generated DNA (FIG. 2). The gene codes for a polypeptide of 871 amino acid residues corresponding to the entire light-chain (LC, 448 amino acids) and 423 residues of the amino terminus of the heavy-chain (Hc) of botulinum neurotoxin type A. This recombinant product is designated the LH423/A fragment (SEQ ID NO: 2).
 Construction of the Recombinant Product
 The first 918 base pairs of the recombinant gene were synthesised by concatenation of short oligonucleotides to generate a coding sequence with an E. coli codon bias. Both DNA strands in this region were completely synthesised as short overlapping oligonucleotides which were phosphorylated, annealed and ligated to generate the full synthetic region ending with a unique KpnI restriction site. The remainder of the LH423/A coding sequence was PCR amplified from total chromosomal DNA from Clostridium botulinum and annealed to the synthetic portion of the gene.
 The internal PCR amplified product sequences were then deleted and replaced with the native, fully sequenced, regions from clones of C. botulinum chromosomal origin to generate the final gene construct. The final composition is synthetic DNA (bases 1-913), polymerase amplified DNA (bases 914-1138 and 1976-2616) and the remainder is of C. botulinum chromosomal origin (bases 1139-1975). The assembled gene was then fully sequenced and cloned into a variety of E.coli plasmid vectors for expression analysis.
 Expression of the Recombinant Gene and Recovery of Protein Product
 The DNA is expressed in E. coli as a single nucleic acid transcript producing a soluble single chain polypeptide of 99,951 Daltons predicted molecular weight. The gene is currently expressed in E. coli as a fusion to the commercially available coding sequence of glutathione S-transferase (GST) of Schistosoma japonicum but any of an extensive range of recombinant gene expression vectors such as pEZZ18, pTrc99, pFLAG or the pMAL series may be equally effective as might expression in other prokaryotic or eukaryotic hosts such as the Gram positive bacilli, the yeast P. pastoris or in insect or mammalian cells under appropriate conditions.
 Currently, E. coli harbouring the expression construct is grown in Luria-Bertani broth (L-broth pH 7.0, containing 10 g/l bacto-tryptone, 5 g/l bacto-yeast extract and 10 g/l sodium chloride) at 37° C. until the cell density (biomass) has an optical absorbance of 0.4-0.6 at 600 nm and the cells are in mid-logarithmic growth phase. Expression of the gene is then induced by addition of isopropylthio-β-D-galactosidase (IPTG) to a final concentration of 0.5 mM. Recombinant gene expression is allowed to proceed for 90 min at a reduced temperature of 25° C. The cells are then harvested by centrifugation, are resuspended in a buffer solution containing 10 mM Na2HPO4, 0.5 M NaCl, 10 mM EGTA, 0.25% Tween, pH 7.0 and then frozen at −20° C. For extraction of the recombinant protein the cells are disrupted by sonication. The cell extract is then cleared of debris by centrifugation and the cleared supernatant fluid containing soluble recombinant fusion protein (GST-LH423/A) is stored at −20° C. pending purification. A proportion of recombinant material is not released by the sonication procedure and this probably reflects insolubility or inclusion body formation. Currently we do not extract this material for analysis but if desired this could be readily achieved using methods known to those skilled in the art.
 The recombinant GST-LH423/A is purified by adsorption onto a commercially prepared affinity matrix of glutathione Sepharose and subsequent elution with reduced glutathione. The GST affinity purification marker is then removed by proteolytic cleavage and reabsorption to glutathione Sepharose; recombinant LH423/A is recovered in the non-adsorbed material.
 Construct Variants
 A variant of the molecule, LH423/A (Q2E,N26K,A27Y) (SEQ ID NO: 26) has been produced in which three amino acid residues have been modified within the light chain of LH423/A producing a polypeptide containing a light chain sequence different to that of the published amino acid sequence of the light chain of BoNT/A.
 Two further variants of the gene sequence that have been expressed and the corresponding products purified are 23LH423/A (Q2E,N26K,A27Y) (SEQ ID NO: 4) which has a 23 amino acid N-terminal extension as compared to the predicted native L-chain of BoNT/A and 2LH423/A (Q2E,N26K,A27Y) (SEQ ID NO: 6) which has a 2 amino acid N-terminal extension (FIG. 4).
 In yet another variant a gene has been produced which contains a Eco 47 III restriction site between nucleotides 1344 and 1 345 of the gene sequence given in (SEQ ID NO: 1). This modification provides a restriction site at the position in the gene representing the interface of the heavy and light chains in native neurotoxin, and provides the capability to make insertions at this point using standard restriction enzyme methodologies known to those skilled in the art. It will also be obvious to those skilled in the art that any one of a number of restriction sites could be so employed, and that the Eco 47 III insertion simply exemplifies this approach. Similarly, it would be obvious for one skilled in the art that insertion of a restriction site in the manner described could be performed on any gene of the invention. The gene described, when expressed, codes for a polypeptide, L/4H4231A (SEQ ID NO: 10), which contains an additional four amino acids between amino acids 448 and 449 of LH423/A at a position equivalent to the amino terminus of the heavy chain of native BoNT/A.
 A variant of the gene has been expressed, LFXa/3H423/A (SEQ ID NO: 12), in which a specific proteolytic cleavage site was incorporated at the carboxy-terminal end of the light chain domain, specifically after residue 448 of L/4H423/A. The cleavage site incorporated was for Factor Xa protease and was coded for by modification of SEQ ID NO: 1. It will be apparent to one skilled in the art that a cleavage site for another specified protease could be similarly incorporated, and that any gene sequence coding for the required cleavage site could be employed. Modification of the gene sequence in this manner to code for a defined protease site could be performed on any gene of the invention.
 Variants of LFXa/3H423/A have been constructed in which a third domain is present at the carboxy-terminal end of the polypeptide which incorporates a specific binding activity into the polypeptide.
 Specific examples described are:
 (1) LFXa/3H423/A-IGF-1 (SEQ ID NO: 14), in which the carboxy-terminal domain has a sequence equivalent to that of insulin-like growth factor-1 (IGF-1) and is able to bind to the insulin-like growth factor receptor with high affinity;
 (2) LFXa/3H423/A-CtxA14 (SEQ ID NO: 16), in which the carboxy-terminal domain has a sequence equivalent to that of the 14 amino acids from the carboxy-terminus of the A-subunit of cholera toxin (CtxA) and is thereby able to interact with the cholera toxin B-subunit pentamer; and
 (3) LFXa/3H423/A-ZZ (SEQ ID NO: 18), in which the carboxy-terminal domain is a tandem repeating synthetic IgG binding domain. This variant also exemplifies another modification applicable to the current invention, namely the inclusion in the gene of a sequence coding for a protease cleavage site located between the end of the clostridial heavy chain sequence and the sequence coding for the binding ligand. Specifically in this example a sequence is inserted at nucleotides 2650 to 2666 coding for a genenase cleavage site. Expression of this gene produces a polypeptide which has the desired protease sensitivity at the interface between the domain providing HN function and the binding domain. Such a modification enables selective removal of the C-terminal binding domain by treatment of the polypeptide with the relevant protease.
 It will be apparent that any one of a number of such binding domains could be incorporated into the polypeptide sequences of this invention and that the above examples are merely to exemplify the concept. Similarly, such binding domains can be incorporated into any of the polypeptide sequences that are the basis of this invention. Further, it should be noted that such binding domains could be incorporated at any appropriate location within the polypeptide molecules of the invention.
 Further embodiments of the invention are thus illustrated by a DNA of the invention further comprising a desired restriction endonuclease site at a desired location and by a polypeptide of the invention further comprising a desired protease cleavage site at a desired location.
 The restriction endonuclease site may be introduced so as to facilitate further manipulation of the DNA in manufacture of an expression vector for expressing a polypeptide of the invention; it may be introduced as a consequence of a previous step in manufacture of the DNA; it may be introduced by way of modification by insertion, substitution or deletion of a known sequence. The consequence of modification of the DNA may be that the amino acid sequence is unchanged, or may be that the amino acid sequence is changed, for example resulting in introduction of a desired protease cleavage site, either way the polypeptide retains its first and second domains having the properties required by the invention.
FIG. 10 is a diagrammatic representation of an expression product exemplifying features described in this example. Specifically, it illustrates a single polypeptide incorporating a domain equivalent to the light chain of botulinum neurotoxin type A and a domain equivalent to the HN domain of the heavy chain of botulinum neurotoxin type A with a N-terminal extension providing an affinity purification domain, namely GST, and a C-terminal extension providing a ligand binding domain, namely an igG binding domain. The domains of the polypeptide are spatially separated by specific protease cleavage sites enabling selective enzymatic separation of domains as exemplified in the Figure. This concept is more specifically depicted in FIG. 11 where the various protease sensitivities are defined for the purpose of example.
 Assay of Product Activity
 The LC of botulinum neurotoxin type A exerts a zinc-dependent endopeptidase activity on the synaptic vesicle associated protein SNAP-25 which it cleaves in a specific manner at a single peptide bond. The 2LH423/A (Q2E,N26K,A27Y) (SEQ ID NO: 6) cleaves a synthetic SNAP-25 substrate in vitro under the same conditions as the native toxin (FIG. 3). Thus, the modification of the polypeptide sequence of 2LH423/A (Q2E,N26K,A27Y) relative to the native sequence and within the minimal functional LC domains does not prevent the functional activity of the LC domains.
 This activity is dependent on proteolytic modification of the recombinant GST-2LH423/A (Q2E,N26K,A27Y) to convert the single chain polypeptide product to a disulphide linked dichain species. This is currently done using the proteolytic enzyme trypsin. The recombinant product (100-600 μg/ml) is incubated at 37° C. for 10-50 minutes with trypsin (10 μg/ml) in a solution containing 140 mM NaCl, 2.7 mM KCl, 10 mM Na2HPO4, 1.8 mM KH2PO4, pH 7.3. The reaction is terminated by addition of a 100-fold molar excess of trypsin inhibitor. The activation by trypsin generates a disulphide linked dichain species as determined by polyacrylamide gel electrophoresis and immunoblotting analysis using polyclonal anti-botulinum neurotoxin type A antiserum.
2LH423/A is more stable in the presence of trypsin and more active in the in vitro peptide cleavage assay than is 23LH423/A. Both variants, however, are fully functional in the in vitro peptide cleavage assay. This demonstrates that the recombinant molecule will tolerate N-terminal amino acid extensions and this may be expanded to other chemical or organic moieties as would be obvious to those skilled in the art.
 As a further exemplification of this invention a number of gene sequences have been assembled coding for polypeptides corresponding to the entire light-chain and varying numbers of residues from the amino terminal end of the heavy chain of botulinum neurotoxin type B. In this exemplification of the disclosure the gene sequences assembled were obtained from a combination of chromosomal and polymerase-chain-reaction generated DNA, and therefore have the nucleotide sequence of the equivalent regions of the natural genes, thus exemplifying the principle that the substance of this disclosure can be based upon natural as well as a synthetic gene sequences.
 The gene sequences relating to this example were all assembled and expressed using methodologies as detailed in Sambrook J, Fritsch E F & Maniatis T (1989) Molecular Cloning: A Laboratory Manual (2nd Edition), Ford N, Nolan C, Ferguson M & Ockler M (eds), Cold Spring Harbor Laboratory Press, New York, and known to those skilled in the art.
 A gene has been assembled coding for a polypeptide of 1171 amino acids corresponding to the entire light-chain (443 amino acids) and 728 residues from the amino terminus of the heavy chain of neurotoxin type B. Expression of this gene produces a polypeptide, LH728/B (SEQ ID NO: 20), which lacks the specific neuronal binding activity of full length BoNT/B.
 A gene has also been assembled coding for a variant polypeptide, LH417/B (SEQ ID NO: 22), which possesses an amino acid sequence at its carboxy terminus equivalent by amino acid homology to that at the carboxy-terminus of the heavy chain fragment in native LHN/A.
 A gene has also been assembled coding for a variant polypeptide, LH107/B (SEQ ID NO: 24), which expresses at its carboxy-terminus a short sequence from the amino terminus of the heavy chain of BoNT/B sufficient to maintain solubility of the expressed polypeptide.
 Construct Variants
 A variant of the coding sequence for the first 274 bases of the gene shown in SEQ ID NO: 21 has been produced which whilst being a non-native nucleotide sequence still codes for the native polypeptide.
 Two double stranded, a 268 base pair and a 951 base pair, gene sequences have been created using an overlapping primer PCR strategy. The nucleotide bias of these sequences was designed to have an E.coli codon usage bias.
 For the first sequence, six oligonucleotides representing the first (5′) 268 nucleotides of the native sequence for botulinum toxin type B were synthesised. For the second sequence 23 oligonucleotides representing internal sequence nucleotides 691-1641 of the native sequence for botulinum toxin type B were synthesised. The oligonucleotides ranged from 57-73 nucleotides in length. Overlapping regions, 17-20 nucleotides, were designed to give melting temperatures in the range 52-56° C. In addition, terminal restriction endonuclease sites of the synthetic products were constructed to facilitate insertion of these products into the exact corresponding region of the native sequence. The 268 bp 5′ synthetic sequence has been incorporated into the gene shown in SEQ ID NO: 21 in place of the original first 268 bases (and is shown in SEQ ID NO: 27). Similarly the sequence could be inserted into other genes of the examples.
 Another variant sequence equivalentto nucleotides 691 to 1641 of SEQ ID NO: 21 and employing non-native codon usage whilst coding for a native polypeptide sequence, has been constructed using the internal synthetic sequence. This sequence (SEQ ID NO: 28) can be incorporated, alone or in combination with other variant sequences, in place of the equivalent coding sequence in any of the genes of the example.
 An exemplification of the utility of this invention is as a non-toxic and effective immunogen. The non-toxic nature of the recombinant, single chain material was demonstrated by intraperitoneal administration in mice of GST-2LH423/A. The polypeptide was prepared and purified as described above. The amount of immunoreactive material in the final preparation was determined by enzyme linked immunosorbent assay (ELISA) using a monoclonal antibody (BA11) reactive against a conformation dependent epitope on the native LHN/A. The recombinant material was serially diluted in phosphate buffered saline (PBS; NaCl 8 g/l, KCl 0.2 g/l, Na2HPO4 1.15 g/l, KH2PO4 0.2 g/l, pH 7.4) and 0.5 ml volumes injected into 3 groups of 4 mice such that each group of mice received 10, 5 and 1 micrograms of material respectively. Mice were observed for 4 days and no deaths were seen.
 For immunisation, 20 μg of GST-2LH423/A in a 1.0 ml volume of water-in-oil emulsion (1:1 vol:vol) using Freund's complete (primary injections only) or Freund's incomplete adjuvant was administered into guinea pigs via two sub-cutaneous dorsal injections. Three injections at 10 day intervals were given (day 1, day 10 and day 20) and antiserum collected on day 30. The antisera were shown by ELISA to be immunoreactive against native botulinum neurotoxin type A and to its derivative LHN/A. Antisera which were botulinum neurotoxin reactive at a dilution of 1:2000 were used for evaluation of neutralising efficacy in mice. For neutralisation assays 0.1 ml of antiserum was diluted into 2.5 ml of gelatine phosphate buffer (GPB; Na2HPO4 anhydrous 10 g/l, gelatin (Difco) 2 g/l, pH 6.5-6.6) containing a dilution range from 0.5 μg (5×10−6 g) to 5 picograms (5×10−12 g). Aliquots of 0.5 ml were injected into mice intraperitoneally and deaths recorded over a 4 day period. The results are shown in Table 1 and Table 2. It can clearly be seen that 0.5 ml of 1:40 diluted anti-GST-2LH423/A antiserum can protect mice against intraperitoneal challenge with botulinum neurotoxin in the range 5 pg-50 ng (1-10,000 mouse LD50; 1 mouse LD50=5 pg).
 Expression of Recombinant LH107/B in E. coli.
 As an exemplification of the expression of a nucleic acid coding for a LHN of a clostridial neurotoxin of a serotype other than botulinum neurotoxin type A, the nucleic acid sequence (SEQ ID NO: 23) coding for the polypeptide LH107/B (SEQ ID NO: 24) was inserted into the commercially available plasmid pET28a (Novogen, Madison, Wis., USA). The nucleic acid was expressed in E. coli BL21 (DE3) (New England BioLabs, Beverley, Mass., USA) as a fusion protein with a N-terminal T7 fusion peptide, under IPTG induction at 1 mM for 90 minutes at 37° C. Cultures were harvested and recombinant protein extracted as described previously for LH423/A.
 Recombinant protein was recovered and purified from bacterial paste lysates by immunoaffinity adsorption to an immobilised anti-T7 peptide monoclonal antibody using a T7 tag purification kit (New England bioLabs, Beverley, Mass., USA). Purified recombinant protein was analysed by gradient (4-20%) denaturing SDS-polyacrylamide get electrophoresis (Novex, San Diego, Calif., USA) and western blotting using polyclonal anti-botulinum neurotoxin type antiserum or anti-T7 antiserum. Western blotting reagents were from Novex, immunostained proteins were visualised using the Enhanced Chemi-Luminescence system (ECL) from Amersham. The expression of an anti-T7 antibody and anti-botulinum neurotoxin type B antiserum reactive recombinant product is demonstrated in FIG. 13.
 The recombinant product was soluble and retained that part of the light chain responsible for endopeptidase activity.
 The invention thus provides recombinant polypeptides useful inter alia as immunogens, enzyme standards and components for synthesis of molecules as described in WO-A-94/21300.