WO2009001058A1 - Modified chondroitinase polypeptides - Google Patents

Abstract

This invention relates to modified forms of bacterial glycosaminoglycan (GAG) polysaccharide lyase enzymes, such as chondroitinase ABC, that possess a reduced number of active N-glycosylation motifs. These modified enzymes can be synthesised and secreted by mammalian cells in an active form and may be useful, for example, in the treatment of nervous system damage, including human spinal cord injury.

Description

Modified Chondroitinase Polypeptides

This invention relates to modified bacterial chondroitinases which are modified for expression and secretion in an active form in mammalian cells. This may be useful, for example, in the treatment of nervous system damage, including human spinal cord injury.

Spinal cord injuries leave large numbers of people permanently paralysed every year. Most patients are young when injured and, in the face of the well-known regenerative failure of the mammalian CNS, can be expected to live for several more decades with little prospect of recovery. Spinal injury causes a glial 'scar' to form in the spinal cord, and cells in this scar zone express molecules that are inhibitory to axon regrowth [Fawcett et al . , 2001; Fawcett & Asher, 1999; Morgenstern et al . (2002); Silver & Miller, 2004; Yiu & He, 2006] . Together with growth-inhibitory proteins released by degenerating myelin, these are thought to provide a significant brake on regenerative growth by injured axons, as a result of which functional recovery fails to take place. Measures designed to stimulate recovery have had modest success in rodent models, but have not yet made the major improvements hoped for [Ramer et al . , 2000; Silver & Miller, 2004; Ramer et al . , 2005; Thuret et al . , 2006; Bradbury & McMahon, 2006] .

An important contribution to axon growth-inhibition in the glial scar is made by chondroitin sulphate proteoglycans (CSPGs) in the extracellular matrix, such as NG2 [Dou & Levine, 1994; Levine & Nishiyama, 1996; Fawcett & Asher, 1999; Bovolenta & Fernaud-Espinosa, 2000; Silver & Miller, 2004; Tan et al . , 2006; Yiu & He, 2006]. These molecules have long sulphated glycosaminoglycan (GAG) chains attached to their central protein component, and the GAG chains are responsible for much of the inhibitory activity [Grimpe & Silver, 2004] . Chondroitinase ABC is a bacterial enzyme that degrades these inhibitory carbohydrate chains. The use of chondroitinase to promote axon regeneration, using rats as an experimental model, was first demonstrated by Moon et al . (2001) in the nigrostriatal tract and by Bradbury et al . (2002) in spinal cord injury (SCI) . Injection of bacterial chondroitinase ABC into mammalian CNS tissue in vivo was found to deplete GAG immunoreactivity surrounding an injury site [Pizzorusso et al . , 2002] and concomitantly generate carbohydrate 'stub' products which are absent from normal tissue [Moon et al . , 2001; Bradbury et al . , 2002] . Repeated injection of the chondroitinase ABC at the site of spinal injury was shown to promote extensive regeneration of both corticospinal and sensory axons, accompanied by significant functional motor recovery (Bradbury et al . , 2002) . These results have been confirmed and extended by others (Yick et al . , 2003; Caggiano et al . , 2005; Steinmetz et al . , 2005). Chondroitinase also enhances axon regeneration in combination with grafts of Schwann cells and/or olfactory ensheathing cells or neural stem/progenitor cells (Chau et al . , 2004; Ikegami et al . , 2005; Fouad et al . , 2005; Houle et al. , 2006) .

The improvement seen after chondroitinase treatment is generally considered to be due to the enzyme removing the block to regeneration by CSPGs present in the glial scar. In addition, chondroitinase may also promote recovery by other mechanisms. One is to promote sprouting of spared axons at some distance from an injury site, which can then find their way through or round the CSPG-depleted injury region (Corvetti & Rossi, 2005; Barritt et al . , 2006; Massey et al . , 2006) . Likewise, chondroitinase can increase the synaptic plasticity of undamaged neurons, possibly by removal of the CSPGs present in peri-neuronal nets (Pizzorusso et al . , 2002, 2006; Rhodes & Fawcett, 2004) . Such anatomical and synaptic plasticity can allow undamaged regions of CNS to take over the function of severed neurons and axons, and is particularly relevant to the case of SCI where two thirds of patients have incomplete lesions with axons surviving through the site of injury. Finally, the degradation of CSPGs by chondroitinase results in the generation of disaccharides that support neural repair by- inducing neurite outgrowth and preventing neuronal death and axonal collapse, both in vitro and in vivo (Rolls et al . , 2004) . Whatever the predominant mechanism, chondroitinase is clearly a promising treatment for spinal cord injury.

However, the use of chondroitinase in human CNS will require considerable modification of present techniques. Delivery of the enzyme by local injection into the spinal cord region is technically problematic, for several reasons. From the cerebrospinal fluid, the enzyme would be required to penetrate a considerable distance through the spinal cord parenchyma. Also, chronic infusion or repeated injections would be required to relieve axon growth-inhibition for the extended periods needed for functional recovery, especially in view of the limited stability of the enzyme (Chau et al . , 2004). Chronic delivery carries risks of tissue damage, infection, and immunogenicity. Moreover, there could be deleterious effects of widespread CSPG degradation, as CSPG is present normally in perineuronal nets (Pizzorusso et al . , 2002; Rhodes & Fawcett, 2004

The present inventors have discovered that bacterial chondroitinase

ABC can be modified without reducing its activity so that mammalian cells are able synthesize and secrete active enzyme. This may be useful, for example, for the transfection of neurons and/or glia at an injury site, so the cells secrete the chondroitinase and facilitate repair of neural tissue without the need for chronic infusion or repeated injections.

An aspect of the invention provides a chondroitinase polypeptide comprising a modified bacterial glycosaminoglycan (GAG) polysaccharide lyase sequence, said modified sequence having a reduced number of active N-glycosylation motifs relative to the unmodified sequence, and said chondroitinase polypeptide being secretable in an active form by mammalian cells.

Glycosaminoglycan (GAG) polysaccharide lyases are secreted bacterial lyase enzymes which degrade glycosaminoglycans , such as hyaluronan and chondroitin, in the extracellular matrix. GAG lyases recognize uronyl residues in polysaccharides and cleave their glycosidic bonds via a beta-elimination reaction to form a double bond between C-4 and C-5 of the non-reducing terminal uronyl residues of released products. Glycosaminoglycan (GAG) polysaccharide lyases include chondroitin AC lyases, chondroitin ABC lyases, xanthan lyases, and hyalurate lyases.

Glycosaminoglycan (GAG) polysaccharide lyases are also known as polysaccharide lyase family 8 and are listed in the NCBI Conserved Domain Database under the reference number cd01083.1.

Examples of Glycosaminoglycan (GAG) polysaccharide lyases are provided in table 3.

A chondroitinase polypeptide as described herein may comprise an amino acid sequence having at least 50% sequence identity to the sequence of a bacterial glycosaminoglycan (GAG) polysaccharide lyase listed in Table 3 and having a reduced number of active N-glycosylation motifs relative to said sequence. Sequence identity is described in more detail below.

The number of N-glycosylation motifs may be reduced in the amino acid sequence by mutating a residue in one or more N-glycosylation motifs which are present in the bacterial GAG lyase sequence.

The mutation of N-glycosylation motifs is described in more detail below. In preferred embodiments, the bacterial GAG lyase sequence is a chondroitinase ABC sequence or a chondroitinase AC sequence. Chondroitinase ABC sequences include the P. vulgaris sequence shown in SEQ ID NO: 1 and the Y. pestis sequence of database entry ZP_00797056.1 GI: 77634965. Chondroitinase AC sequences include the Flavobacterium columnare sequence of database entry AY912281.1 GI: 59800135 and the Pedobacter heparinus sequence of database entry U27583.1 GI: 1002524.

In some preferred embodiments, a chondroitinase polypeptide may comprise an amino acid sequence having at least 20% sequence identity to the sequence of SEQ ID NO: 1 or a fragment thereof having chondroitinase activity, the amino acid sequence having a reduced number of active N- glycosylation motifs relative to the sequence of SEQ ID NO: 1 or the fragment thereof , and the chondroitinase polypeptide being secretable in an active form by mammalian cells.

A fragment of the chondroitinase having chondroitinase activity may- comprise residues 282 to 960 of SEQ ID NO: 1. For example, a chondroitinase polypeptide may comprise an amino acid sequence having at least 20% sequence identity to the P. vulgaris Chondroitinase ABC sequence of residues 282 to 960 of SEQ ID NO: 1, said amino acid sequence having a reduced number of active N- glycosylation motifs relative to the sequence of residues 282 to 960 of SEQ ID NO: 1, and said chondroitinase polypeptide being secretable in an active form by mammalian cells.

Other suitable fragments may comprise residues 282 to 963 of SEQ ID NO: 1. A chondroitinase polypeptide may comprise an amino acid sequence having at least 20% sequence identity to the sequence of residues 282 to 963 of SEQ ID NO: 1 and having a reduced number of active N-glycosylation motifs relative to the sequence of residues 282 to 963 of SEQ ID NO: 1.

Other suitable fragments may comprise residues 25 to 1021 of SEQ ID NO: 1. A chondroitinase polypeptide may comprise an amino acid sequence having at least 20% sequence identity to the sequence of residues 25 to 1021 of SEQ ID NO: 1 and having a reduced number of active N-glycosylation motifs relative to the sequence of residues 25 to 1021 of SEQ ID NO: 1.

Reducing the number of N-glycosylation motifs in the amino acid sequence reduces the amount of glycosylation which is undergone by the chondroitinase polypeptide when expressed in mammalian cells.

The number of N-glycosylation motifs may be reduced in the amino acid sequence by mutating a residue in one or more N-glycosylation motifs which are present in the corresponding sequence of SEQ ID NO: 1 and SEQ ID NO: 3, for example, the sequence of residues 282 to 960, residues 282 to 963 or residues 25 to 1021 of SEQ ID NO: 1 or SEQ ID NO: 3. Mutation of a residue within a N-glycosylation motif as described herein prevents N-glycosylation of the motif during expression in a mammalian cell.

An N-glycosylation motif consists of the sequence N-X-S/T, where X is any amino acid.

An N-glycosylation motif may be mutated in the chondroitinase polypeptide by means of a conservative or semi-conservative substitution of the N or S/T residue of the sequence to produce a mutated N-glycosylation motif which is not glycosylated in mammalian cells.

A conservative substitution is an amino acid replacement that preserves the structure and functional properties of the native protein. Similarity matrices such as the PAM 250 or Blosum 45 are commonly used in the art to determine conservative or semi- conservative variation.

In some embodiments, the N residue of the N-glycosylation motif may be mutated, for example by substitution for another amino acid, such as Q, K, D or A.

In some embodiments, the S or T residue of the N-glycosylation motif may be mutated, for example by substitution for another amino acid, such as A, G or L.

The Proteus vulgaris chondroitinase ABC amino acid sequence which is shown in residues 25 to 1021 of SEQ ID NO: 1 or SEQ ID NO: 3 has N- glycosylation motifs at positions N72, N171, N174, N202, N282, N338, N345, N515, N603, N675, N751, N773 , N836, N856, N918, N963, N976.

The amino acid sequence between residues 282 and 960 of SEQ ID NO: 1 or SEQ ID NO: 3 has N-glycosylation motifs at positions N282, N338, N345, N515, N603, N675, N751, N773 , N836, N856 and N918.

The sequence between residues 282 and 963 of SEQ ID NO: 1 or SEQ ID NO: 3 has N-glycosylation motifs at positions N282, N338, N345, N515, N603, N675, N751, N773, N836, N856, N918 and N963.

However, the N-glycosylation motif at position N751 has unfavourable flanking sequences and is shown herein to be unglycosylated in mammalian cells.

One or more of these N-glycosylation motifs may be mutated in the chondroitinase polypeptides described herein. For example, a chondroitinase polypeptide may comprise an amino acid sequence having at least 20% sequence identity to residues 282 to 960 of SEQ ID NO: 1 which has a mutation in one or more N-glycosylation motifs selected from the group consisting of the N-glycosylation motifs at positions N282, N338, N345, N515, N675, N773 and N856 in SEQ ID NO: 1 or an amino acid sequence having at least 20% sequence identity to residues 282-963 or residues 25 to 1021 of SEQ ID NO: 1 which has a mutation in one or more N-glycosylation motifs selected from the group consisting of the N-glycosylation motifs at positions N282, N338, N345, N515, N675, N773, N856 and N963 in SEQ ID NO: 1.

More preferably, the amino acid sequence has a mutation at one, two or three N-glycosylation motifs selected from the group consisting of the N-glycosylation motifs at positions N282, N345, and N515 of SEQ ID NO: 1.

More preferably, the amino acid sequence has a mutation at one, two, three, four or all five N-glycosylation motifs selected from the group consisting of the N-glycosylation motifs at positions N282, N345, N336, N515 and N675 of SEQ ID NO: 1.

A chondroitinase polypeptide as described herein may comprise an amino acid sequence having a mutation in one or more, two or more, three or more, four or more or five or more N-glycosylation motifs of a bacterial GAG lyase sequence .

The mutation of the N-glycosylation motifs at N751 in SEQ ID NO: 1 are shown herein to have no effect on expression and secretion of active chondroitinase polypeptide. In some embodiments, this motif may be unmutated.

Examples of chondroitinase polypeptides include the Bl, X12, X30, Y13, AlO and Y133 mutants listed in Table 1.

A chondroitinase polypeptide catalyses the depolymerization of chondroitin sulfate (EC 4.2.2.4) to unsaturated disaccharides . In particular, the chondroitinase polypeptide degrades the sulphated glycosaminoglycan (GAG) chains of chondroitin sulfate proteoglycans (CSPGs) , such as NG2.

The chondroitinase polypeptide described herein is active under physiological conditions and is able to be synthesised and secreted in an active form by mammalian host cells. Secretion of active chondroitinase polypeptide as described herein may be useful in neuronal regeneration or repair, for example at a site of neural injury or damage, such as a spinal cord lesion.

The Proteus vulgaris chondroitinase ABC sequence (Genbank Ace No; AAB43331.1 GI: 1828877) is shown in SEQ ID NO: 1. Residues 1 to 24 represent the signal peptide and residues 25 to 1021 represent the active enzyme. A modified Proteus vulgaris chondroitinase ABC sequence comprising a mammalian signal peptide is shown in SEQ ID NO 3. The first 29 residues (here numbered -4 to 24) represent the signal peptide and residues 25 to 1021 represent the active enzyme.

Chondroitinase polypeptides described herein may comprise at least 20% sequence identity or sequence similarity with the sequence of residues 282 to 960, residues 282 to 963 or residues 25 to 1021 of SEQ ID NO: 1 or SEQ ID NO: 3 or other reference GAG lyase sequence. Preferably, the chondroitinase polypeptide comprises an amino acid sequence which shares greater than 30%, greater than 40%, greater than 50%, greater than 60%, greater than 70%, greater than 80%, greater than 85%, greater than 90%, greater than 95% or greater than 98% sequence identity or sequence similarity with the reference GAG lyase sequence.

Chondroitinase polypeptides described herein include mutations in one or more N-glycosylation motifs which are present in the sequence of a reference bacterial GAG lyase sequence, for example SEQ ID NO: 1 or a fragment thereof, such as residues 282 to 960, residues 282 to 963 or residues 25 to 1021 of SEQ ID NO: 1 or SEQ ID NO: 3. The sequence of the chondroitinase polypeptide may also differ from the reference bacterial GAG lyase sequence by one or more of addition, insertion, deletion or substitution of one or more amino acids at positions other than N-glycosylation motifs.

In some embodiments, up to about 5, 10, 15, 20, 30 or 40 amino acids may be altered in a chondroitinase polypeptide described herein. Such alterations may be caused by one or more of addition, insertion, deletion or substitution of one or more nucleotides in the encoding nucleic acid.

Preferably the sequence of the chondroitinase polypeptide does not differ from the sequence of residues 282 to 960 of SEQ ID NO: 1 or SEQ ID NO: 3 at residues H501, Y508, R560 and E653 as set out in SEQ ID NO : 1 or SEQ ID NO : 3.

Sequence similarity and identity are commonly defined with reference to the algorithm GAP (Wisconsin GCG package, Accelerys Inc, San Diego USA) . GAP uses the Needleman and Wunsch algorithm to align two complete sequences that maximizes the number of matches and minimizes the number of gaps. Generally, default parameters are used, with a gap creation penalty = 12 and gap extension penalty = 4. Use of GAP may be preferred but other algorithms may be used, e.g. BLAST (which uses the method of Altschul et al. (1990) J. MoI. Biol. 215: 405-410), FASTA (which uses the method of Pearson and Lipman (1988) PNAS USA 85: 2444-2448) , or the Smith-Waterman algorithm (Smith and Waterman (1981) J. MoI Biol. 147: 195-197), or the TBLASTN program, of Altschul et al . (1990) supra, generally employing default parameters. In particular, the psi-Blast algorithm (Nucl. Acids Res. (1997) 25 3389-3402) may be used. Sequence identity and similarity may also be determined using Genomequest™ software (Gene-IT, Worcester MA USA) .

Sequence comparisons are preferably made over the entire length of the relevant sequence described herein. In some preferred embodiments, a chondroitinase polypeptide described herein may comprise an amino acid sequence which is the sequence of SEQ ID NO:1 or a fragment thereof with a mutation in an N- glycosylation motif thereof, said mutation preventing N-glycosylation of the motif.

Suitable fragments of SEQ ID NO: 1 may comprise residues 282 to 960, residues 282 to 963 or residues 25 to 1021 of SEQ ID NO: 1.

One or more heterologous amino acids may be joined or fused to a chondroitinase polypeptide set out herein and a polypeptide may- comprise a polypeptide sequence as described above linked or fused to one or more heterologous amino acids. One or more heterologous amino acids may include sequences from a source other than a bacterial chondroitinase protein.

In preferred embodiments, the chondroitinase polypeptide described above is coupled to a signal sequence which directs secretion of the chondroitinase polypeptide. Preferably, the signal sequence is a eukaryotic signal sequence, which directs secretion of the chondroitinase polypeptide in eukaryotic cells. Many suitable eukaryotic signal sequences are known in the art. In some preferred embodiments, a mammalian signal sequence may be employed, for example the mouse MMP2 signal sequence shown in the first 29 residues of SEQ ID NO: 3 (residues -4 to 25) . Other suitable signal sequences include GDNF, prolactin, or immunoglobulin signal sequences.

In some embodiments, bacterial signal sequences, including the chondroitinase ABC signal sequence, may be employed.

Heterologous peptides might be attached at the N- or C-terminus of the chondroitinase polypeptide, for example to allow recognition by an antibody, or to direct secretion to a particular part of the cell such as the neuronal growth cone . The invention also encompasses nucleic acids, vectors and cells which are suitable for use in methods of producing a chondroitinase polypeptide, as described above.

Accordingly, a nucleic acid may comprise a nucleic acid sequence which encodes a chondroitinase polypeptide as described above.

A suitable nucleic acid may comprise a nucleotide sequence having at least 40% sequence identity to a reference bacterial GAG lyase nucleotide sequence, for example SEQ ID NO: 2, SEQ ID NO: 4 or a fragment thereof, such as a sequence comprising nucleotides 191-3181 of SEQ ID NO: 2 or nucleotides 297-3287 of SEQ ID NO: 4 , wherein the nucleotide sequence includes one or more mutations relative to the reference bacterial GAG lyase nucleotide sequence which reduce the number of N-glycosylation motifs in the encoded amino acid sequence relative to the amino acid sequence of the reference bacterial GAG lyase amino acid sequence, for example SEQ ID NO: 1 or SEQ ID NO: 3 or a fragment thereof, such as a sequence comprising residues 282 to 963 of SEQ ID NO: 1 or SEQ ID NO: 3.

A mutation which reduces the number of N-glycosylation motifs in the encoded amino acid sequence may be any mutation, for example the addition, insertion, deletion or substitution of one or more nucleotides, which leads to a substitution of the N or S/T residue of an N-glycosylation motif to produce a mutated N-glycosylation motif which is not glycosylated in mammalian cells. In some embodiments, the substitution may be a conservative substitution.

A suitable nucleic acid may share greater than 50% sequence identity, greater than 60% sequence identity, greater than 70%, greater than about 80%, greater than 90% or greater than 95% with the reference bacterial GAG lyase nucleotide sequence, for example all or part of the nucleic acid sequence from SEQ ID NO: 2 or SEQ ID NO: 4 as described above .

A suitable nucleic acid may have one or more of addition, insertion, deletion or substitution of one or more nucleotides in the sequence of the reference bacterial GAG lyase nucleotide sequence. For example, up to 10, 20, 30, 50, 100, 500 or 1000 nucleotides may be added, inserted, deleted or substituted as described.

In some embodiments, a nucleic acid may comprise one or more codons optimized for eukarytic expression, preferably mammalian cell expression. The optimisation of codons for expression in a particular host, for example by replacing preferred bacterial codons with codons preferred in mammalian cells, is well-known in the art. An optimised nucleic acid sequence may comprise the nucleic acid sequence from nucleotides 297-3287 of SEQ ID NO: 4 with one or more mutations which reduce the number of N-glycosylation motifs.

In some embodiments, the nucleic acid may comprise an optimized Kozak sequence for improved expression in a eukaryotic mammalian host cell. For example, the nucleic acid may comprise the kozak sequence which is shown in SEQ ID NO : 4.

The nucleotide sequence encoding the chondroitinase polypeptide may be operably linked to a heterologous regulatory sequence.

Suitable regulatory sequences to drive the expression of heterologous nucleic acid coding sequences in expression systems are well known in the art and include constitutive promoters, for example viral promoters such as CMV or SV40, and inducible promoters, such as Tet- on, ecdysone or tamoxifen controlled promoters. In some preferred embodiments, promoters specific to neurons (such as neuron-specific enolase) or to astrocytes (such as GFAP) or to other glial cells may be employed.

The term "heterologous" indicates that the gene/sequence of nucleotides in question or a sequence regulating the gene/sequence in question, is a recombinant sequence which has been introduced into a construct, vector or cell, artificially, using genetic engineering or recombinant means, i.e. by human intervention. Heterologous nucleotide sequences are sequences which do not naturally occur together in nature. Nucleotide sequences which are heterologous to a cell may be non-naturally occurring in cells of that type, variety or species (i.e. exogenous or foreign) or may be sequences which are non- naturally occurring in that sub-cellular or genomic environment of the cells or may be sequences which are non-naturally regulated in the cells i.e. operably linked to a regulatory element other than the natural regulatory element for the sequence .

Nucleic acid sequences and constructs as described above may be comprised within a vector. Suitable vectors can be chosen or constructed, containing appropriate regulatory sequences, including promoter sequences, terminator fragments, polyadenylation sequences, enhancer sequences, marker genes and other sequences as appropriate. Preferably, the vector contains appropriate regulatory sequences to drive the expression of the nucleic acid in mammalian, in particular human, cells. A vector may also comprise sequences, such as origins of replication and selectable markers, which allow for its selection and replication in bacterial hosts such as E. coli and/or in eukaryotic cells .

Vectors may be plasmids, viral e.g. 'phage, or phagemid, as appropriate. For further details see, for example, Molecular Cloning: a Laboratory Manual: 3rd edition, Russell et al . , 2001, Cold Spring Harbor Laboratory Press. Many known techniques and protocols for manipulation of nucleic acid, for example in preparation of vectors and the introduction of DNA into cells are described in detail in Protocols in Molecular Biology, Second Edition, Ausubel et al . eds . John Wiley & Sons, 1992.

In some embodiments, the vector is a viral vector suitable for expression in mammalian cells, in particular human cells. Suitable viral vectors include adenovirus, adeno-associated virus (AAV) , for example AAV serotype 2 virus, retrovirus, lentivirus, recombinant adenovirus, 'gutless' adenovirus, herpes simplex virus, and poliovirus vectors.

A viral vector may be packaged into a viral particle comprising one or more capsid proteins prior to transfection of host cells.

In some embodiments, the viral vector may be packaged into a heterologous viral particle. For example, a lentiviral vector may be pseudotyped with a rabies glycoprotein, such as rabies-G envelope protein, which provides for transduction of neurons in vivo (Mazarakis et al Human Molecular Genetics (2001) 10 2109-2121)

A nucleic acid or vector as described herein may be introduced into a host cell. This may occur, for example, in vitro or in vivo or ex vivo.

Suitable host cells include neural cells, such as neurons and glial cells, preferably mammalian neurons and glial cells, including Schwann cells, oligodendrocytes, and astrocytes. Bacterial cells such as E. coli may also be useful as host cells for some purposes, for example in the production of nucleic acid for use as described herein.

Techniques for the introduction of nucleic acid into cells are well established in the art and any suitable technique may be employed, in accordance with the particular circumstances. For eukaryotic cells, suitable techniques may include DEAE-dextran, polyethyleneimine, electroporation, liposome-mediated transfection and transduction using retrovirus or other virus, e.g. adenovirus, AAV, lentivirus or vaccinia. For bacterial cells, suitable techniques may include transformation of chemically competent cells, electroporation and transfection using bacteriophage.

Marker genes such as antibiotic resistance or sensitivity genes may be used in identifying clones containing nucleic acid of interest, as is well known in the art. In some embodiments, a marker such as green fluorescent protein may be used to identify cells expressing the introduced nucleic acid. Marker genes may be comprised in the same vector as the nucleic acid encoding the chondroitinase polypeptide, or on separate vectors, which may be co-transfected into a host cell.

The introduced nucleic acid may be on an extra-chromosomal vector within the cell or the nucleic acid may be integrated into the genome of the host cell. Integration may be promoted by inclusion of sequences within the nucleic acid or vector which promote recombination with the genome, in accordance with standard techniques.

The introduction may be followed by expression of the nucleic acid to produce the encoded chondroitinase polypeptide.

In some embodiments, host cells (which may include cells actually transformed although more likely the cells will be descendants of the transformed cells) may be cultured in vitro under conditions for expression of the nucleic acid, so that the encoded chondroitinase polypeptide is produced. When an inducible promoter is used, expression may require the activation of the inducible promoter.

A host cell which expresses the nucleic acid and secretes the expressed chondroitinase polypeptide may be isolated and/or purified. In some embodiments, host cells, in particular glial cells, may be obtained from an individual, preferably an individual requiring treatment for neural damage. Nucleic acid encoding the chondroitinase polypeptide may be then be introduced into the host cells ex vivo using standard transfection or transduction techniques and, optionally, cultured, isolated and/or purified, prior to implantation or administration to the individual, for example at a site of neural damage .

In some embodiments, nucleic acid may be introduced into a cell in vivo. For example, a nucleic acid or vector may be administered to the individual such that one or more cells of the individual incorporate the nucleic acid or vector. Preferably, the nucleic acid or vector is administered at or adjacent to the site of neural tissue damage to facilitate uptake of the nucleic acid or vector by neural cells e.g. neurons and glial cells, at the damage site. Neurons and/or glial cells, including astrocytes, oligodendrocytes and Schwann cells at the damage site express the nucleic acid and secrete the chondroitinase polypeptide. The secreted chondroitinase degrades extracellular CSPGs, for example in glial scars and this may be useful in the repair of neural damage and the improvement of nerve function.

A method of producing a chondroitinase polypeptide as described herein may comprise; introducing a nucleic acid encoding a chondroitinase polypeptide as described above into a host cell, for example a mammalian neural cell, and; expressing said nucleic acid in the cell to produce the chondroitinase polypeptide.

Nucleic acid may be expressed by allowing the host cells (which may include cells actually transformed, although more likely the cells will be descendants of the transformed cells) to grow under conditions under which they express the introduced nucleic acid to produce the chondroitinase polypeptide. Following expression, the chondroitinase polypeptide is secreted by the host cells.

As described above, the nucleic acid may be introduced into the host cell in vitro, in vivo or ex vivo. Preferably the cell produces and secretes the chondroitinase polypeptide in vivo at the site of nervous tissue damage.

Further aspects of the invention provide a host cell comprising a vector or nucleic acid construct as described herein, and a method of producing a host cell as described herein may comprise introducing a nucleic acid as described herein into the host cell.

A host cell may contain a nucleic acid sequence encoding a chondroitinase polypeptide as a result of the introduction of the nucleic acid sequence into an ancestor cell.

A range of host cells suitable for the production of chondroitinase polypeptides are known in the art. Suitable host cells may include mammalian cells, in particular neural cells, for example neurons, astrocytes, oligodendrocytes, oligodendrocyte precursor cells, olfactory ensheathing cells, and Schwann cells. For therapeutic applications, the cells are preferably human cells. As described above, cells which express and secrete the chondroitinase polypeptide may be transplanted to a site of neural tissue damage or cells in situ at site of neural tissue damage may be transfected with a nucleic acid encoding the chondroitinase polypeptide.

Aspects of the invention provide a nucleic acid, viral particle, vector or cell as described herein for use in a method of treatment, for example, a method of treating neural tissue damage, in particular spinal cord injury, and a pharmaceutical composition comprising a nucleic acid, viral particle, vector or cell as described herein. Other aspects of the invention provide the use of a nucleic acid, viral particle, vector or cell in the manufacture of a medicament for use in treating neural tissue damage comprising administering a nucleic acid, viral particle, vector or cell as described herein to an individual in need thereof .

The methods described herein may be useful in promoting nerve repair at a site of neural damage, for example in the treatment of spinal cord injury.

In some embodiments, the damaged or injured nerve is within the central nervous system (CNS) . The CNS includes the brain, the spinal cord, and the retina. It includes neurons whose cell bodies lie within, or have a primary synapse in, the brain or spinal cord, including neurons of origin of the corticospinal tract. The CNS also includes axons of sensory neurons whose cell bodies lie outside the CNS.

The neural injury or damage may be a spinal cord injury, for example an injury caused by assault, accident, tumour, intervertebral disc or bone abnormality, or surgery, e.g. surgery for spinal problems and/or surgery to remove tumours .

In preferred embodiments, the nucleic acid, vector or cell is administered directly to the site of neural tissue damage, for example by injection or infusion. Suitable infusion techniques are known in the art and commonly used in therapy (see, e.g., Rosenberg et al . , New Eng. J. of Med., 319:1676, 1988).

The term "treatment", as used herein in the context of treating a condition, pertains generally to treatment and therapy, whether of a human or an animal (e.g. in veterinary applications), in which some desired therapeutic effect is achieved, for example, the inhibition of the progress of the condition, and includes a reduction in the rate of progress, a halt in the rate of progress, amelioration of the condition, and cure of the condition.

Whether it is a nucleic acid, vector, viral particle or cell according to the present invention that is to be given to an individual, administration is preferably in a "therapeutically effective amount" this being sufficient to show benefit to the individual. The actual amount administered, and rate and time-course of administration, will depend on the nature and severity of what is being treated. Prescription of treatment, e.g. decisions on dosage etc, is within the responsibility of medical practitioners.

Pharmaceutical compositions according to the present invention, and for use in accordance with the present invention, may include, in addition to active ingredient, a pharmaceutically acceptable excipient, carrier, buffer, stabiliser or other materials well known to those skilled in the art. The precise nature of the carrier or other material will depend on the route of administration, which may be oral, or by injection, e.g. cutaneous, subcutaneous or intravenous. A pharmaceutical composition may thus comprise a nucleic acid encoding a chondroitinase polypeptide or a vector or cell comprising such a nucleic acid as described herein and a pharmaceutically acceptable excipient .

A pharmaceutical composition may be produced by admixing or formulating the nucleic acid, viral particle, vector, or cell with a pharmaceutically acceptable excipient.

Formulations suitable for administration (e.g. by infusion), include aqueous and non-aqueous isotonic, pyrogen-free, sterile injection solutions which may contain anti-oxidants, buffers, preservatives, stabilisers, bacteriostats, and solutes which render the formulation isotonic with the blood of the intended recipient; and aqueous and non-aqueous sterile suspensions which may include suspending agents and thickening agents. Examples of suitable isotonic vehicles for use in such formulations include Sodium Chloride Injection, Ringer's Solution, or Lactated Ringer's Injection. Suitable vehicles can be found in standard pharmaceutical texts, for example, Remington's Pharmaceutical Sciences, 18th edition, Mack Publishing Company, Easton, Pa., 1990.

For intravenous, cutaneous or subcutaneous injection, or injection at the site of affliction, the active ingredient will be in the form of a parenterally acceptable aqueous solution which is pyrogen-free and has suitable pH, isotonicity and stability. Those of relevant skill in the art are well able to prepare suitable solutions using, for example, isotonic vehicles as described above. Preservatives, stabilisers, buffers, antioxidants and/or other additives may be included, as required.

Pharmaceutical compositions comprising nucleic acids for use in gene therapy can be stored and administered in a sterile physiologically- acceptable carrier, where the nucleic acid is dispersed in conjunction with any agents which aid in the introduction of the nucleic acid into cells .

The term "pharmaceutically acceptable" as used herein pertains to compounds, materials, compositions, and/or dosage forms which are, within the scope of sound medical judgement, suitable for use in contact with the tissues of a subject (e.g., human) without excessive toxicity, irritation, allergic response, or other problem or complication, commensurate with a reasonable benefit/risk ratio. Each carrier, excipient, etc. must also be "acceptable" in the sense of being compatible with the other ingredients of the formulation.

Administration of nucleic acid, vectors or cells in vivo can be effected in one dose, continuously or intermittently (e.g., in divided doses at appropriate intervals) throughout the course of treatment. The determination of the most effective means and dosage of administration is within the remit of the medical practitioner and will vary with the formulation used for therapy, the purpose of the therapy, the target cell being treated, and the subject being treated. Single or multiple administrations can be carried out with the dose level and pattern being selected by the treating physician.

Neural cells expressing the chondroitinase polypeptide may be administered directly at the site of neural damage by cell-therapy techniqes. The administered cells at the site of neural damage produce and secrete the chondroitinase protein locally. In some embodiments, the neural cells may be contained within an implant or nerve guide which is positioned at a site of neural tissue damage. In some embodiments, the implant may be fixed in position. Any convenient technique may be employed, for example, the implant may be sutured or glued in place. Suitable implants include, for example collagen gel implants .

For gene therapy to be practical, it is desirable to employ a nucleic acid transfer method that: (1) directs the therapeutic sequence into specific target cell types (e.g. neurons, or glial cells), (2) is highly efficient in mediating uptake of the therapeutic polynucleotide into the target cell population, and (3) is suited for use in vivo for therapeutic application.

Nucleic acid encoding the chondroitinase polypeptide may be delivered using any of a variety of techniques. Current techniques may be divided into three major groups: transfection with a viral vector, such as a replication-defective retroviral vector (Cornetta et al, Hum. Gene. Ther. 2:215 (1991), adenoviral vector (Rosenfeld et al, Cell 68:143 (1992)), or adenoassociated virus,- fusion with a lipid, for example in a liposome or immunoliposome (Litzinger and Huang, Biochem. Biophys . Acta 1113:201 (1992); Gao and Huang, Biochem. Biophys . Res. Commun. 179:280 (1991) Wang and Huang, Proc . Natl. Acad. Sci. USA 84:7851 (1987); Trubetskoy et al, Biochem. Biophys. Acta 1131:311 (1992)); and DNA transfer with a cation such as polyethyleneimine or polylysine. The selection of which technique to use depends upon the particular circumstances and is within the remit of skilled persons in the field.

Any suitable delivery method may be used to deliver nucleic acid as described herein to cells at a site of neural tissue damage. In some embodiments, direct physical application of naked or liposome encapsulated nucleic acid to cells at the site of neural tissue damage may be preferred.

Various further aspects and embodiments of the present invention will be apparent to those skilled in the art in view of the present disclosure. All documents mentioned in this specification are incorporated herein by reference in their entirety.

"and/or" where used herein is to be taken as specific disclosure of each of the two specified features or components with or without the other. For example "A and/or B" is to be taken as specific disclosure of each of (i) A, (ii) B and (iii) A and B, just as if each is set out individually herein.

Unless context dictates otherwise, the descriptions and definitions of the features set out above are not limited to any particular aspect or embodiment of the invention and apply equally to all aspects and embodiments which are described.

Certain aspects and embodiments of the invention will now be illustrated by way of example and with reference to the figures and tables described below. Figures 1 and 2 show recombinant chondroitinase ABC is active in vitro, after multi-site mutagenesis to correct mutations as described in the methods section below.

Figure 1 shows products of 35S-labelled in vitro transcription and translation (IVTT) on SDS-PAGE. Lane 1 shows a truncated product made by a plasmid with an internal stop codon. Lane 2 shows a full-length product after correction of the stop codon. L represents Luciferase (from control mRNA) . The double band in each lane is apparently a result of initiation of protein synthesis at an internal AUG which is an artifact of in vitro translation.

Figure 2 shows in vitro activity assay using the Morgan-Elson reaction. Left panel, calibration curve for commercial Chondroitinase ABC (0 to 2 mU) . Right panel, results of chondroitinase-ABC activity assay on products of IVTT reactions with chondroitinase clones. Clone 02 has a Leu745-Pro mutation and is inactive. Correction of the proline mutation, as in clones B5 and C4, generates enzyme activity. B5, C4, Chondroitinase ABC clones: C4 has the fully normalised sequence, and B5 additionally has one potential N-glycosylation site mutated.

Figure 3 shows an SDS-PAGE of IVTT reactions with microsomes to assess glycosylation. IVTT reactions were labeled with biotinylated lysine and incubated with canine microsomes. After SDS-PAGE and blotting, the products were detected with streptavidin-linked HRP. The lower arrow indicates unglycosylated chondroitinase-ABC; the upper arrow indicates the highest glycosylated band (from clone C4) , which is not produced without microsomes. The glycosylated band migrates faster in products of clones with several mutations of N-glycosylation sites. The grey bar denotes a non-specific band, masked out for clarity, which obscures the glycosylated band from clones X12 and X30. Figure 4 shows proteinase K protection after IVTT reaction with microsomes. The lower arrow indicates unglycosylated chondroitinase- ABC; the upper arrow, the glycosylated form. Only the glycosylated form is protected from Proteinase K, indicating that it has been internalised in the microsomes.

Figures 5 to 7 show western blots of medium from non-transfected and chondroitinase-transfected Neu7 cells, for the proteoglycan NG2. Each lane is labelled according to the clone used for transfection; those marked "+Ch'ase" were digested with commercial chondroitinase in vitro before SDS-PAGE.

Figure 5 shows, in lanes 1-4, positive and negative controls. In medium from untransfected or GFP-transfected cells, NG2 appears largely as a characteristic 'smear¹ as expected due to the GAG chains, and this is all converted to core protein by digestion with commercial chondroitinase (lanes 1 and 3). After transfection with modified chondroitinase AlO, there is a reduction of the GAG smear and compensating intensification of the core protein band. After transfection with unmodified chondroitinase (clone C4*, in this case with the original bacterial signal sequence) , there is no such digestion of the GAG chains.

Figure 6 shows a similar experiment to Figure 5 in which the NG2 also appears to have been digested by protease in the medium, showing complete removal of the immunoreactive smear by commercial chondroitinase (lane 4) or by transfection with clone Y13 (lanes 1 and 2) . (In this case the medium was collected from 16 to 22 h post- transfection, and the western blot was probed with Chemicon antibody against immuno-purified NG2. )

Figure 7 shows results from conditioned medium from untransfected Neu7 cells placed onto COS7 cells from 24 to 48 hrs after transfection with the indicated clones . Clones B5 and C4 show no secreted chondroitinase activity, but clone Y13 shows complete digestion of the CSPG smear to core band. COS7 cells do not produce NG2 , which may be why the digestion was complete in this experiment but incomplete in figure 5.

Figure 8 shows detachment of Neu7 cells following transfection with the modified chondroitinase clone Y13. All were photographed after 48h in culture. Cell rounding and detachment is not seen using a similar construct that lacks a signal sequence for secretion (d) , nor with GFP (a,b), indicating that the effect is a consequence of chondroitinase secretion.

Figure 9 shows western blots of Neu7 conditioned medium (a source of CSPGs) placed on transfected COS7 cells, probed with antibody 1B5 against the carbohydrate 'stub' epitope that represents the residue of chondroitinase digestion. Figure 9 (a) shows that medium incubated with GFP-transfected control cells shows little immunoreactivity (lane 2) , but digestion with commercial chondroitinase in vitro generates extensive reactivity (lane 1) . Lane 3 shows medium incubated with COS7 cells transfected with chondroitinase AlO, and shows reactivity as great as that produced by chondroitinase in vitro. (*) Saturated bands which have bleached before autoradiography. Figure 9 (b) shows the same experiment repeated with additional clones. Neu7 medium for control lanes 1 and 2 was not incubated with COS7 cells.

Figure 10 shows western blots prepared as in Figure 9, showing that chondroitinase activity is routinely produced by several mutated genes with 3 or more selected glycosylation sites changed. The upper panels show 'stub' immunoreactivity as in Figure 9; the lower panels show NG2 immunoreactivity as in Figures 5-7.

Figure 11 shows western blots as in Figure 10, showing the activity of transfected chondroitinase mutants Y13 and Y133 relative to unmodified sequence C4 in glial cell lines SCTM and Neu7. Figure 12 shows western blot lanes, probed for 'stub' immunoreact- ivity, showing the activity of transfected chondroitinase mutant AlO (with a signal sequence from human prolactin replacing the previous signal sequence; see Methods) in a primary astrocyte cell culture.

Figure 13 shows western blot lanes, probed for NG2 , of Neu7 conditioned medium that was placed on COS7 cells after transfection with the chondroitinase mutant AlO coupled to different signal sequences (see Methods) . Chondroitinase activity is revealed by degradation of NG2 glycan as in Figures 5-7. It is produced by AlO with three different signal sequences, but not by the negative control (C4-Csig) .

Experiments Materials and methods

The sequence of the gene for Proteus vulgaris chondroitinase ABC was reported by Ryan et al (1994) [Entrez accession number AAB43331, gil828877, SEQ ID NO: 2] and confirmed by Prabhakar et al . (2005a).

The encoded sequence was also confirmed from the protein crystal structure by Huang et al (2003) . These studies showed that an independent sequence reported by Sato et al . (1994) contained several errors .

A cDNA for Proteus vulgaris chondroitinase ABC, in a prokaryotic expression vector, was obtained from Glyko Inc. Sequencing revealed that this construct had the prokaryotic signal sequence removed, and contained two mutations, at nucleotide 1671 (C to T, glutamine to stop codon) and at nucleotide 2458 (T to C, leucine-745 to proline) . Apart from these codons, the clone encoded the same sequence reported by Ryan et al (1994) . We corrected the two mutations by site-directed mutagenesis, and added a eukaryotic signal sequence from mouse matrix metalloprotease 2 (GenBank accession no. NM008610 = gi47271505) (Reponen et al . , 1992), to direct enzyme secretion from transfected cells. An optimized Kozak sequence was also inserted to allow recognition by eukaryotic ribosomes and to maximize protein yield. We further changed some of the codons that are unfavourable for translation by eukaryotic ribosomes, replacing them with those used more frequently by eukaryotic cells. The resulting coding sequence is set out in SEQ ID NO: 4 and the sequence of the encoded protein is set out in SEQ ID NO: 3.

The numbering of nucleotides and codons in the chondroitinase coding region (nucl. 297 onwards) is as in Sato et al . (1994); the section that was frame-shifted in Sato et al . (1994) is marked. The sequence up to nucl. 297 is replaced by a Notl site (underlined: used for cloning into pcDNA3.1) overlapping a Kozak initiation sequence (bold, purple) , then the coding sequence for mouse MMP-2 signal sequence, which is joined to nucl. 297 of the chondroitinase ABC sequence. (Nucls 297-300, and the encoded Ala-Ala sequence, comprise overlap of identical residues representing the cleavage site of the signal sequence in both MMP-2 and chondroitinase ABC.) From then on the sequence is identical to that of the P. vulgaris chondroitinase ABC gene, except for individual nucleotides in bold where synonymous substitutions have been made to alter restriction sites or to substitute codons more frequently used in mammalian genes. Underlined sections labelled Gl to G9 indicate matches to oligonucleotides used to introduce mutations in N-glycosylation sites.

This modified cDNA was subcloned into the eukaryotic expression vector pcDNA 3.1 (Invitrogen) , in which transcription is directed by the viral CMV promoter, which directs high-level expression in a wide range of eukaryotic cells.

Site-directed mutagenesis

Mutagenesis was carried out using the QuikChange Multi -Site-Directed Mutagenesis kit (Stratagene) . All mutagenesis was carried out using constructs inserted in pcDNA 3.1. Primers containing the desired mutations are shown in SEQ. ID. NO.5. They were designed where possible to insert or delete a restriction site to allow easy identification of mutant clones. The primers were modified with a 5' phosphate and PAGE-purified to improve mutation efficiency. All clones were sequenced to confirm successful mutagenesis. Each construct was assayed using the TNT system to assess the effect of the mutation on enzyme activity. This allows us to distinguish the direct effect of the mutation on enzyme activity separately from effects due to glycosylation inside the cell.

Variants of clone AlO were made in which the MMP-2 signal sequence was replaced with signal sequences from the following genes: 1) original bacterial chondroitinase [Ryan et al, 1994 see above; Entrez ref.no. AAB43331, gi : 1828877] 2) rat GDNF [Lin, L. F. et al (1993) Science 260, 1130-1132 Entrez ref . no. NM_019139, gi : 9506720]

3) human prolactin [Jungnickel, B. and Rapoport, TM (1995) . Cell 71, 489-503; Entrez ref. no. BC088370; gi : 56971865]

4) mouse immunoglobulin kappa [Coloma, M. J. et al (1992) . J. Imm. Methods 152, 89-104]

In vitro transcription-translation

In vitro transcription/translation (IVTT) reactions were carried out with rabbit reticulocyte lysate in a coupled reaction with T7 polymerase, using the TNT Quick Coupled Transcription/Translation kit (Promega) . Each 25 μl reaction contained 1 μg of plasmid. Labelled reactions also included 1 μl biotinylated lysine (Transcend) or 1 μl 35S-methionine (Redivue L-methionine, 37 MBq/mmol, Amersham) . Reactions to assess glycosylation also contained 1 μl of canine microsomes (Promega) . The samples were incubated at 3O⁰C for 90mins. The products of the IVTT reactions were then separated by sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) , as follows :

35S-labeled samples: 1-2 μl samples were run on Nupage 4-12% bis/tris gradient gels (Invitrogen) , fixed in methanol/ acetic acid, and incubated in Amplify (Amersham) for 30 min prior to drying. The dried gels were then exposed to X-ray film for 2 h.

Biotin-labeled samples: 1-2 μl samples of IVTT reactions labeled with biotinylated lysine were run on 10% Tris/glycine gels, transferred to nitrocellulose membrane using a semi-dry-blot (Invitrogen) , and then probed with strepavidin-linked horse radish peroxidase (Promega) , prior to development using chemiluminescence (Promega) . All chemiluminescence products were detected using chemiluminsescence film (Amersham) .

Protease protection assay (Schmidt-Rose & Jentsch, 1997) : The translation mixture was brought to 10 mM CaC12 and chilled on ice. Aliquots of 10 μl were incubated with proteinase K, 30 μg/ml (Roche) , in the presence or absence of 1% Triton XlOO. Controls remained without proteinase and detergent. Proteolysis proceeded on ice for 60min and was stopped by adding 5 mM phenylmethylsulfonyl fluoride. After 10 min on ice, 50 μl of preheated sample buffer (95°C) was added and the sample was boiled for 15 min to inactivate the protease. The samples were then run on a gel .

Morgan-Elson reaction

This reaction measures chondroitinase activity by the N-acetylation of product disaccharides and subsequent reaction to give a coloured product [Morgan and Elson, 1934; Reissig et al . , 1955] . The procedure was adapted from standard protocols [Yamagata, T. et al J. Biol. Chem. , 243, 1523 (19680] . Initially the reaction resulted in unstable Colour intensity due to the formation of precipitate during the reaction, as previously noted by others [Takahashi et al . , 2003] . Therefore, activity could only be estimated qualitatively. We then refined the procedure by centrifuging at two stages to minimise the presence of precipitate at the end, thereby allowing spectrophotometry to be used to quantify activity.

The reaction was made up of 100 μl of 40 mM NaAc, 40 mM TrisCl pH 8.0, lOmg/ml chondroitin-6-sulphate (Sigma) , mixed with 20 μl enzyme sample (IVTT product or standard) . Proteus vulgaris chondroitinase ABC (Sigma) was used as standard. The reaction was incubated at 37°C for 20 min, then stopped by boiling for 1 minute. Potassium borate solution (0.8 M, pH 9.1, 100 μl) was added and the mixture was boiled for 7 mins . It was chilled on ice then centrifuged in a microfuge at 13000 rpm for 10 min. To the supernatant, 1 ml glacial acetic acid was added and mixed before centrifugation for a further 20-30 min. To 1 ml of supernatant, 0.4 ml of Morgan-Elson Reagent (1Og para- dimethylaminobenzaldehyde in 100 ml in glacial acetic acid with 12.5% concentrated HCl) was added and incubated at 37°C for 20 min. Product was measured by absorbance at 550 nm. (This wavelength gives higher and more consistent absorption values than the standard 585 nm.)

The Morgan-Elson reaction is not sensitive enough to detect enzyme secreted by transfected cells in unconcentrated medium, and does not work in concentrated conditioned medium. We determined this by mixing commercial chondroitinase ABC with Neu7 conditioned medium that had been concentrated ten-fold with a Centricon-50 unit: activity in the Morgan-Elson reaction was abolished. Therefore, other methods had to be used to assay secretion of the enzyme.

Cell transfection COS7 cells were grown in Dulbecco's modified Eagle's medium (DMEM) (Gibco-Invitrogen) plus 10% fetal bovine serum plus standard concentrations of penicillin, streptomycin and fungazone (CDMEM) . Neu7 cells [Smith-Thomas et al . , 1994; Fok-Seang et al . , 1995] were grown in CDMEM plus 10% horse serum. Transfection was performed in 25-cm² flasks with the cells ~60-70% confluent, with polyethyleneimine (PEI) and pAdVAntage vector to increase translation efficiency (Promega) , as follows. For two flasks, 10 μl of chondroitinase or GFP plasmid at 1 μg/μl in

Tris/EDTA, plus 1 μg of pAdVAntage, was mixed with 50 μl of 0.15 M NaCl, then rapidly mixed with 53 μl of 0.1 M PEI in 0.15 M NaCl, and incubated at room temperature for 30 min. The cells were rinsed with DMEM with ITS3+ (Sigma) . Each flask then received 0.5 ml of DMEM with ITS3+ and 55 μl of the DNA/PEI mixture, and was incubated at 37° for 4 hours. The mixture was then replaced with CDMEM and incubation continued. After 24 hr, the medium was replaced with DMEM with ITS3+, or (for COS7 cells) with medium of the same composition which had been incubated for 48 hours with just-confluent Neu7 cells (Neu7 conditioned medium) . This conditioned medium was collected after 48 hrs, centrifuged to remove detached cells, and concentrated 7- to -10- fold by centrifugation in a Centricon-50 unit (Millipore) , mixed with protease inhibitor cocktail (Sigma P8340) , and stored at -20° for subsequent electrophoresis.

In every round of transfections, one transfection was performed with GFP to assess transfection efficiency by fluorescence microscopy of the live cells, and in some cases by fixation and counterstaining with bisbenzamide (Hoechst 33258) . Transfection efficiencies were 17-30% for COS cells and 24-50% for Neu7 cells.

Western blots

Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) and western blotting were done by standard techniques. For detection of proteoglycans, samples were 50 μl of concentrated conditioned medium. Controls were similar samples from non-transfected cells, one of which was digested with chondroitinase ABC (Sigma, 20 mU) at 37° for 3 hrs. Samples were mixed with 10 μl of 5x non-reducing Laemmli sample buffer, boiled for 2 min, separated by SDS-PAGE (5% acrylamide gel) , and electroblotted in a Transblot Semi-dry Transfer Cell blotter (Bio-Rad) to Hybond-ECL membrane.

IVTT product samples consisted of 5 μl IVTT product, 6 μl 5x reducing Laemmli sample buffer, 19 μl chondroitinase buffer (40 rtiM NaAc, 40 mM TrisCl pH 8.0). Proteins were separated by SDS-PAGE (6% acrylamide gel) and transferred to Hybond-ECL membrane.

Antibodies were: mouse anti-NG2 (Santa Cruz sc33666 = mcAb 132.38), diluted 1:1000 and mouse anti-'stub' (Seikagaku, mcAb 1B5 , 1:250).

Membranes were incubated in 2% ECL Advanced Blocking Agent (Amersham) in Tris-buffered saline with 0.1% Tween-20 (TBS-T) at room temperature for 3-4 hours, then incubated with primary antibody in blocking solution, overnight at 4°C. Membranes were washed in TBS-T before incubation with secondary antibody (peroxidase-labelled anti-mouse) , 1:10,000 in blocking solution) for 1 hour at room temperature. Membranes were washed in TBS-T before reaction with ECL chemiluminescence detection reagent and visualisation on Hyperfilm (Amersham) .

Results

1. Recombinant chondroitinase ABC is active in vitro.

As described in Methods, we constructed a cDNA clone that encodes the natural sequence of chondroitinase ABC, with a Kozak sequence to initiate translation in mammalian cells, the mouse MMP2 signal sequence, and some codons modified to be more favourable for mammalian expression. This 'normalised' cDNA, subcloned into the expression vector pcDNA 3.1, was named clone C4. The sequence is set out in SEQ. ID.NOS 3 and 4.

These initial modifications were sufficient for active enzyme to be produced when testing the construct in an in vitro transcription/translation (IVTT) system using rabbit reticulocyte lysate. A band of the expected size was synthesised (Fig.l) and active enzyme could be detected by colorimetric assay (Fig.2) .

However, no secreted enzyme activity was detected 48 hours after transfection of COS 7 cells with the construct, even though RT-PCR confirmed that the cDNA was efficiently transcribed by the cells.

One explanation for the lack of detectable secreted product may be the presence of cryptic signals in the bacterial protein which are inappropriately recognized and processed by eukaryotic cells. Computer prediction programs enabled us to screen for such signals. We established the absence of the endoplasmic reticulum (ER) retention signal KDEL and signal sequences directing proteins to organelles, but we did find 17 predicted sites for N-glycosylation in the protein sequence. Such modifications could interfere with enzyme folding, impeding transit through the secretory pathway, and/or causing loss of activity by alteration of the active site or by sterically hindering substrate binding. Misfolding of a protein could also direct it to the proteosome for degradation.

The IVTT results are consistent with the possibility that abnormal N- glycosylation causes loss of chondroitinase activity. N-glycosylation does not occur with the rabbit reticulocyte lysate system, which produces active enzyme. However, when canine microsomes, which carry out N-glycosylation, were added to our IVTT, enzyme activity was completely suppressed.

We used a protein 3D-imaging program to identify the location of potential N-glycosylation sites in chondroitinase ABC, and found several in regions very likely to affect protein folding and/or substrate binding. We first performed multi-site-directed mutagenesis on the five sites marked (Table 1 and SEQ ID NOS: 5 to 13) . The glycosylation recognition sequences were abolished by conservative substitutions using either Asn->Gln if the residue was on the surface or Ser->Val, if it was buried or slightly buried.

2. Mutations at most N-glycosylation sites do not impair enzyme activity in vitro.

The different constructs were tested for activity by transcription and translation in the rabbit reticulocyte lysate system (IVTT) , and the products assayed by the colorimetric assay (Morgan-Elson reaction) . All constructs produced similar amounts of protein. This was demonstrated either by western blots using an antibody to the bacterial enzyme or by labeling the products with biotinylated lysine which was detected by strepavidin-linked HRP followed by chemiluminescence .

All the clones listed produced similar levels of activity in the

Morgan-Elson reaction, within a factor of two, like clones B5 and C4 (Fig.2) .

3. Mutations at N-glycosylation sites reduce glycosylation in vitro. Our hypothesis is that these sites are being inappropriately glycosylated in a eukaryotic system. To test this, and identify which of the sites are actually glycosylated, we compared different constructs in the reticulocyte lysate system in the presence of canine microsomes (endoplasmic reticulum preparation) , which incorporate and N-glycosylate the nascent polypeptide as in an animal cell.

Fig.3 shows that the unmodified enzyme (C4) is heavily glycosylated in the presence of microsomes. The unglycosylated form produced in the absence of microsomes is seen as a band of around 110 kDa. In the presence of microsomes, an additional band is seen at higher molecular weight, as expected for glycosylation. To confirm that it was internalised by the microsomes, a Proteinase K protection experiment was performed (Fig.4) . N-linked glycosylation occurs only within intact microsomes, and this assay makes use of the protection afforded the translocated protein domain by the microsomal membrane. Thus, translocated proteins are protected from exogenously added protease, and only the glycosylated form remains.

The glycosylated products of clones C4 and B5 comigrate (Fig.3), so the mutation of Asn-751 in B5 made no visible difference to the glycosylation. This indicates that Asn-751 is not detectably glycosylated, consistent with the sequence flanking Asn-751 which makes N-glycosylation unlikely [Petrescu et al . , 2004] . Conversely, the glycosylated products of clones with mutations at other sites show increased mobility on the gel (Fig.3) . Comparison of these products with the mutations they contain (Table 1) indicates that Asn-515, 345, 338 and 282 are all glycosylated as predicted. The glycosylation of Asn-515 is of particular significance because it lies in the cleft which constitutes the enzyme's active site.

Similar experiments with other constructs have shown that Asn-675, 856, and 963 are also glycosylated, but Asn-836 is not. Thus 7 out of 9 sites so far tested are glycosylated as predicted.

4. Chondroitinase ABC with reduced N-glycosylation can degrade NG2 proteoglycan from transfected Neu7 cells.

We have transfected these constructs, containing up to five mutations, into the astrocyte cell line Neu7. These cells were derived by transforming rat brain glia with the Neu oncogene [Fok-Seang et al . , 1995] , and secrete substantial amounts of the axon-inhibitory CSPG, NG2, into the medium and the matrix [Smith-Thomas et al . , 1994; Fidler et al . , 1999] . NG2 consists of a core protein of -290 kDa, and a proportion of the molecules are glycanated and therefore run as a higher-Mr smear on a gel [Levine & Nishiyama, 1996] . Chondroitinase activity would be demonstrated by reduction or loss of the glycanated smear, which is converted to the core protein. To assess whether our clones would generate chondroitinase activity, we transfected Neu7 cells with each clone, incubated them with serum-free medium from 24 hr to 72 hr after transfection, and analysed the conditioned medium by- western blotting for NG2.

The results are shown in Figs 5 to 7 and Table 2. There was no significant GAG degradation by cells transfected with GFP, or with clones C4, B5, Bl, or X12, which have from 0 to 3 glycosylation sites mutated. Conversely, GAG degradation was clearly seen following transfection with clones AlO or Y13, which have additional glycosylation mutations.

In one experiment, a higher-resolution view of the GAG digestion was fortuitously obtained as the NG2 core protein was cleaved by- endogenous protease activity in the medium (Fig.6) . The antibody detected a broad smear in control (GFP-transfected) medium which was removed by treatment with commercial chondroitinase . Medium from Y13- transfected cells showed the same appearance as chondroitinase-treated control medium .

In the same series of transfections, the Neu7 cells reproducibly detached from the substratum by 48 hrs after transfection with clone Y13, but not with GFP, nor with a Y13 clone lacking a signal sequence (Fig.8) . The same effect was produced by adding commercial chondroitinase ABC to the cultures. This effect may be due to degradation of the matrix, as a consequence of chondroitinase secretion, in a batch of Neu7 cells which were also secreting large amounts of proteases. However other batches of Neu7 cells did not show this effect, demonstrating that chondroitinase synthesis does not have a toxic effect on the cells.

5. Chondroitinase ABC with reduced N-glycosylation is secreted from transfected COSl cells.

The preceding results showed that abolishing selected glycosylation sites resulted in chondroitinase activity in transfected astrocytes in culture. However, the sensitivity of the assays was limited because the cells do not produce much CSPG for the first 48 hrs after transfection. Moreover, it did not exclude the possibility that the enzyme remained cell-associated and acted during the secretion of the CSPG rather than within the medium. We therefore transfected the same constructs into COS7 cells, which do not synthesise CSPGs, and added conditioned medium from newly confluent Neu7 cells as a reliable source of CSPGs. After 24-48 hrs, this medium was then analysed by- western blotting for NG2 as before (Fig.9) . In addition, western blots were probed with the antibody 1B5 which recognizes the carbohydrate 'stub' epitope remaining after chondroitinase has cleaved off the CSPG chains (Fig.9). This offers greater sensitivity because it detects the appearance of a novel product rather than the diminution of a substrate.

The western blots (Figs.9 & 10) show that several of the modified clones produce as much stub reactivity as in vitro treatment with commercial chondroitinase ABC. The results parallel those with antibody against NG2 (above and Fig.10) but as expected are more sensitive; thus clones Bl and X12 produced little or no visible degradation of NG2, but generated distinct stub immunoreactivity. Activity of the enzymes following transfection correlates with the number of glycosylation sites which have been ablated (Table 2) .

A range of mutant chondroitinase constructs were also transfected into cell lines representing glial cells of the nervous system, viz. SCTM cells (Schwann cell line) and Neu7 cells (astrocyte cell line) and incubated in Neu7 conditioned medium as a source of CSPGs. Active chondroitinase was secreted from both SCTM cells and Neu7 cells (Fig.11) . Active chondroitinase was also secreted following transfection of primary astrocytes from rat brain (Fig.12).

The level of secretion of active chondroitinase was found to be further improved by mutagenesis of site Asn-675, generating clone Y133, which is the most active mutant chondroitinase identified so far (Figs. 10 and 11) .

All these constructs used the signal sequence from the mouse MMP2 gene to direct secretion, but there is evidence that signal sequences show differential activity in different cell types. Therefore, for one of the more active mutants (AlO) , alternative signal sequences were substituted, from the MMP2, GDNF, immunoglobulin kappa, and prolactin genes. The resulting constructs were transfected into COS cells and incubated in Neu7 conditioned medium for 24-48 hrs . The medium was then analysed by western blotting for NG2. Several of the various signal sequences were shown to direct secretion of active chondroitinase from cells transfected with our mutagenised genes (Fig. 13) .

Little is known about how a bacterial protein is trafficked in a eukaryotic cell. The data herein shows that bacterial chondroitinase is inappropriately glycosylated by eukaryotic cells and that this has a detrimental effect on secretion of active enzyme.

Mutations of many N-glycosylation sites are permissible without impairing enzyme activity. However, a construct equivalent to X12 plus a Thr-Ala substitution at Thr-338 was inactive in the in vitro assay, whereas a mutation at Asn-336, eliminating the same N-glycosylation site in clone AlO, was well tolerated.

An in vitro assay was also useful to show whether the enzyme could enter the secretory pathway. In the presence of microsomes, a protein destined for secretion is internalised and N-glycosylated, and this was observed to happen with our chondroitinase constructs. However, a positive result in this assay does not guarantee that the enzyme will be adequately secreted by transfected cells. All clones, including the native sequence C4, showed internalisation and glycosylation of a proportion of their products in microsomes in vitro, whereas some of them failed to produce detectable activity in medium from transfected cells .

We have shown that ablation of strategic glycosylation sites results in secretion of active enzyme from transfected mammalian cells.

First, transfection of an astrocyte cell line, Neu7, with two of our modified constructs led to the degradation of the matrix, accompanied by the removal of the GAG chains from NG2 , the main inhibitory- component of the matrix. These results were consistent with active chondroitinase being produced and secreted. This conclusion was further supported by experiments on COS7 cells (which do not produce NG2 nor other CSPGs) . COS7 cells were transfected with several of the modified cDNAs, and conditioned medium containing CSPGs was incubated with transfected cells. Western blots showed that GAG chains were cleaved from NG2 molecules, and that the stub epitope was generated. The latter assay was more sensitive because it depended on detection of a new product rather than diminution of a substrate. It also conclusively showed that the chondroitinase was secreted, as the only source of CSPG substrate was the conditioned medium added to the transfected cells.

Recently, Cafferty et al . (2007) have reported production of Proteus chondroitinase ABC in transgenic mice, using the GFAP promoter and the original bacterial signal sequence to express it in astrocytes, and demonstrating activity in the injured CNS by staining with a stub antibody. However the stub antibody can detect small amounts of chondroitinase product, and indeed it could have detected products of intracellular chondroitinase activity. Moreover, the astrocytic expression was not sufficient to allow corticospinal axons to regenerate beyond the lesion, unlike the injection experiments of

Bradbury et al . (2002), and neuronal expression was not attempted; it remains to be seen whether the bacterial sequence will be adequate for secretion from axons. As we have shown that mutations of glycosylation sites can greatly enhance secretability, it is clearly important to use these mutants in future transgenic experiments for optimal expression, to improve the efficacy of chondroitinase as a treatment for spinal cord injury.

Curinga et al . (2007) have recently reported expression of a different bacterial enzyme, chondroitinase AC, in mammalian cells, using an immunoglobulin signal sequence and adenovirus vector. However the specific activity of the secreted product was very low and it was ineffective in vivo.

Bacterial chondroitinase ABC has shown considerable promise in animal models as a treatment for spinal cord injury. It not only promotes regeneration of injured axons and functional recovery following spinal injury, but also increases plasticity so new axons can grow and take over the function of damaged axons. The strategy of making cells at the site of injury secrete the enzyme will have two major advantages: it delivers the enzyme to the precise place it is needed, and it circumvents the requirement of repeated injections.

The above data shows that ablation of strategic glycosylation sites results in secretion of active enzyme from transfected mammalian cells. Therefore, it is now possible to transfect it into cells such as Schwann cells which could be grafted into the CNS to form a bridge across the injury site. We have previously produced Schwann cell lines secreting GDNF and shown that bridge grafts of these promoted axon regeneration in a rat model of Parkinson's disease (Wilby et al . , 1999) . The modified chondroitinase gene could also be expressed from viral vectors to target it to neurons and/or glia in the injured CNS (Murray & Fischer, 2001) . C4 B5 Bl X12 X30 Y13 AlO Y133

Clone no:

Positions changed:

Asn-751 (N-Q) - + + + + + + +

Asn-515 (S-A) + + + + + +

Asn-345 (N-Q) - - - + + + +

Asn-336 (N-Q) - - + -

Asn-282 (N-K, or N-Q) N-K N-K N-Q N-K

Asn-675 - - - +

The table indicates whether each clone had a mutation at each of 5 Asn-X- (Ser/Thr) sites. Mutations were Asn-Gln (N-Q), Asn-Lys (N-K), or Ser-Ala (S-A; at position +2 relative to the Asn) , as indicated.

Table 1: Mutations of N-glycosylation sites

Clone no: C4 B5 Bl X12 X30 Y13 AlO Y133

No. of positions changed: 0 1 2 3 3 4 5 5

Activity after transfection:

(Neu7 cells) - (+/-) nd + + + (COS7 cells) ( +/-) + ++ ++ ++ ++ +++

The table summarises the level of chondroitinase activity observed in media from transfected cells, either Neu7 cells generating endogenous CSPGs, or COS7 cells provided with Neu7 conditioned medium as a source of CSPGs. They were scored using both NG2 glycan degradation, and appearance of stub immunoreactivity. (nd, not determined) .

Table 2: Results of transfections

Table 3 : Examples of bacterial GAG lyases

References

Barritt AW et al.(2006) J.Neurosci. 26:10856-67.

Bovolenta P et al . (2000) Progress in Neurobiology 61: 113-132.

Bradbury EJ et al (2006) Nature Reviews Neuroscience 7, 644-653. Bradbury, E.J et al (2002) Nature 416: 589-590.

Caggiano AO et al (2005). J. Neurotrauma . 22 (2) : 226-239.

Chau, CH. et al (2004) FASEB J. 18: 194-196.

Corvetti, L. & Rossi, F. (2005) J. Neurosci. 25:7150-8

Curinga GM et al [2007 Mar 29; Epub] . J Neurochem. [PMID: 17394547] de Ia Cueva-Mendez, G. et al (2003) EMBO J. 22: 246-51.

Dempski RE et al (2002) Current Opinion in Chemical Biology 6, 844-850

Dou C-L & Levine JM (1994) J. Neurosci. 14, 7616-7628.

Fawcett, J. W. et al (1999) Brain Res. Bull. (1999) 49: 377-391.

Fawcett JW et al (eds.), 'Brain Damage, Brain Repair¹ (Oxford Univ. Press, 2001)

Fidler PS et al (1999) J. Neurosci. 19, 8778-8788.

Fok-Seang J et al (1995), Brain Research 689, 207-223.

Fouad K et al (2005) J. Neurosci. 25, 1169-1178.

Gahmberg CG et al (1996) Trends in Biochemical Science 21, 308-311. Grimpe, B. et al (2004) J. Neurosci. 24: 1393-1397.

Houle JD et al (2006 July). J. Neurosci. 26: 7405 - 7415.

Ikegami T et al (2005 Dec.) Eur J Neurosci. 22 (12) : 3036-3046.

Levine JM et al (1996) Perspectives in Developmental Neurobiology 3,

245-259. Lowrie DB et al (1999) Nature 400: 269-71.

Massey JM et al (2006) J. Neurosci. 26 (16) : 4406-4414.

Moon LDF et al (2001) Nature Neuroscience 4, 465-466.

Moon LDF et al (2003) J. Neurosci. Res. 71: 23-37.

Morgan WTJ & Elson LA (1934) Biochem. J. 28, 988-995. Morgenstern DA et al (2002) Progress in Brain Research 137, 313-332.

Murray M & Fischer I (2001) The Neuroscientist 7: 28-41.

Petrescu AJ et al (2004) Glycobiology 14:103-114.

Pizzorusso T et al (2002) Science 298, 1248-1251.

Pizzorusso T et al (2006) PNAS 103, 8517-8522. Prabhakar V et al (2005a). Biochem J. 386: 103-112.

Prabhakar V et al (2005b) Biochem. J. 390: 395-405.

Ramer MS et al (2000) Spinal Cord 38, 449-472.

Ramer LM et al (2005) Spinal Cord 43, 134-161. Reissig JL et al (1955) J. Biol. Chem. 217, 959-966.

Reponen P. et al (1992). J. Biol. Chem. 267, 7856-7862.

Rhodes K & Fawcett J (2004) J. Anat . 204, 33-48.

Rolls,A. et al (2004). Eur. J Neurosci . 20:1973-1983.

Ryan MJ et al (1994) Patent WO 94/25567. Sato N et al (1994) Appl . Microbiol. Biotechnol . 41, 39-46.

Schmidt-Rose T and Jentsch,T.J. (1997) PNAS 94; 7633-7638.

Silver, J. S. et al (2004) Nature Rev. Neurosci. 5: 146-161.

Smith-Thomas LC et al (1994) J. Cell Sci. 107, 1687-1695.

Steinmetz MP et al (2005) J. Neurosci. 25:8066-8076. [Erratum p. 8590] Takahashi T et al (2003) Anal. Biochem. 322:257-263.

Tan AM et al (2006) J. Neurosci. 26, 4729-4739.

Thuret S et al (2006) Nat. Rev. NS. 7, 628-643.

Wilby MJ et al (1999). J. Neurosci. 19, 2301-2312.

Wilson, R. L. et al (2000) J. Immunol. Methods 234: 137-47. Yiu G & He Z (2006) Nature Reviews Neuroscience 7, 617-627.

Yick LW et al (2003) Exp Neurol. 182:160-8.

Yamagata,T. et al (1968) J. Biol. Chem. 243: 1536-40.

Sequences

1 mpifrftala mtlgllsapy namaatsnpa fdpknlmqse iyhfaqnnpl adfssdknsi

61 ltlsdkrsim gnqsllwkwk ggssftlhkk livptdkeas kawgrsstpv fsfwlynekp 121 idgyltidfg eklistseaq agfkvkldft gwravgvsln ndlenremtl natntssdgt 181 qdsigrslga kvdsirfkap snvsqgeiyi drimfsvdda ryqwsdyqvk trlsepeiqf 241 hnvkpqlpvt penlaaidli rqrlinefvg geketnlale enisklksdf dalnihtlan 301 ggtqgrhlit dkqiiiyqpe nlnsqdkqlf dnyvilgnyt tlmfnisray vlekdptqka 361 qlkqmyllmt khlldqgfvk gsalvtthhw gyssrwwyis tllmsdalke anlqtqvyds 421 llwysrefks sfdmkvsads sdldyfntls rqhlalllle pddqkrinlv ntfshyitga 481 ltqvppggkd glrpdgtawr hegnypgysf pafknasqli yllrdtpfsv gesgwnnlkk 541 amvsawiysn pevglplagr hpfnspslks vaqgyywlam saksspdktl asiylaisdk 601 tqnestaifg etitpaslpq gfyafnggaf gihrwqdkmv tlkayntnvw sseiynkdnr 661 ygryqshgva qivsngsqls qgyqqegwdw nrmqgattih lplkdldspk phtlmqrger 721 gfsgtssleg qygmmafdli ypanlerfdp nftakksvla adnhlifigs ninssdknkn 781 vettlfqhai tptlntlwin gqkienmpyq ttlqqgdwli dsngngylit qaekvnvsrq 841 hqvsaenknr qptegnfssa widhstrpkd asyeymvfId atpekmgema qkfrenngly 901 qvlrkdkdvh iildklsnvt gyafyqpasi edkwikkvnk paivmthrqk dtlivsavtp 961 dlnmtrqkaa tpvtinvtin gkwqsadkns evkyqvsgdn teltftsyfg ipqeiklspl 1021

SEQ ID NO: 1 (AAB43331.1 Gl:1828877)

1 ggaattccat cactcaatca ttaaatttag gcacaacgat gggctatcag cgttatgaca 61 aatttaatga aggacgcatt ggtttcactg ttagccagcg tttctaagga gaaaaataat

121 gccgatattt cgttttactg cacttgcaat gacattgggg ctattatcag cgccttataa

181 cgcgatggca gccaccagca atcctgcatt tgatcctaaa aatctgatgc agtcagaaat

241 ttaccatttt gcacaaaata acccattagc agacttctca tcagataaaa actcaatact

301 aacgttatct gataaacgta gcattatggg aaaccaatct cttttatgga aatggaaagg 361 tggtagtagc tttactttac ataaaaaact gattgtcccc accgataaag aagcatctaa

421 agcatgggga cgctcatcta cccccgtttt ctcattttgg ctttacaatg aaaaaccgat

481 tgatggttat cttactatcg atttcggaga aaaactcatt tcaaccagtg aggctcaggc

541 aggctttaaa gtaaaattag atttcactgg ctggcgtgct gtgggagtct ctttaaataa

601 cgatcttgaa aatcgagaga tgaccttaaa tgcaaccaat acctcctctg atggtactca 661 agacagcatt gggcgttctt taggtgctaa agtcgatagt attcgtttta aagcgccttc

721 taatgtgagt cagggtgaaa tctatatcga ccgtattatg ttttctgtcg atgatgctcg

781 ctaccaatgg tctgattatc aagtaaaaac tcgcttatca gaacctgaaa ttcaatttca

841 caacgtaaag ccacaactac ctgtaacacc tgaaaattta gcggccattg atcttattcg

901 ccaacgtcta attaatgaat ttgtcggagg tgaaaaagag acaaacctcg cattagaaga 961 gaatatcagc aaattaaaaa gtgatttcga tgctcttaat attcacactt tagcaaatgg

1021 tggaacgcaa ggcagacatc tgatcactga taaacaaatc attatttatc aaccagagaa

1081 tcttaactcc caagataaac aactatttga taattatgtt attttaggta attacacgac

1141 attaatgttt aatattagcc gtgcttatgt gctggaaaaa gatcccacac aaaaggcgca

1201 actaaagcag atgtacttat taatgacaaa gcatttatta gatcaaggct ttgttaaagg 1261 gagtgcttta gtgacaaccc atcactgggg atacagttct cgttggtggt atatttccac

1321 gttattaatg tctgatgcac taaaagaagc gaacctacaa actcaagttt atgattcatt

1381 actgtggtat tcacgtgagt ttaaaagtag ttttgatatg aaagtaagtg ctgatagctc

1441 tgatctagat tatttcaata ccttatctcg ccaacattta gccttattat tactagagcc

1501 tgatgatcaa aagcgtatca acttagttaa tactttcagc cattatatca ctggcgcatt 1561 aacgcaagtg ccaccgggtg gtaaagatgg tttacgccct gatggtacag catggcgaca

1621 tgaaggcaac tatccgggct actctttccc agcctttaaa aatgcctctc agcttattta

1681 tttattacgc gatacaccat tttcagtggg tgaaagtggt tggaataacc tgaaaaaagc

1741 gatggtttca gcgtggatct acagtaatcc agaagttgga ttaccgcttg caggaagaca 1801 cccttttaac tcaccttcgt taaaatcagt cgctcaaggc tattactggc ttgccatgtc 1861 tgcaaaatca tcgcctgata aaacacttgc atctatttat cttgcgatta gtgataaaac 1921 acaaaatgaa tcaactgcta tttttggaga aactattaca ccagcgtctt tacctcaagg 1981 tttctatgcc tttaatggcg gtgcttttgg tattcatcgt tggcaagata aaatggtgac 2041 actgaaagct tataacacca atgtttggtc atctgaaatt tataacaaag ataaccgtta 2101 tggccgttac caaagtcatg gtgtcgctca aatagtgagt aatggctcgc agctttcaca 2161 gggctatcag caagaaggtt gggattggaa tagaatgcaa ggggcaacca ctattcacct 2221 tcctcttaaa gacttagaca gtcctaaacc tcatacctta atgcaacgtg gagagcgtgg 2281 atttagcgga acatcatccc ttgaaggtca atatggcatg atggcattcg atcttattta 2341 tcccgccaat cttgagcgtt ttgatcctaa tttcactgcg aaaaagagtg tattagccgc 2401 tgataatcac ttaattttta ttggtagcaa tataaatagt agtgataaaa ataaaaatgt 2461 tgaaacgacc ttattccaac atgccattac tccaacatta aatacccttt ggattaatgg 2521 acaaaagata gaaaacatgc cttatcaaac aacacttcaa caaggtgatt ggttaattga 2581 tagcaatggc aatggttact taattactca agcagaaaaa gtaaatgtaa gtcgccaaca 2641 tcaggtttca gcggaaaata aaaatcgcca accgacagaa ggaaacttta gctcggcatg 2701 gatcgatcac agcactcgcc ccaaagatgc cagttatgag tatatggtct ttttagatgc 2761 gacacctgaa aaaatgggag agatggcaca aaaattccgt gaaaataatg ggttatatca 2821 ggttcttcgt aaggataaag acgttcatat tattctcgat aaactcagca atgtaacggg 2881 atatgccttt tatcagccag catcaattga agacaaatgg atcaaaaagg ttaataaacc 2941 tgcaattgtg atgactcatc gacaaaaaga cactcttatt gtcagtgcag ttacacctga 3001 tttaaatatg actcgccaaa aagcagcaac tcctgtcacc atcaatgtca cgattaatgg 3061 caaatggcaa tctgctgata aaaatagtga agtgaaatat caggtttctg gtgataacac 3121 tgaactgacg tttacgagtt actttggtat tccacaagaa atcaaactct cgccactccc 3181 ttgatttaat caaaagaacg ctcttgcgtt ccttttttat ttgcaggaaa tctgattatg 3241 ctaataaaaa accctttagc ccacgcggtt acattaagcc tctgtttatc attacccgca 3301 caagcattac ccactctgtc tcatgaagct ttcggcgata tttatctttt tgaaggtgaa 3361 ttacccaata cccttaccac ttcaaataat aatcaattat cgctaagcaa acagcatgct 3421 aaagatggtg aacaatcact caaatggcaa tatcaaccac aagcaacatt aacactaaat 3481 aatattgtta attaccaaga tgataaaaat acagccacac cactcacttt tatgatgtgg 3541 atttataatg aaaaacctca atcttcccca ttaacgttag catttaaaca aaataataaa 3601 attgcactaa gttttaatgc tgaacttaat tttacggggt ggcgaggtat tgctgttcct 3661 tttcgtgata tgcaaggctc tgcgacaggt caacttgatc aattagtgat caccgctcca 3721 aaccaagccg gaacactctt ttttgatcaa atcatcatga gtgtaccgtt agacaatcgt 3781 tgggcagtac ctgactatca aacaccttac gtaaataacg cagtaaacac gatggttagt 3841 aaaaactgga gtgcattatt gatgtacgat cagatgtttc aagcccatta ccctacttta 3901 aacttcgata ctgaatttcg cgatgaccaa acagaaatgg cttcgattta tcagcgcttt 3961 gaatattatc aaggaattcc

SEQ ID NO: 2 (129953.1 GI: 1820744)

gcg gcc gcc atg gag gca cga gtg Kozak

M E A R V Notl, Ncol gcc tgg gga gcg ctg gcc gga cct ctg egg gtt etc tgc gtc ctg A W G A L A G P L R V L C V L tgc tgc ctg ttg ggc cgc gcc ate gcc gcc act agt aat cct gca NT 297 C C L L G R A I A A T S N P A ttt gat cct aaa aat ctg atg cag tea gaa att tac cat ttt gca F D P K N L M Q S E I Y H F A caa aat aac cca tta gca gac ttc tea tea gat aaa aac tea ata Q N N P L A D F S S D K N S I eta acg tta tct gat aaa cgt age att atg gga aac caa tct ctt L T L S D K R S I M G N Q S L tta tgg aaa tgg aaa ggt ggt agt age ttt act tta cat aaa aaa L W K W K G G S S F T L H K K ctg att gtc ccc ace gat aaa gaa gca tct aaa gca tgg gga cgc L I V P T D K E A S K A W G R tea tct ace ccc gtt ttc tea ttt tgg ctt tac aat gaa aaa ccg S S T P V F S F W L Y N E K P att gat ggt tat ctt act ate gat ttc gga gaa aaa etc att tea I D G Y L T I D F G E K L I S ace agt gag get cag gca ggc ttt aaa gta aaa tta gat ttc act T S E A Q A G F K V K L D F T ggc tgg cgt get gtg gga gtc tct tta aat aac gat ctt gaa aat G W R A V G V S L N N D L E N cga gag atg ace tta aat gca ace aat ace tec tct gat ggt act R E M T L N A T N T S S D G T caa gac age att ggg cgt tct tta ggt get aaa gtc gat agt att Q D S I G R S L G A K V D S I cgt ttt aaa gcg cct tct aat gtg agt cag ggt gaa ate tat ate R F K A P S N V S Q G E I Y I gac cgt att atg ttt tct gtc gat gat get cgc tac caa tgg tct D R I M F S V D D A R Y Q W S gat tat caa gta aaa act cgc tta tea gaa cct gaa att caa ttt D Y Q V K T R L S E P E I Q F cac aac gta aag cca caa eta cct gta aca cct gaa aat tta gcg H N V K P Q L P V T P E N L A gcc att gat ctt att cgc caa cgt eta att aat gaa ttt gtc gga 256 A I D L I R Q R L I N E F V G 1035 ggt gaa aaa gag aca aac etc gca tta gaa gag aat ate age aaa GIb

271 G E K E T N E Α E E El K I 5 K 1080 tta aaa agt gat ttc gat get ctt aat att cac act tta gca aat

286 L K S D F D A L N I H T L A N

1125 ggt gga acg caa ggc aga cat ctg ate act gat aaa caa ate att

301 G G T Q G R H L I T D K Q I I

1170 att tat caa cca gag aat ctt aac tec caa gat aaa caa eta ttt

316 I Y Q P E N L N S Q D K Q L F

1215 gat aat tat gtt att tta ggt aat tac acg aca tta atg ttt aat G2E

331 D N Y V I L G N Y T T L M F N

1260 att age cgt get tat gtg ctg gaa aaa gat ccc aca caa aag gcg

346 I S R Α ! SJ E E K D P T Q K A

1305 caa eta aag cag atg tac tta tta atg aca aag cat tta tta gat

361 Q L K Q M Y L L M T K H L L D

1350 caa ggc ttt gtt aaa ggg agt get tta gtg aca ace cat cac tgg

376 Q G F V K G S A L V T T H H W

1395 gga tac agt tct cgt tgg tgg tat att tec acg tta tta atg tct

391 G Y S S R W W Y I S T L L M S

1440 gat gca eta aaa gaa gcg aac eta caa act caa gtt tat gat tea

406 D A L K E A N L Q T Q V Y D S

1485 tta ctg tgg tat tea cgt gag ttt aaa agt agt ttt gat atg aaa

421 L L W Y S R E F K S S F D M K

1530 gta agt get gat age tct gat eta gat tat ttc aat ace tta tct P2

436 V S A D S S D L D Y F N T L S

1575 cgc caa cat tta gcc ttg ctg tta ctg gag cct gat gat caa aag

451 R Q H L "K E E E E E P D D Q IC

1620 cgt ate aac tta gtt aat act ttc age cat tat ate act ggc gca pi

466 R I N L V N T F S H Y I T G A

1665 ttg acg caa gtg cca ccg ggt ggt aaa gat ggt tta cgc cct gat

48i L T Q V P P G G K D G L R P D Prame^{^}Shlfi_,

1

1

1800 aca cca ttt tea gtg ggt gaa agt ggt tgg aat aac ctg aaa aaa Frame 'Shift:

526 T P F S V G E S G W N N L K K je5g_: remove

1845 gcg atg gtt tea gcg tgg ate tac agt aat cca gaa gtt gga tta t @ NT 1812

541 A M V S A W I Y S N P E V G L 1890 ccg ctt gca gga aga cac cct ttt aac tea cct teg tta aaa tea

556 P L A G R H P F N S P S L K S

1935 gtc get caa ggc tat tac tgg ctt gee atg tct gca aaa tea teg

571 V A Q G Y Y W L A M S A K S S

1980 cct gat aaa aca ctt gca tct att tat ctt gcg att agt gat aaa

586 P D K T L A S I Y L A I S D K

2025 aca caa aat gaa tea act get att ttt gga gaa act att aca cca

601 T Q N E S T A I F G E T I T P

2070 gcg tct tta cct caa ggt ttc tat gee ttt aat ggc ggt get ttt

616 A S L P Q G F Y A F N G G A F

2115 ggt att cat cgt tgg caa gat aaa atg gtg aca ctg aaa get tat

631 G I H R W Q D K M V T L K A Y

2160 aac ace aat gtt tgg tea tct gaa att tat aac aaa gat aac cgt

646 N T N V W S S E I Y N K D N R

2205 tat ggc cgt tac caa agt cat ggt gtc get caa ata gtg agt aat G5b

661 Y G R Y Q S H G V A Q I V S N 2250 ggc teg cag ctt tea cag ggc tat cag caa gaa ggt tgg gat tgg

676 G S Q L S Q G Y Q Q E G W D W

2295 aat aga atg caa ggg gca ace act att cac ctt cct ctt aaa gac

690 N R M Q G A T T I H L P L K D

2340 tta gac agt cct aaa cct cat ace tta atg caa cgt gga gag cgt

706 L D S P K P H T L M Q R G E R

2385 gga ttt age gga aca tea tec ctt gaa ggt caa tat ggc atg atg

721 G F S G T S S L E G Q Y G M M

2430 gca ttc gat ctt att tat ccc gcc aat ctg gag cgt ttt gat cct G4 & G4b

736 A F D L I Y F A^" N E F, R^" F D I?

2475 aac ttc act gcg aaa aag agt gta tta gcc get gat aat cac tta

751 N F T A K K S V L A A D N H L

2520 att ttt att ggt age aat ata aat agt agt gat aaa aat aaa aat G9

766 I F I G S N I N S S D K N K N

2565 gtt gaa acg ace tta ttc caa cat gcc att act cca aca tta aat

781 V E T T L F Q H A I T P T L N

2610 ace ctt tgg att aat gga caa aag ata gaa aac atg cct tat caa

796 T L W I N G Q K I E N M P Y Q

2655 ^aca aca ctt caa caa ggt gat tgg tta att gat age aat ggc aat

811 T T L Q Q G D W L I D S N G N

2700 ggt tac tta att act caa gca gaa aaa gta aat gta agt cgc caa G6

826 G Y L I T Q A E K V N V S R Q

2745 cat cag gtt tea gcg gaa aat aaa aat cgc caa ccg aca gaa gga G7 841 H Q V S A E N K N R Q P T E G

2₇90 aac ttt age teg gca tgg ate gat cac age act cgc cec aaa gat

856 N F S S A W I D H S T R P K D

2835 gec agt tat gag tat atg gtc ttt tta gat gcg aca cct gaa aaa

8₇1 A S Y E Y M V F L D A T P E K

2880 ^atg gga gag atg gca caa aaa ttc cgt gaa aat aat ggg tta tat

886 M G E M A Q K F R E N N G L Y

2925 cag gtt ctt cgt aag gat aaa gac gtt cat att att etc gat aaa

901 Q V L R K D K D V H I I L D K

29₇o etc age aat gta acg gga tat gcc ttt tat cag cca gca tea att

916 L S N V T G Y A F Y Q P A S I

3015 9^aa 9ac aaa tgg ate aaa aag gtt aat aaa cct gca att gtg atg

931 E D K W I K K V N K P A I V M

3060 act cat cga caa aaa gac act ctt att gtc agt gca gtt aca cct G8

946 T H R Q K D T L I V S A V T P

3105 9^at tta aat atg act cgc caa aaa gca gca act cct gtc ace ate

961 D L N M T R Q K A A T P V T I

3150 ^aat gtc acg att aat ggc aaa tgg caa tct get gat aaa aat agt g₇6 N V T I N G K W Q S A D K N S

3195 9^aa gtg aaa tat cag gtt tct ggt gat aac act gaa ctg acg ttt

991 E V K Y Q V S G D N T E L T F

3240 ^ac9 ^a9ft tac ttt ggt att cca caa gaa ate aaa etc teg cca etc

1006 T S Y F G I P Q E I K L S P L

3285 cct tga SEQ ID NO; 4

1021 P * SEQ ID NO; 3

SEQ ID NOS: 3 and 4

GIb Start NT 1052, End NT 1099

N₂₈₂ → Q

5' 3' c etc gca tta gaa gag CaG ate age aaa tta aaa agt gat ttc gat gc

L A L E E Q I S K L K S D F D SEQ ID NO: 5 G2E

Start NT 1224, End NT 1280

^N33β ^→ Q/ ^N345 ^→ Q' ^L342 improved codon.

5' 3' gtt att tta ggt CaG tac acg aca Cta atg ttt CaG att age cgt get tat gtg ctg

V I L G Q Y T T L M F Q I S R A Y V L

SEQ ID NO: 6

G3

Start NT 1756, End NT 1804

(N₅₁₅) S₅₁₇ — » A, L₅₁₉ and L₅₂₂ codons improved.

5' PvuII 3' ca gcc Ttt aaa aat gcc Get cag ctG att tat CtG tta cgc gat aca cc A F K N A A Q L I Y L L R D T

SEQ ID NO: 7

G4

Starts NT 2446 Ends NT 2486

P₇₄₅ → L, N₇₅₁ → Q

5' Xhol 3' at ccc gcc aat cTC gag cgt ttt gat cct CaG ttc act gcg P A N L E R F D P Q F T A

SEQ ID NO: 8 G5b

Start NT 2237 End NT 2277

N,

SEQ ID NO: 9

G6

Start NT 2718 End NT 2757 N₈₃₆ ^{→ D} (Aspartic acid)

SexAi /

SEQ ID NO: 10

G7b

Start NT 2761 End NT 2800

N₈₅₆ → ^Q

5' 3 g aca gaa gga CaG ttt age teg gca tgg ate gat cac agg

SEQ ID NO: 11 G8b

Start NT 3081 End NT 3123

N_q

Sfcl

3 ' gtc agt gca gtt aca cct gat Cta CaG atg act cgc c

SEQ ID NO: 12

G9

Start NT 2529 End NT 2573

SEQ ID NO: 13

Claims

Claims :

1. A chondroitinase polypeptide comprising a modified bacterial glycosaminoglycan (GAG) polysaccharide lyase sequence, said sequence having a reduced number of active N-glycosylation motifs relative to the unmodified sequence, and; said chondroitinase polypeptide being secretable in an active form by mammalian cells.

2. A chondroitinase polypeptide according to claim 1 comprising an amino acid sequence having at least 20% sequence identity to the sequence of SEQ ID NO: 1 or a fragment thereof having chondroitinase activity, the amino acid sequence having a reduced number of active N- glycosylation motifs relative to the sequence of SEQ ID NO: 1 or the fragment thereof , and the chondroitinase polypeptide being secretable in an active form by mammalian cells.

3. A chondroitinase polypeptide according to claim 2 wherein said fragment comprises residues 282 to 963 of SEQ ID NO: 1.

4. A chondroitinase polypeptide according to claim 2 or claim 3 wherein said amino acid sequence comprises a mutation in one or more N-glycosylation motifs which are present in the sequence of SEQ ID NO: 1 or the fragment thereof, said mutation preventing N-glycosylation of the motif.

5. A chondroitinase polypeptide according to claim 4 wherein the amino acid sequence comprises a conservative substitution of the N or S/T residue of an N-glycosylation motif.

6. A chondroitinase polypeptide according to claim 5 wherein the amino acid sequence comprises an N-Q or N-K or N-D substitution of the N residue of a N-glycosylation motif.

7. A chondroitinase polypeptide according to claim 5 wherein the mutation is an S/T to A substitution of the S/T residue of the N- glycosylation motif.

8. A chondroitinase polypeptide according to any one of claims 4 to 7 wherein said amino acid sequence comprises a mutation in one or more N-glycosylation motifs selected from the group consisting of the N- glycosylation motifs positioned at N282, N345, N336, N515, N675, N773, N856 and N963 of SEQ ID N0:l.

9. A chondroitinase polypeptide according to any one of claims 4 to 8 wherein said amino acid sequence comprises a mutation in two or more N-glycosylation motifs.

10. A chondroitinase polypeptide according to claim 9 wherein said amino acid sequence comprises mutations in the N-glycosylation motifs positioned at N282, N336, N345 and N515 of SEQ ID NO: 1.

11. A chondroitinase polypeptide according to claim 10 wherein said amino acid sequence further comprises a mutation in the N- glycosylation motif positioned at N675 of SEQ ID NO: 1.

12. A chondroitinase polypeptide according to any one of claims 2 to 11 wherein the amino acid sequence comprises the sequence of SEQ ID NO: 1 or a fragment thereof with mutations in one or more N- glycosylation motifs thereof, said mutations preventing N-glycosylation of the one or more motifs .

13. A chondroitinase polypeptide according to claim 12 comprising mutations in one or more N-glycosylation motifs selected from the group consisting- of the N-glycosylation motifs positioned at N282, N345, N336, N515, N675, N773, N856 and N963 of SEQ ID NO:1.

14. A polypeptide comprising a chondroitinase polypeptide according to any one of the preceding claims coupled to a signal sequence.

15. A polypeptide according to claim 14 wherein said signal sequence is a mammalian signal sequence.

16. A nucleic acid encoding a chondroitinase polypeptide according to any one of claims 1 to 13 or a polypeptide according to claim 14 or claim 15.

17. A nucleic acid according to claim 16 having a sequence which is optimised for eukaryotic codon usage.

18. A nucleic acid according to claim 16 or claim 17 comprising a kozak sequence optimised for eukaryotic expression.

19. A nucleic acid according to any one of claims 16 to 18 operably linked to a heterologous regulatory element.

20. A nucleic acid according to any one of claims 16 to 19 wherein said heterologous regulatory element is suitable for expression in mammalian neural cells.

21. A nucleic acid according to claim 20 wherein said heterologous regulatory element is suitable for expression in mammalian glial cells .

22. A vector comprising a nucleic acid according to any one of claims 16 to 21.

23. A vector according to claim 22 which is a viral vector.

24. A vector according to claim 23 which is a adenovirus, adeno- associated virus (AAV) , for example AAV serotype 2 virus, retrovirus, lentivirus, recombinant adenovirus, 'gutless' adenovirus, herpes simplex virus, or poliovirus vector.

25. A host cell comprising a nucleic acid according to any one of claims 16 to 21 or a vector according to any one of claims 22 to 24.

26. A host cell according to claim 25 which is a mammalian neural cell.

27. A host cell according to claim 26 which is a human neural cell.

28. A host cell according to claim 27 which is a human glial cell.

29. A nucleic acid according to any one of claims 16 to 21, a vector according to any one of claims 22 to 24 or a host cell according to any one of claims 25 to 28 for use in a method of treatment of the human or animal body.

30. A nucleic acid according to any one of claims 16 to 21, a vector according to any one of claims 22 to 24 or a host cell according to any one of claims 25 to 28 for use in a method of treatment of neural damage .

31. Use of a nucleic acid according to any one of claims 16 to 21, a vector according to any one of claims 22 to 24 or a host cell according to any one of claims 25 to 28 in the manufacture of a medicament for use in the treatment of neural damage .

32. A method of treating neural damage comprising administering a nucleic acid according to any one of claims 16 to 21, a vector according to any one of claims 22 to 24 or a host cell according to any one of claims 25 to 28 to an individual in need thereof .