WO1996023003A1 - A therapeutic molecule - Google Patents

Abstract

The present invention is directed generally to a therapeutic molecule. More particularly, the present invention provides a therapeutic molecule comprising a heparan sulfate polymer or a derivative thereof obtainable from heparan sulfate proteoglycan and which is capable of interacting with a specific cytokine. The therapeutic molecule of the present invention will be useful in promoting cytokine function in vitro and in vivo. The present invention is particularly directed to the effects of the heparan sulfate polymers on Fibroblast Growth Factors.

Description

A THERAPEUTIC MOLECULE

The present invention is directed generally to a therapeutic molecule. More particularly, the present invention provides a therapeutic molecule comprising a heparan sulfate polymer or its derivative obtainable from heparan sulfate proteoglycan and which is capable of interacting with a specific cytokine. The therapeutic molecule of the present invention will be useful in promoting cytokine function in vitro and in vivo. The present invention is particularly directed to the effects of the heparan sulfate polymers on Fibroblast Growth Factors.

Bibliographic details of the publications numerically referred to in this specification are collected at the end of the description. Sequence Identity Numbers (SEQ ID NOs.) for the amino acid sequences referred to in the specification are defmed following the bibliography.

Throughout this specification, unless the context requires otherwise, the word "comprise", or variations such as "comprises" or "comprising", will be understood to imply the inclusion of a stated element or integer or group of elements or integers but not the exclusion of any other element or integer or group of elements or integers.

Many cytokines act as growth factors, promotants and/or stimulants for precursor cells. For example, the mammalian central nervous system is developmentally derived from the cells of the neural tube. Fibroblast Growth Factor- 1 (FGF-1; also known as acidic (a) FGF) and Fibroblast Growth Factor-2 (FGF-2; also known as basic (b) FGF) are heparin-binding growth factors which are potent developmental regulators of the proliferation, migration and differentiation of neural precursor cells (la;b). FGF activity is regulated in part by heparin analogues such as heparan sulfate proteoglycan [HSPG](2a;b). The HSPGs are a highly diverse group of macromoiecules, each of which consists of a core protein to which highly sulfated glycosaminoglycan (GAG) side chains of heparan sulfate are covalently attached (3a;b). They are ubiquitous constituents of mammalian cell surfaces and of most extracellular matrices including the specialised basal laminae that surround neural tissue (2a;b;4).

In order to develop therapeutic molecules capable of influencing precursor cell development, and in particular cells derived from the neural tube, it is important to understand the interaction between cytokines and other effector molecules and their ligands and to determine whether other molecules are required for, or to promote, such interaction. It is these "other molecules" which may form the basis of potentially useful therapeutics.

For example, an HSPG-FGF interaction is requisite for the presentation and subsequent binding of FGF to signal transducing receptors (5a;b;c;d). However, the exact nature of the interaction is unclear. There is a need, therefore, to elucidate the mechanism of HSPG interaction with FGF-1 and FGF-2.

In accordance with the present invention, it has been surprisingly discovered that the GAG side chains or various derivatives thereof on HSPGs exhibit specificity for particular cytokines such as FGF-1 or FGF-2. A common core protein is synthesised with a particular species of polymeric GAG side chains which exhibit specificity, for example, to FGF-1 or FGF-2. This discovery will result in a new generation of therapeutic molecules capable of modulating cytokine-ligand interaction and more particularly FGF-ligand interaction. More particularly, the present invention provides a range of derivatives and more specifically functional fragments of the GAG side chains which are especially useful in generating a range of therapeutic molecules.

Accordingly, one aspect of the present invention is directed to an isolated GAG polymer or derivative thereof from an HSPG wherein said GAG polymer or its derivative is capable of interaction with a cytokine. More particularly, the present invention provides an isolated GAG polymer or a derivative thereof from an HSPG wherein said GAG polymer or its derivative is capable of interaction separately with either FGF-1 or FGF-2 but not both. Even more particularly, the present invention contemplates an isolated GAG polymer or a derivative thereof from an HSPG wherein said GAG polymer or its derivative is capable of interaction with FGF-2 and wherein said HSPG is obtainable from murine cells at approximately embryonic day 8-10. In a related embodiment, the present invention relates to an isolated GAG polymer or a derivative thereof from an HSPG wherein said GAG polymer or its derivative is capable of interaction with FGF-1 and wherein said HSPG is obtainable from murine cells at approximately embryonic day 11 or older such as day 11-13.

In a most preferred embodiment, the murine cells are embryonic day 10 neuroepithelial cells transformed with an oncogene on a retroviral vector. An example of a suitable cell line is the 2.3D cell line which is FGF sensitive (la;b). The 2.3D cell line was deposited at the PHLS Centre for Applied Microbiology and Research, European Collection of Animal Cell Cultures (ECACC), Division of Biologies, Porton Down, Salisbury, Wiltshire SP4 0JG on 16 May, 1995 under Provisional Accession Number 95061001; the confirmed Accession Number is 95051601.

Accordingly, a particularly preferred embodiment of the subject invention is directed to an isolated GAG polymer or a derivative thereof from an HSPG wherein said GAG polymer or its derivative is capable of interaction with FGF-2 but not FGF-1 and wherein said HSPG is obtainable from the 2.3D cell line dividing freely in tissue culture. This GAG polymer is referred to herein as "GAGB". A fragment or derivative includes molecules capable of promoting FGF binding to its receptor or inhibiting binding to its receptor. The derivatives may act, therefore, as agonists or antagonists of FGF binding to its receptor or of FGF-HSPG interaction.

In a related embodiment, the present invention is directed to an isolated GAG polymer or a derivative thereof from an HSPG wherein said GAG polymer or its derivative is capable of interaction with FGF-1 but not FGF-2 and wherein said HSPG is obtainable from the 2.3D cell line grown in culture under contact inhibiting conditions. This GAG polymer is referred to herein as "GAGA".

The term "polymer" or like derivations includes molecules comprising at least four sugars or derivatives thereof such as from four to 400. Accordingly, another aspect of the present invention provides an isolated molecule comprising: (i) a repeating disaccharide structure (X-Y)_n wherein: X is hexuronic acid; Y is glucosamine; and n is 2 to 200; (ii) an ability to bind either FGF-1 or FGF-2 but not both; and (iii) being isolatable from an HSPG which in one form comprises GAG polymers capable of binding FGF-1 and in another form comprises GAG polymers capable of binding FGF-2.

Preferably, X-Y are α,β-linked glucosamine and hexuronic acid in linkage sequence [(l→ 4)α-D-glucosaminyl-(l→ 4)β-D-hexuronosyl]_n. The glucosamine may be N- acetylated or N-sulfated and the hexuronate may be glucuronate or iduronate. Preferably in relation to FGF-2 binding, n is 2 to 20. More preferably, n is 4 to 15. Even more preferably n is 8 to 12 such as 9. The specificity of a particular GAG polymer for a particular cytokine is created by the specific pattern of carboxyl and sulfate groups attached to the glucosamine and hexuronic acid.

In a particularly preferred embodiment of the present invention, there is provided a GAG polymer having the identifying characteristics of GAGB and comprising the following disaccharides in percentage amounts given in parentheses: iduronic acid N-acetylated glucosamine [UA-GlcNAc](55.4%), iduronic acid N-sulfated glucosamine [UA-GlcNSO₃] (22.2%), iduronic acid N-acetylated glucosamine 6-sulfate [UA-GlcNAc(όS)] (3.2%), iduronic acid 2-sulfate N-acetylated glucosamine [UA-(2S GlcNAc] (1.8%), iduronic acid N-sulfated glucosamine 6-sulfate [UA-GlcNSO₃(6S)] (2.5%), iduromc acid 2-sulfate N-sulfated glucosamine [UA-(2S)-GlcNSO₃] (9.0%) and iduronic acid 2-sulfate N-sulfated glucosamine 6-sulfate [UA-(2S)-GlcNSO₃(6S)] (5.1 %). Other disaccharides up to about 0.7-0.8% may also be found in GAGB.

In an alternative embodiment, there is provided a GAG polymer having the identifying characteristics of GAGA and comprising the following disaccharides in percentage amounts given in parentheses: iduronic acid N-acetylated glucosamine [UA-GlcNAc] (50.7%), iduromc acid N-sulfated glucosamine [UA-GlcNSO₃] (19.1%), iduronic acid N-acetylated glucosamine 6-sulfate [UA-GlcNAc (6S)] (4.7%), iduronic acid 2-sulfate N-acetylated glucosamine [UA-(2S)- GlcNAc] (2.6%), iduronic acid N-sulfated glucosamine 6-sulfate [UA-GlcNSO₃(6S)] (2.8%), iduronic acid 2-sulfate N-sulfated glucosamine [UA-(2S)-GlcNSO₃] (9.1%) and iduronic acid 2-sulfate N-sulfated glucosamine 6-sulfate [UA-(2S)-GlcNSO₃(6S)] (5.8%). Other disaccharides up to about 5.1-5.2% may also be found in GAGA.

The identifying characteristics of GAGB or GAGA include the preferential interaction with FGF2 and FGF1, respectively.

By "isolated" is meant a preparation of a GAG polymer or a derivative thereof which has undergone at least one purification or separation step away from a core protein. Preferably, the term "isolated" extends to a biologically pure preparation of the polymer comprising at least 35%, preferably at least 45%, more preferably at least 55%, still more preferably at least 65%, even more preferably at least 75-80% and even more preferably greater than 95% of the GAG polymer as determined by weight, activity (e.g. cytokine binding activity), immunoreactivity (e.g. antibody interactivity), sugar content or other convenient means.

Conveniently, the GAG polymer or derivative thereof is purified from HSPG derived from conditioned medium produced by either the neuroepithelial cell line 2.3D which expresses the c-myc oncogene in cloned embryonic primary neuroepithelial cells or from primary neuroepithelial brain. As is described in Example 2, both neuroepithelial tissue at embryonic day 9 (referred to herein as 9") or 2.3 D cells grown in non-confluent culture produces HSPG capable of binding FGF-2. Embryonic day 11 tissue (referred to herein as "El l") or 2.3 D cells grown in continuously confluent culture [13] (i.e. under contact inhibiting conditions) produces HSPG capable of binding FGF-1. The present invention is predicated in part on the discovery that the heparan sulfate side chains on 2.3 D non-confluent cells or E9 HSPGs bind FGF-2 while the heparan sulfate side chains on 2.3 D continuously confluent cells or El 1 HSPGs bind FGF-1. The 2.3 D cell line provides, therefore, a particularly useful source of HSPG side chains and which can be readily upgraded to large scale commercial production.

Accordingly, another aspect of the present invention contemplates a method of purifying a GAG polymer or a derivative thereof capable of binding either FGF-1 or FGF-2, said method comprising generating a neuroepithelial cell line expressing an oncogene and growing and/or maintaining the cell line for a time and under conditions sufficient for said cell line to secrete HSPG molecules into the conditioned medium; collecting the HSPG at predetermined time intervals and subjecting same to HSPG isolating means; subjecting isolated HSPG to GAG polymer purification means.

A suitable protocol for purifying GAG polymers away from HSPG includes but is not limited to subjecting the HSPG to one or more proteolytic enzymes to destroy or substantially remove the protein core. Pronase is a particularly useful enzyme in this respect. Alternatively, the protein core may be removed by sonic disruption, shearing or via various forms of hydrolysis. HPLC or other suitable means may then be used to purify the GAG polymers.

Preferably the neuroepithelial cell line is cell line 2.3D which is made by expressing the c-myc oncogene in cloned embryonic day 10 primary neuroepithelial cells.

Preferably, the neuroepithelial cell line is grown to approximately 50-90% and more preferably about 70% confluency and then the conditioned medium is collected at predetermined intervals. These intervals are those sufficient for HSPGs to be synthesized with a specificity for FGF-2 and then, following a change in GAG polymer composition and or structure, HSPGs are synthesized with specificity for FGF-1 at a later time. In one embodiment, FGF-2 specific HSPG is produced by 2.3D cells grown under non-confluent conditions whereas FGF-1 specific HSPG is produced by 2.3D cells grown under contact inhibiting conditions. In another embodiment, E9 or El 1 primary neuroepithelial brain cells are used, respectively.

The purification of the HSPG can be by any convenient means such as DEAE-Sepharose chromatography, affinity chromatography or immunosorbant chromatography amongst others.

Another aspect of the present invention contemplates an isolated core protein of a heparan sulfate proteoglycan (HSPG) wherein said core protein is capable of being substituted with GAG side chains such that one species of side chains is capable of preferentially binding to FGF-1 and another species is capable of preferentially binding to FGF-2.

By "isolated" is meant a preparation of polypeptide or protein which has undergone at least one purification or separation step away from the naturally occurring environment of the polypeptide or protein. Preferably, "isolated" extends to a biologically pure preparation comprising at least 35%, preferably at least 45%, more preferably at least 55%, still more preferably at least 65%, even more preferably at least 75-80% and even more preferably greater than 95% of the polypeptide or protein as determined by weight, activity, immunoreactivity (e.g. antibody reactivity), cytokine binding activity or other convenient means.

Alternatively, or in addition to, the isolated polypeptide may be recombinant or synthetic or may be a non-full length molecule relative to the naturally occurring protein.

The polypeptide or protein of the present invention has, in a preferred embodiment, a molecular weight determined on SDS-PAGE of between 30 to 55 kDa. More specifically, the molecular weight is between 35 and 50 kDa and even more specifically is approximately 45±5 kDa.

Preferably, the polypeptide or protein comprises a region having the amino acid sequence: G A S C E D C Q T F Y Y G D A Q R G T P Q D [SEQ ID NO:l] and/or a region having the amino acid sequence: G T P Q D C Q P C P C Y G A P R R T T P A [SEQ ID NO:2], or an amino acid sequence having at least 60%, more preferably at least 70%, even more preferably at least 80% and still more preferably at least 90% similarity to either or both of the above sequences or to a portion or region thereof. The core protein bears some homology to the basement membrane protein proteoglycan, perlecan, although is of considerably smaller size (400 kDa versus 45 kDa) and has considerable higher glycosylation density. It also carries unique peptide domains and is encoded in an mRNA of approximately 3.5 kb.

The polypeptide or protein of the present invention is useful, for example, as a core substrate for GAG polymer synthesis to produce a specific cytokine binding molecule, such as an FGF-1 or FGF-2 binding molecule. Additionally, the polypeptide or protein may be used to generate antibodies against itself or related molecules or to generate agonists or antagonists to a naturally occurring form of the molecule. Most preferably, however, the polypeptide or protein will be in glycosylated form.

Accordingly, another aspect of the present invention relates to an isolated proteoglycan having one of at least two species of GAG polymer side chains such that one species binds preferentially to FGF-1 and another species binds preferentially to FGF-2.

The term "isolated" and the protein moiety of the proteoglycan are as hereinbefore defined. Conveniently, the proteoglycan is isolatable from conditioned medium of a neuroepithelial cell line such as cell line 2.3D as hereinbefore described. Additionally, the neuroepithelial cell line may be transgenic for other genetic sequences and in particular those which modify or assist in the expression of the proteoglycan of the present invention.

The present invention as described herein is predicated in part on the discovery that different cytokines bind to the same proteoglycan depending on the composition and nature of the GAG side chains bound to the proteoglycan. In particular, one form of an HSPG binds preferentially FGF-1 and another form of the same molecule binds preferentially to FGF-2. The term "binds", however, is not to be construed as imparting any limitation and extends to association, aggregation or any other form of interaction between molecules including tripartite interaction between an FGF, its receptor and the GAG side chain.

The proteoglycan or heparan sulfate polymer of the present invention will be useful in promoting, stimulating and/or enhancing activation of cytokines. In this regard, it is particularly exemplified by HSPG-FGF interaction required for presentation of FGF to the appropriate signal transducing receptors on neural precursor cells or any other cell type bearing the appropriate FGF receptor. It is apparent, however, that the present invention extends to the promotion, stimulation and/or enhancement of interaction between other cytokines and their target sequences. The GAG chains bearing specificity for FGF-2 also bear a carbohydrate subdomain which specifically bind a region of the FGFR1 receptor. In a particular embodiment, the GAG chains bind to the FGFRlIIIc receptor. The heparan sulfate interaction with FGF-2 thus serves to activate the cytokine and directly couple it to its appropriate receptor by formation of a ternary complex. The GAG chain with specificity for FGF-1 works in an analogous fashion with its particular FGF receptor.

In many circumstances, it may be preferable to use a non-full length GAG polymer in promoting, stimulating and/or enhancing activation of cytokines such as FGF-1 or FGF-

2. It has now been surprisingly discovered that the GAG polymers may be derivatised into smaller, functional fragments which are particularly efficacious in mediating FGF interaction. Accordingly, the present invention further contemplates mutants, derivatives, fragments, parts, homologues, analogues and chemical equivalents of the GAG polymers. Such forms are referred to collectively herein as "derivatives".

Particularly preferred derivatives are fragments of GAG polymers obtainable by any number of means such as by chemical disruption and in particular acid hydrolysis with nitrous acid or by enzymatic cleavage with heparanase I and/or heparanase III

(heparitinase). The derivatives contemplated herein may also act as antagonists and inhibit or reduce FGF-receptor interaction. Such antagonists may also have important therapeutic utility.

Non-full length GAG derivatives are particularly preferred as they are readily diffusable into tissues, have greater bioavailability, potentially exhibit greater specificity, tend to reduce adverse side effects and reduce the likelihood of adverse host immune reactivity. The latter is particularly important since GAG polymers of non-human origin (e.g. from murine sources) are operative in humans and up to the present time, the preferred source of GAG polymers is from HSPGs isolated from murine sources. Accordingly, the present invention contemplates GAG polymers or derivatives thereof from HSPGs of non-human origin (e.g. murine source) used in humans or non-murine animals (a heterologous system) as well as the use of a GAG polymer from an HSPG from the same species origin as the recipient of therapy (a homologous system).

An example of a derivative of a GAG polymer bearing both an FGF-2 binding domain and an FGF-2 receptor-binding domain is a 9 disaccharide unit fragment of a GAG polymer isolatable from an HSPG obtainable from 2.3D cells, grown under non- confluent conditions. When this 9 disaccharide units is further split into smaller fragments with heparanase I, the fragments are incapable of promoting cell-FGF-2 interaction, but are capable of blocking the formation of an activating FGF-2-heparan sulfate-FGF-2 receptor ternary complex. The present invention also contemplates a similar fragment capable of interactivity with FGF-1. In this context, "interactivity" includes functional interaction to facilitate FGF-receptor binding or alternatively antagonistic interaction to inhibit or reduce FGF-receptor interaction.

Accordingly, the present invention further contemplates a fragment of a GAG polymer, said fragment being at least about 5 disaccharides in length and obtained from a GAG polymer isolated from a HSPG from 2.3D cells grown under non-confluent conditions wherein said fragment is capable of interaction with FGF-2.

In a related embodiment the fragment is isolated from a GAG polymer isolated from an HSPG from 2.3D cells grown under contact-inhibiting conditions and wherein said fragment is capable of interaction with FGF-1.

Preferably, the fragment is at least about 7 disaccharides in length. More preferably, the fragment is at least about 9 disaccharides in length.

Another aspect of the present invention is directed to an antagonist of FGF-receptor interaction, said antagonist comprising a fragment of a GAG polymer, said fragment being at least about 3 disaccharides in length and obtained from a GAG polymer isolated from an HSPG.

Preferably, where the antagonist affects FGF-2-receptor interaction, it is from an HSPG from 2.3D cells grown under non-confluent conditions.

Preferably, where the antagonist affects FGF-1 -receptor interaction, it is from an HSPG from 2.3D cells grown under contact-inhibiting conditions.

The antagonistic fragment may also be at least about 5 or 7 disaccharides in length.

Furthermore, FGF-interactive derivatives may be readily detected by a number of convenient assays. One such assay consists of a mitogenic assay on embryonic neuroepithelial cells, or the 2.3D cell line. Another such assay is where labelled GAG fragments are chromatographed on FGF or FGF receptor (FGFR) peptide fragments coupled to an Affi-Gel 10 affinity support column and monitored for their ability to be retained. Another procedure is a plate assay whereby appropriate amino acid fragments derived from either the FGF or the FGF receptor GAG-binding domains are derivatized to plastic and checked for their ability to bind appropriate [³H]- or [³⁵S]- labelled GAG sequence.

Another aspect of the present invention contemplates a method of promoting, stimulating and/or enhancing interaction between a particular cytokine and a target site on a cell in an animal, said method comprising administering to said animal a GAG polymer or derivative thereof which preferentially binds to said cytokine, for a time and under conditions sufficient for said GAG polymer or its derivatives to promote binding of said cytokine with said target sequence.

Preferably, the cytokine is FGF-1 or FGF-2 and the GAG polymer is GAGA and GAGB, respectively. In one embodiment, the effect of the GAG polymer is to maintain cells in a viable state. In another embodiment, the effect of the GAG polymer is to prevent or delay cell death. In a related embodiment, two GAG polymers or derivatives thereof are administered simultaneously or sequentially to thereby promote interaction of at least two different cytokines with target sequences in the cell.

In an alternative embodiment, the present invention contemplates a method for inhibiting or reducing interaction between a particular cytokine and a target site on a cell in an animal, said method comprising administering to said animal an antagonist of cytokine- receptor interaction for a time and under conditions sufficient to inhibit or reduce said interaction.

In one embodiment, the antagonist is a fragment of GAGB and inhibits or reduces FGF- 2-receptor interaction. An example of such a fragment is a fragment of the 9 disaccharide unit fragment of GAGB. In another embodiment, a fragment of GAGA is used to inhibit FGF-1 -receptor interaction.

In another preferred embodiment, there is provided a method for promoting, stimulating and/or enhancing cell proliferation, migration and/or differentiation of any tissue which bears the appropriate FGF receptors or in an animal said method comprising the administration of a GAG polymer or derivatives thereof wherein said GAG polymer or its derivative interacts with FGF-1 or FGF-2 but not both. This embodiment relates particularly to non-neuronal tissue. In a related aspect, the present invention provides a method for promoting or facilitating maintenance and survival of neuronal cells in an animal, said method comprising the administration of a GAG polymer or derivative thereof wherein said GAG polymer or its derivative interacts with FGF-1 or FGF-2 but not both. "Interacts" in this context is to facilitate FGF binding to its receptor.

In a preferred embodiment, the cells are motor neurons and the effect of the GAG polymer or its derivatives in combination with FGF 1 or 2 is to rescue motor neurons during the period of cell death. However, the present invention extends to all neurons and in particular large neurons.

In yet another preferred embodiment, the GAG polymer or derivatives thereof are used in vitro to maintain or stimulate growth υf suitable cell lines, such as neuroepithelial cells.

The route of in vivo administration may be by any convenient means but is generally by intravenous administration. Other forms of administration are possible, however, modification of the active molecules may be required to, for example, protect same from host enzymes or to facilitate passage through the blood vessel walls.

The effective amount of GAG polymer or derivative thereof will depend on the preparation, condition and host but, may generally be from at least about O.OOlμg/kg body weight to at least about lOmg/kg body weight. A more preferred range is at least about O.Olμg/kg body weight to at least about lmg/kg body weight. Alternatively, a range of at least about lμg/kg body weight to about 500μg/kg body weight. Administration may be a single dose or a series of doses. Additionally, the GAG polymer or derivatives thereof may also be complexed with an FGF.

Another aspect of the present invention is directed to a pharmaceutical composition comprising a GAG polymer or derivative thereof capable of interactivity with FGF-2, said composition further comprising one or more pharmaceutically acceptable carriers and/or diluents. In a related embodiment, the pharmaceutical composition comprises a GAG polymer or derivative thereof capable of binding FGF-1. In a further embodiment, the composition comprises at least two species of GAG polymers or derivatives thereof or one species of GAG polymer and a cytokine or other active molecule.

Methods for preparing pharmaceutical compositions are well known and reference can conveniently be made to Remington's Pharmaceutical Sciences, 17th Edition, Elsivier Publishing Co., Eaton, Pennsylvania, U.S.A.

In yet a further embodiment of the present invention, there is provided a heparan sulfate polymer linked to a particular core molecule capable of targeting the polymer to a specific site or group of sites. Such a hybrid molecule will be particularly useful, for example, for localized FGF treatment. The present invention extends to heterologous and homologous systems in relation to the species from which the HSPG is purified and the intended recipient, for example, murine heparan sulfate polymer is active both in human cells and in chick embryos amongst other animal tissues. Preferably, however, murine or human HSPGs are used as a source of GAG polymers.

Still another aspect of the present invention contemplates a method for rescuing neurons during the period of cell death in a mammal, said method comprising administering to said mammal an effective amount of a GAG polymer or a derivative thereof from an HSPG obtainable from embryonic day 8-10 cells and wherein said GAG polymers or derivative is capable of interaction with FGF-2 but not FGF-1. Preferably, the neurons are large neurons. More particularly, the neurons are motor neurons.

A further aspect of the present invention provides a method for promoting the viability of cells carrying an FGFRlIIIc receptor for FGF-2 in a mammal, said method comprising administering to said mammal an effective amount of a GAG polymer or a derivative thereof from an HSPG obtainable from embryonic day 8-10 cells and wherein said GAG polymers or derivative is capable of interaction with FGF-2 but not FGF-1.

Preferably, the mammal is a human and preferably the cells are 2.3D cells grown under non-confluent conditions.

The present invention extends to the use of the GAG polymers of the present invention in situations where cells producing FGF-1 and or FGF-2 are transplanted into brain parenchyma to relieve the symptoms of neurological disorders such as Huntington' s Disease or Parkinson's Disease or afflictions which involve Parkinsonism. In one embodiment, the cells are embryonic cells and the GAG polymers facilitate FGF interaction with its receptors. In another embodiment the embryomc cells are genetically engineered to express an FGF such as FGF-1 and or FGF-2. For example, E14 embryonic cells may be used. The cells are then transplanted into brain tissue to an area generally occupied by cells of the substantial nigra which are dopaminergic. After, during or prior to grafting, GAGA and/or GAGB is supplied which will greatly increase the benefits of the transplanted cells both anatomically and behaviourally.

Accordingly, another aspect of the prevent invention contemplates a method of ameliorating the effects or symptoms of neurological disorders in a mammal, said method comprising transplanting cells which synthesize FGF-2 to a neurological environment and contacting said transplanted cells with an effective amount of a GAG polymer capable of interaction with FGF-2 but not FGF-1 and obtainable from an HSPG from murine cells at approximately embryomc day 8-10. Preferably the GAG polymer is GAGB.

A further aspect of the present invention contemplates a method of ameliorating the effects or symptoms of neurological disorders in a mammal, said method comprising transplanting cells which synthesize FGF-1 to a neurological environment and contacting said transplanted cells with an effective amount of a GAG polymer capable of interaction with FGF-1 but not FGF-2 and obtainable from an HSPG from murine cells at approximately embryomc day 11-13. Preferably, the GAG polymer is GAGA.

In a particularly preferred embodiment, the cells are first genetically engineered to express increased amounts of an FGF or a derivative thereof.

Yet another aspect of the present invention contemplates a method of ameliorating the effects or symptoms of neurological disorders in a mammal, said method comprising transplanting cells genetically engineered to express FGF-2 to a neurological environment and contacting said transplanted cells with an effective amount of a GAG polymer capable of interaction with FGF-2 but not FGF-1 and obtainable from an HSPG from murine cells at approximately embryonic day 8-10. Preferably, the GAG polymer is GAGB.

In an alternative embodiment, the present invention provides a method of ameliorating the effects or symptoms of neurological disorders in a mammal, said method comprising transplanting cells genetically engineered to express FGF-1 to a neurological environment and contacting said transplanted cells with an effective amount of a GAG polymer capable of interaction with FGF-1 but not FGF-2 and obtainable from an HSPG from murine cells at approximately embryonic day 1 1-13. Preferably, the GAG polymer is GAGA.

In accordance with these and other aspects of the present invention, preferred animals for treatment are mammals such as humans, livestock animals (e.g. sheep, cows, pigs, horses), companion animals (e.g. dogs, cats) or laboratory test animals (e.g. mice, rats, rabbits). Most preferably, the mammal is a human.

The present invention is further described by reference to the following non-limitating Figures and/or Examples.

In the figures:

Figure 1 is a graphical representation of the results of affinity chromatography using HSPG coupled to an Affi-Gel 10 (see Example 1). Values presented are the means and SDs of six determinations from two to four experiments.

Figure 2 is a graphical representation showing gel filtration analysis of GAG chains from E9 and El l HSPGs on BioGel P-2 columns. CS, chondroitin sulfate

Figure 3 shows [³H]thymidine incorporation into neuroepithelial cells maintained in either FGF1 or FGF2 with supplemental glycosaminoglycan. E10 neuroepithelial cells were trypsinized (0.1% w/v trypsin) to remove surface and adherent proteoglycans, allowed 2 hours to recover, and then plated onto HL-a plates in the presence of 5 ng/ml of FGF-1 or FGF-2 at a cell density of 1,500 cells per well (1). After 36 hours the cells were pulsed for 16 hours with [³H]thymidine, harvested, washed, and counted. Cultures remained unsupplemented, or were supplemented with either heparin, postnatal day 1 brain HSPG preparation (PI), commercially obtained HSPG from liver (Sigma), E9 HSPG or El l HSPG, all at 10 μg/ml. In some experiments, the cells were either not pretreated with trypsin (NT-control) or were not supplemented with growth factors (E9 HSPG, El l HSPG, no FGF). In preparations where the purified HSPG preparations were immunodepleted (ID), procedures were similar to those of Edgar et al. (6). E9 and El 1 HSPG preparations were mixed with anti-2.3D core protein antibody (10:1 v/v for two hours), mixed with Pansorbin (CalBiochem, 10:1 v/v for two hours), clarified, and the procedure repeated. Experiments with ^SO^labelled HSPGs demonstrated that greater than 88% of HSPGs are removed by this procedure. Equivalent volumes of the immunodepleted HSPGs were then added back to the cultures. Values are the means and standard deviations (SDs) of six determinations.

Figure 4 is a graphical representation of the dose-response relationship between increasing amounts of glycosaminoglycan fragments (X axis) and their effects on 2.3D cell proliferation (Y-axis) as monitored by [³H]-thymidine uptake (measured in cpm) after 24 hour exposure in tissue culture to a fixed concentration of FGF-2(5ng ml. in the dish), A DDS, no activity; ■ — ■ ODS, activity; • ^■ " ^• - ^• • HSPG , heparan sulphate.

Figure 5 is a graphical reproduction demonstrating that both the 6 disaccharide unit (12 mer, "DDS") and the 3 disaccharide unit (6 mer, "HeS") are capable of competitively inhibiting the mitogenic effects (as monitored by thymidine uptake in cpm) of the 9 disaccharide unit (18 mer, "DDS") on 2.3D cells in culture grown in a fixed concentration of FGF-2(5ng/ml).

Figure 6 is a graphical representation demonstrating further the specificity of the interaction between the 9 disaccharide unit (ODS; 18 mer) and the FGF receptor type 1 isoform. 2.3D cells were grown as before in culture in the presence of FGFs, sugars and receptor-blocking peptides. The curve marked FGF-2/ODS establishes the baseline positive control.

Figure 7 is a graphical representation showing rotational behaviour of grafted rats following amphetamine administration. Only rats implanted with FGF-2 in combination with GAGB showed a significant drop in the turning response induced by amphetamine.

♦ = PBS vehicle alone; • = BDNF alone; o = FGF2 alone; Δ = FGF2 with GAGB. The following abbreviations are used in the specification:

Abbreviation Meaning

SEQ ID NO Sequence Identity Number

FGF Fibroblast Growth Factor

FGF-1 Acidic FGF

FGF-2 Basic FGF

GAG Glycosaminoglycan HSPG Heparan sulfate proteoglycan

E9 Embryonic day 9

E10 Embryonic day 10

El l Embryonic day 11

ODS Octadecasaccharide (18 mer) HeS Hexasaccharide

DDS Dodecasaccharide

GAGA GAG polymer with binding specified for FGF 1

GAGB GAG polymer with binding specified for FGF 2

UA iduronic acid UA-GlcNAc iduronic acid N-acetylated glucosamine

UA-GlcNSO₃ iduronic acid N-sulfated glucosamine

UA-GlcNAc(6S) iduronic acid N-acetylated glucosamine 6-sulfate

UA-(2S)-GlcNAc iduronic acid 2-sulfate N-acetylated glucosamine

UA-GlcNSO₃(6S) iduronic acid N-sulfated glucosamine 6-sulfate UA-(2S)-GlcNSO₃ iduronic acid-2-sulfate N-sulfated glucosamine

UA-(2S)-GlcNSO₃(6S) iduronic acid 2-sulfate N-sulfated glucosamine 6-sulfate

UA-(2S)-GlcNAc(6S) iduronic acid 2-sulfate N-acetylglucosamine 6-sulfate The following single and three letter abbreviations are used for amino acid residues:

Amino Acid Three-letter One-letter Abbreviation Symbol

Alanine Ala A

Arginine Arg R

Asparagine Asn N Aspartic acid Asp D Cysteine Cys C Glutamine Gin Q Glutamic acid Glu E Glycine Gly G Histidine His H Isoleucine He I Leucine Leu L Lysine Lys K Methionine Met M Phenylalanine Phe F Proline Pro P Serine Ser S Threonine Thr T Tryptophan Trp w Tyrosine Tyr Y Valine Val V Any residue Xaa X EXAMPLE 1 PURIFICATION OF HEPARAN SULFATE PROTEOGLYCAN

Purification Protocol

FGF-2 specific HSPG is derived from the conditioned medium produced by the neuroepithelial cell line 2.3D, previously made by expressing the c-myc oncogene in embryonic day 10 primary neuroepithelial cells. The 2.3D neuroepithelial cell line is grown to 70% confluency, the cells pulsed with

overnight, the medium conditioned collected and clarified by centrifugation, and then passed through DEAE- Sepharose equilibrated in Tris-buffered saline (pH7.4). The column is washed with 10 column volumes of 0.25 M NaCi 0.1% v/v Triton X-100, the same wash supplemented with 8 M Urea, then 0.3 M sodium formate (pH 3.5)/8 M Urea, and then with 0.05 M Tris-HCl (pH 8)70.01% v/v Triton X-100. HSPGs are released from the column with an increasing gradient of NaCl (0.15-1.0 M) in 0.01 M Tris-HCl (pH 8)/0.01% v/v Triton X-100. A similar purification protocol is employed in the preparation of FGF-1 - specific HSPG except that the 2.3D cells are cultured under contact-inhibiting conditions by growing the cells to 100% confluency and then maintaining the cells under these conditions for 6 days before collecting the conditioned medium. A suitable method for growing cells to confluency is described in reference 13.

The proteoglycans were sized after Sepharose CL-6B gel chromatography. The proteoglycans present eluted as a single peak at approximately MW 450,000. The proteoglycan peak disappeared after both nitrous acid (pH 1.5) and heparanase III treatment, but not chondroitinase ABC. The proteoglycan is thus a heparan sulfate. When stripped of its side chains the core protein of the proteoglycan ran on SDS-PAGE gels at approximately 45,000. The side chains derived from non-confluent 2.3D cells averaged 20,000 daltons.

The core protein from this proteoglycan was treated with trypsin, carboxymethylated in 6M guanidine HC1 (pH 8.6), reduced with beta-mercaptoethanol (50°C, N2, 1 hour), alkylated, dialyzed against 5% v/v acetic acid, chromatographed on Biogel P10 and run on a reverse phase HPLC Zorbax OD5 and the fragments sequenced for amino acids on a gas phase sequenator. The core protein was also subjected to V8 protease for 6 hours, run on Sepharose 4B (0.2 M NaC170.02 M Tris-HCl (pH 8)), then on DEAE-5PW HPLC columns, reduced with dithiothreitol, alkylated with iodoacetamide, rerun on DEAE-5PW HPLC and the peaks similarly amino acid sequenced. Partial amino acid sequence obtained from the procedure is shown in Table 1.

TABLE 1 Heparan Sulfate Proteoglycan Core Protein Amino Acid Sequencing of Fragments Generated by V8 and Tryptic digestion

V8

V18: GASCEDCQTFYYGDAQRGTPQD [SEQ ID NO:l] Tryptic T3: GTPQDCQPCPCYGAPRRTTPA [SEQ ID NO:2]

For preparation of HSPG in a particular glycosylated form for binding to either FGF-1 or FGF-2, purification can conveniently occur for Ell and E9 neuroepithelial cultures (see Example 2), respectively.

Affinity Chromatography of HSPG using HSPG coupled to an Affi-Gel 10

Serum-free media conditioned over 2.3D cells, E9, or El 1 (see below) neuroepithelial cell cultures (10^ cells per 16-mm well for 24 hours) were filtered through 0.45 μ mesh and chromatographed through a low-pressure Econo-Pac Q Sepharose cartridge (Bio- Rad) at 2 ml/min. The column was washed with 0.15 M TBS, pH 7.4, until the absorbance at 280 nm reached baseline. The bound material was then released with an NaCl gradient from 0.15 to 1.0 M and collected in 3-ml fractions. In some experiments the cells were maintained in DMEM containing

in order to detect GAG side chains; in other experiments the cells were maintained in DMEM containing [³^S]methionine in order to detect HSPG core proteins. Purified HSPG preparations from either E9 or Ell conditioned media, or 2.3D cell media, were ligated to the affinity agarose support Affi-Gel 10 (Bio-Rad) in carbodiimide coupling buffer according to the manufacturers instructions. Approximately 100 μg of each HSPG preparation was bound to each 1 ml column volume of gel bed. The bound support was then decanted into small chromatographic columns and washed in 0.15 M Tris-buffered saline (TBS, pH 7.4). For each experiment, 100 ng of either ^{1 5}l-labelled FGF-1 or * -labelled FGF-2 in 0.15M TBS, was passed over the gel bed at a flow rate of 0.5 ml/min. The flow through material was collected as the unbound fraction and the column was then washed with 20 column volumes of 0.15M TBS. The bound fraction (2 ml) was then eluted from the column with 2.0 M NaCl/TBS, at a rate of 0.5 ml/min. The radioactivity in the applied (1.65 x 10⁶ cpm for FGF-1, 1.83 x 10⁶ cpm for FGF- 2), unbound, and bound fractions was then quantitated. Preliminary control experiments demonstrated that other factors such as transforming growth factor-βl, insulin-like growth factor-I, nerve growth factor, ciliary neurotrophic factor, leukemia inhibitory factor and the heparin-binding factor midkine were not retained by the HSPG columns. When the HSPG columns were pretreated with 10 μg heparitinase (Sigma) for 1 hour at 37°C, they lost their ability to bind either FGF-1 or FGF-2.

Analysis of Core Proteins of HSPG

To reveal the molecular weight of the core proteins, HSPG complexes collected after chromatography were pretreated with heparitinase III (heparanase I, EC 4.2.2.8, Sigma; 10 μg/ml, 37°C, 1 hour) before being applied to a 3-10 w/v SDS-po.yacrylamide gel. To prove the antigenic similarity of the E9 and El 1 core proteins a rabbit polyclonal antibody was raised against the 2.3D core protein (7). Affinity-purified antibodies were then used to immunoprecipitate the core proteins from both E9 and El l HSPG preparations according to the method of Edgar and coworkers (6). EXAMPLE 2 EFFECT OF HSPG ON FGF INTERACTION

The inventors discovered that neuroepithelial cells differentially regulate the expression of FGF during development. Studies were performed on mesencephalic and telencephalic neuroepithelial tissue at embryonic day 9, 10, 11 and 13 and this tissue is referred to herein as E9, E10, El l and El 3, respectively. In particular, the inventors showed FGF-2 expression in E9 and then subsequently FGF-1 expression in El l.

The potential role of HSPG in this differential expression of FGFs was then examined. E9 and El l neuroepithelial-derived GAGs were coupled to Affi-Gel 10 affinity support and decanted into small chromatographic columns. Both ^I-labelled FGF-1 and *^I- labelled FGF-2 were passed over respective column beds to which DEAE-purified HSPG from either E9 or El 1 conditioned medium, or from confluent and nonconfluent 2.3D cells, had been coupled. The amount of radioactivity applied to the columns was monitored, as were the number of counts that flowed through the column (the unbound fraction), and the number of counts recoverable after a 2 M NaCl wash (the bound fraction). The results are shown in Figure 1 and reveal a switch in the binding affinity of the HSPGs isolated from the two ages, and from the two states of confluency. HSPGs derived from E9 preparations were over four times more effective in binding FGF-2 compared with FGF-1, whereas HSPGs derived from El 1 neuroepithelium were six times more effective at binding FGF-1 compared with FGF-2. This change in binding specificity also occurred in 2.3D cells: nonconfluent 2.3D HSPG preferentially bound FGF-2, while the HSPGs isolated from 2.3D cells confluent for 6 days displayed a much greater capacity to bind FGF-1 than FGF-2.

To determine if this change in binding specificity was due to differential expression of HSPG core proteins or differential glycosylation of the same core protein, conditioned media were collected from E9 and El l cells maintained in either [^S]methionine to label the core protein or in -"SO4 to label the GAG side chains. When labelled E9 and El 1 HSPGs were stripped of GAGs with heparitinase, the core proteins appeared to have very similar molecular weights of about 45 kDa (Fig. 2A). This similarity was further substantiated by immunoprecipitation with an affinity-purified rabbit antibody raised against the core protein of the nonconfluent 2.3D HSPG. This antibody was able to precipitate core proteins with an identical molecular weight to those obtained from the original DEAE isolates from both E9 and El l (Fig. 2A) and 2.3D cells. These molecular weights together with preliminary amino acid sequencing of core protein fragments indicate that neuroepithelial cells secrete a single unique species of HSPG. When these core proteins were digested with trypsin, the resultant peptides yielded profiles on SDS-polyacrylamide gels and reversed phase high pressure liquid chromatography that were essentially identical.

Preliminary stoichiometric analysis of

GAG side chains after gel filtration showed that the average side chain from E9 conditioned medium was approximately 20 kDa, as opposed to the size of the GAG side chains at El 1 of 35 kDa (Fig. 2B). These data in association with the molecular weight data on the intact HSPGs and the core proteins (Fig. 2A) indicate that at E9 approximately 20 side chains are attached per core protein synthesized, whereas at El 1 approximately 12 side chains are attached per HSPG core protein. Overall, it appears that the same protein core is differentially glycosylated at the two ages and that the FGF binding specificity resides in the GAG chains of the HSPGs.

To test whether HSPG-binding specificity coincides with the ability of each factor to stimulate cell proliferation, El l neuroepithelial cells were isolated, pretreated with trypsin to exogenous HSPGs and attached growth factors, and then exposed to either FGF-1 or FGF-2 in the presence of HSPGs obtained from E9 or El l neuropithelium (Fig. 3). E l l HSPG was approximately four times more effective with FGF-2. This response to the HSPGs was dose-dependent within the range 0.01 to lOμg/ml.

In order to verify that the HSPG recognised by the anti-core protein antibody is actually mediating the biological activity, purified preparations of E9 or El l HSPGs were immunodepleted with the antibody and protein A before addition to the E10 neuroepithelial cultures. The immunodepleted HSPG preparations of both ages had greatly reduced mitogenic potential, demonstrating that the HSPGs are the biologically active molecules in the conditioned medium. It also indicates that the core proteins that mediate the E9 and El l response are immunologically identical. Control incubations with other antibodies did not significantly reduce the bioactivity. Thus the functional activity of the HSPGs reflect their binding specificities.

EXAMPLE 3 PURIFICATION OF HEPARAN SULFATE POLYMERS

Heparan sulfate polymers in the form of GAG chains of the E9 and El 1 HSPGs were prepared from a 100 μl sample of immunopurified proteoglycan layered onto a BioGel P-2 column equilibrated in TBS, collected in the void volume, and digested with 1 mg ml Pronase for 4 hours at 25°C. The samples were concentrated to 50 μl by dialysis against solid polyethylene glycol at 25°C for 2 hours, adjusted to 4 M guanidinine hydrochloride/50 mM Tris, pH 7.0. The eluted fractions were counted in Aquasol (NEN, Dupont, Sydney). The column was calibrated with samples of ^C-labelled dextran (70 kD), [³H] chondroitin sulphate (50 kD) and [³H]heparin (12 kD). The purification of the FGF-1- and FGF-2- specific heparan sulfate polymers is shown in Figure 2. The GAG side chains were in some cases repurified through Q Sepharose using similar procedures to those for total HSPG following the methods of Cole and Burg (8) and Kojima et al (9).

EXAMPLE 4 EFFECT OF HEPARAN SULFATE POLYMERS ON FGF INTERACTION

Using the model system of Nurcombe et al. (10), the therapeutic possibilities of the heparan sulfate polymers of the present invention were investigated on the normal histogenetic cell death phases of the motor neurons of the embryonic chick spinal cord.

The development of the vertebrate nervous system is characterized by an initial overproduction of neurons in many regions followed by their large-scale elimination. This phenomenon takes place at a particularly important stage during the development of embryonic neurons, the period immediately following the arrival of their axons in the specific target fields. Ideas current in neurotrophic theory place the basis of this cell death on a competition for limiting amounts of crucial trophic factors supplied by the target organ. However, to date only two defined trophic molecules have been shown to support embryonic neuronal survival in vivo - nerve growth factor and brain-derived neurotrophic factor.

Like other neurons, the somatic motor neurons of the spinal cord undergo naturally occurring cell death during embryonic development. Although motor neurons are insensitive to NGF there is circumstantial evidence that the survival of embryomc motor neurons is dependent on trophic substances within developing skeletal muscles. Skeletal muscle contains substances that enhance the survival and development of motor neurons in vitro.

In the present study, the inventors used the optical dissector of Gundersen et al. (11) to estimate the total number of neurons in the developing chick lumbar lateral motor column and to examine the effects of growth factors (FGF-2, CNTF, LIF, NGF) on neuronal number. In particular, the effect of FGF-2 alone or complexed with heparan sulfate polymers from E9 HSPG was investigated in the chick embryo spinal cord model.

Application in vivo of Purified Growth Factors

White Leghorn chick embryos were treated daily in ovo with either 0.9% w/v saline or purified growth factor in saline from E6 to E9. Each growth factor in a volume of 50 μl. was applied to the vascularized chorioallantoic membrane through a window in the shell as described by Oppenheim et al. (12).

Embryos treated with FGF-2 received either daily applications of 2 ug of recombinant human FGF-2 in 50 ul of 0.9% w/v saline, or the same FGF-2 that had been mixed on an orbital shaker with purified E9 HSPG-GAG chains at a molar ratio of 3:1 (GAG:FGF-2) in eppendorf tubes at room temperature for 2 hours prior to application to the vascularized chorioallantoic membrane. Fixation and Tissue Processing

The spinal cords were immersion fixed in Carnoy's fixative for 1 h, dehydrated in 100% w/v ethanol overnight and embedded in glycolmethacrylate (Polaron Embedding Medium, Bio Rad.).

Estimating the Number of Neurons

Neuron numbers were counted using an optical dissector (11).

Effect of GAG-FGF-2 Complex The results are seen in Table 2 and clearly show that heparan sulfate polymers from E9 complexed with FGF-2 (E9 GAG-FGF-2 complex) has the ability to rescue motor neurons during the period of cell death.

TABLE 2

Effects of GAG on the total number of neurons in the embryonic chick lateral motor column at E10

Saline control FGF-2 E9-GAG alone E9 GAG-FGF-2 2JD GAG fragment complex "18mer"+FGF-2 (n - 2) (n - 5) (n - 6) <n - 5) <n - 8)

Number of Motor

Neurons 10, 1221173 10-339i_X-4 10,027± 55 I4,065±l,0 0 14,9241862

EXAMPLE S COMPOSITIONAL ANALYSIS OF GAG POLYMERS

FGF-2-specific GAG polymer was subjected to compositional analysis as follows:

Cell culture, radiolabelling and preparation of intact heparan sulfate chains

Neuroepithelial cells were grown in 0.5 ml 10% v/v FCS/DMEM and 2 ng/ml FGF-2 in 24 well tissue culture plates at a density of 100,000 cells/well. The cells were allowed to settle in a 10% v/v CO/air-humidified incubator for 30-60 min before addition of 20 uCi/ml [H³]glucosamine. Wells were monitored daily for contamination or excessive cell death (over 50%) and those cells and media discarded. Cultures were further incubated for 50-60 hours. The medium was gently removed and centrifuged (1000 rpm for 5 min) to remove any cell debris and stored at -20°C until required. The conditioned media was then subjected to ion-exchange chromatography on a DEAE- Sephacel column (2 ml) which had previously been blocked with heparin and equilibrated in 150 mM NaCl with phosphate buffered saline, pH 7.2. The sample was then washed with ten column volumes of 250 mM NaCl in 50 mM phosphate buffered saline, pH 7.2. The bound material (primarily heparan sulfate, chondroitin sulfate and dermatan sulfate) was eluted in a step elution at 1 M NaCl in 50 mM phosphate buffered saline, pH 7.2 and 2 ml fractions collected.

Fractions containing the tritiated glucosamine (primarily fractions 1-3) were pooled and desalted on Amicon concentration cones, freeze dried and resuspended in miiiimal volume (100-500 ul maximum). Sialic acid was removed with neuraminidase in 25 mM Na-acetate pH 5.0, for 4 hours. Chondroitin sulfate and dermatan sulfate were digested with chondroitin ABC lyase treatment for 4 hours at 37°C and a further digest overnight with fresh enzyme. The core protein and all of the lyases are digested with Pronase at 37°C for 24 h and the sample passed through a 2 ml dEAE-Sephacel column and eluted as previously described while collecting 1 ml fractions. The sample was finally desalted on a 1 cm x 35 cm P2 column and the Vo fraction collected and freeze dried for further analysis.

Gel chromatography

Low pressure gel chromatography of intact chains or treated chains (oligosaccharide fragments) was performed on Sepharose (C1-6B, Bio-Gel P-2 and Bio-Gel P10 columns. Resolving columns were 120 cm x 1 cm, in 0.5 M NH4HCO3 eluted at 4 ml/hr and 1 mi fractions collected. Fractions corresponding to disaccharides and/or tetrasaccharides were pooled and freeze dried if required for further analysis.

Depolymerisation of HSPGs

Heparitinase (heparitinase I), heparitinase II and heparitinase IV were used at a concentration of 25 m units ml in 100 mM-sodium acetate/0.2 mM-calcium acetate, pH 7.0. Samples were incubated at 37°C for 16 h and then a second aliquot added and incubated a further 4 hours. Heparinase was used at a concentration of 50 m units/ml in the same buffer as heparitinase. Sequential digests for recovery of disaccharides for SAX-HPLC containing 100 ug non-labelled heparan sulfate as a carrier and were digested at 37°C as follows: heparinase for 2 h then heparitinase for 1 hour then heparitinase II for 18 hours and finally, an aliquot of each lyase and heparitinase IV for six hours. Samples were then dried down to < 100 ul and run on a P2 to ensure >95% disaccharide recovery. Deaminative cleavage was carried out with low-pH HNO₂ as described by Shively and Conrad (16).

SAX-HPLC of disaccharides and tetrasaccharides

Disaccharide composition was analysed by either complete depolymerisation of the entire heparan sulfate chain with heparitinase, heparitinase II, heparitinase IV and heparinase (yields >/+ 95%) or subjected to nitrous acid (pH 1.5) so that both the disaccharide fraction and the tetrasaccharide fraction could be separately collected. These pools were freeze dried and resuspended in 300 ul water. The disaccharides or tetrasaccharides were then separated by SAX-HPLC on either one or two ProPac PA1 analytical columns (4 x 250 mm; Dionex, Surrey, United Kingdom). After equilibration in mobile phase (double-distilled water adjusted to pH 3.5 with HC1) at 1 ml min, samples were injected and disaccharides eluted with a linear gradient of sodium chloride (0-1M over 45 min) in the same mobile phase. The eluant was monitored in-line for the nitrous acid derived disaccharides (H3 radioactivity by a radiomatic Flo-one/Beta A-200 detector, Canberra packard, Pangbourne, United Kingdom) or 0.5 ml fractions were collected and counted on a scintillation machine for the lyase derived disaccharides and the tetrasaccharides.

Oligosaccharide mapping by gradient page

Radiolabelled heparan sulfate that had been treated with a variety of reagents was mapped by gradient PAGE as described previously by Turnbull and Gallagher (1988) with some modifications. Briefly, 25-33% w/v -polyacrylamide-gradient gels (32cm x 16cm x 0.75mm) were prepared with a 5% w/v stacking gel. Samples were electrophoresed as previously described until the phenol red marker was about 1 cm from the bottom. Gel was equilibrated in 10 mM Tris/acetate buffer containing 0.5 mM- EDTA for 10-20 min. Oligosaccharides were then transferred onto a positively charged nylon membrane (Biotrace RP) in a Trans-blot tank at low voltage in the same buffer for 3-4 hours. The oligosaccharides were detected by fluorography of the membrane by using Enhance surface autoradiography enhancer and Kodak X-Omat AR X-ray film.

The compositional analysis of the FGF-2-specific GAG polymer (GAGB) is shown in Table 3. The disaccharide compositions shown in Table 3 are as derived from HPLC analysis of various preparations of heparan sulfate. The numbers represent the average of two runs for the 2.3D cell-derived samples from lyase-digested samples. Samples were digested extensively with the heparitinase 1 -4 and disaccharides isolated from a P2 column and then freeze dried. Over 97% disaccharides were recovered from each sample. Run on only IX propac column, l-46min=0-l M NaCl (pH 3.0), max. sample volume was 500 ul and the loop was 1 ml. A similar analysis was conducted for the FGF-1 -specific GAG polymer (GAGA) and the results are shown in Table 4.

TABLE 3

PEAK DISACCHARIDE % GAGB NUMBER

1 UA-GlcNAc 55.4

10 3 UA-GlcNSO₃ 22.2

2 UA-GlcNAc(6S) 3.2

7 UA-(2S)-glcNAc 1.8

4 UA-GlcNSO₃(6S) 2.5

5 UA-(2S)-GlcNSO₃ 9.0

15 UA-(2S)-GlcNAc(6S) 0

•

6 UA-(2S)- 5.1 GlcNSO₃(6S)

9 unknown 0.7

TABLE 4

PEAK NUMBER DISACCHARIDE % GAGA

1 UA-GlcNAc 50.7

3 UA-GlcNSO₃ 19.1

2 UA-GlcNAc(6S) 4.7

7 UA-(2S)-GlcNAc 2.6

4 UA-GlcNSO₃(6S) 2.8

5 UA-(2S)-GlcNSO₃ 9.1

8 UA-(2S)-GlcNAc(6S) 0

6 UA-(2S)-GlcNSO₃(6S) 5.8

9 unknown 5.1

EXAMPLE 6

Figure 4 is a graphical representation of the dose-response relationship between increasing amounts of glycosaminoglycan fragments (X axis) and their effects on 2.3D cell proliferation (Y-axis) as monitored by [³H]-thymidine uptake (measured in cpm) after 24 hour exposure in tissue culture to a fixed concentration of FGF-2(5ng/ml, in the dish). As can be seen from the trace marked "HSPG", the full proteoglycan shows classical pharmacokinetics - the presence of increasing amounts of HSPG triggers greater cell proliferation up to a plateau, before inhibition sets in. This can be interpreted to mean that very high doses of sugar shifts the equilibrium such that the FGF-2 becomes more interested in low affinity binding to sugar than to a low number of high affinity receptors on the cells. The trade marked "ODS" (for octadecasaccharide) shows that molecule for molecule, the 9 disaccharide unit carries all the bioactivity of the native HSPG; its kinetics are directly comparable. If the 9 disaccharide unit is cleaved with heparanase I into a 6 disaccharide (dodecasaccharide, "DDS") unit and a 3 disaccharide (hexasaccharide, "HeS") unit, neither of these carries the ability to stimulate cell division by itself.

EXAMPLE 7

Figure 5 demonstrates that both the 6 disaccharide unit (12 mer, "DDS") and the 3 disaccharide unit (6 mer, "HeS") are capable of competitively inhibiting the mitogenic effects (as monitored by thymidine uptake in cpm) of the 9 disaccharide unit (18 mer, "ODS") on 2.3D cells in culture grown on a fixed concentration of FGF-2(5ng/ml). At a molar ratio in culture of one ODS to one DDS or one HeS, there is no dimimshment of effect, but even at the comparatively low challenging ratio of 5 of the smaller molecules to one ODS fragment, we lose substantial activity. This indicates the extreme specificity of all the fragments, and the involvement of separable domains within the ODS to trigger activity.

EXAMPLE 8

Figure 6 demonstrates further the specificity of the interaction between the 9 disaccharide unit and the FGF receptor type 1 isoform. 2.3D cells were grown as before in culture in the presence of FGFs, sugars and receptor-blocking peptides. The curve marked FGF-2/ODS establishes the baseline positive control. The subsequent experiments are performed in the presence of increasing concentrations of the peptide "K22", a 22 amino acid peptide established by Kan et al. (15) to represent the portion of the FGFR1, designated Ig domain 2, which engages the sugar before the FGF docks with the receptor in the Ig3 domain. The second curve shows that cells grown in the presence of the E12 (ie. FGF-1 specific) HSPG and FGF-1 cannot be inhibited in their growth by the presence of the FGFR1 -specific K22K peptide. Therefore the FGF-1- specific GAG is not using this receptor for signal transduction, unlike FGF-2. The next curve, designated ODS/FGF-1/FGFR1 also shows a lack of inhibition of growth, demonstrating both that the ODS does not potentiate FGF-1 (the growth plateau is lower than for the first 2 curves), and that K22K FGFR1 -specific peptides have no effect on this growth. The last curve, marked ODS/FGF-2/FGFR1 shows increasing inhibition of growth in the presence of increasing amounts of the FGFR1 -specific peptide. Thus, the ODS stimulates the effects of FGF-2 specifically, and does so through the FGFR1 receptor.

EXAMPLE 9 AMELIORATION OF AN INDUCED PARKINSONISM IN RATS

Animals

Pathogen-free Sprague Dawley rats (body weight 158-163 g) were kept under regular day and night conditions at constant 23 °C temperature with free access to food pellets and water.

Lesion

Experiments were performed essentially as in Takiyama et al (17). Unilateral destruction of dopaminergeric neurons in the substantial nigra of adult rats was achieved by stereotaxic injection of 6-hydroxydopamine into the medical forebrain bundle. Rats were anaesthetised with a mixture of ketamine (75 mg/kg), xylazine (4 mg/kg) and acepromazine (5.6 mg/kg) for the procedure. The toxin was dissolved at a concentration of 6 mg/ml in saline and 1.5 ul was injected at 4.3 mm posterior to bregma, 1.5 mm laterial and 7.3 mm below dura, the syringe was raised 0.2 mm and another 1.5 ul was injected at -7.1 mm. A canula connected to an osmotic Alzet minipump (model 2002; Alza Corporation, Palo Alto CA) was then implanted in the brainstem immediately following the lesion; the pumps were loaded with one of four test substances listed below. Approximately one week after the lesion, control uninfused but chemically lesioned rats were tested for turning behaviour with amphetamine sulfate (5 mg/kg). Infused rats were divided into groups that were matched for rotational scores: infusion with saline (phosphate-buffered saline: PBS) vehicle alone; infusion with saline plus brain-derived neurotrophic factor (BDNF; 100 ng/ml); infusion with saline plus FGF-2 (100 ng/ml); or infusion with FGF-2 supplemented with GAGB (10 ug ml). Results are shown in Figure 7.

Histology

At the conclusion of behavioural testing rats were deply anaesthetised and perfused with 4% v/v paraformaldehyde in 100 mM PBS. Brains were removed, sectioned coronally at 40 um on a freezing microtome, stained for tyrosine hydroxylase or Nissl substance and quantitated sterologically according to the methods of Janson and Moller (18).

EXAMPLE 10

GAGB may be used in any situation where cells synthesize FGFs including cells genetically engineered to express FGFs, are transplanted into brain parenchyma to relieve the symptoms of neurological disorders such as Huntington' s Disease or Parkinson's Disease, or afflictions which involve parkinsonism. In this disease state, where there is loss of dopaminergic cells of the substantial nigra, transplanted cells have been shown to ameliorate the behavioural motor deficits which ensue (17). One potential method for increasing the viability of dopamine neurons after grafting may be to supply the cells with trophic support such as that supplied by FGF-2. That is, the FGF-2 cDNA, spliced into a retroviral vector under the control of a promoter such as the long terminal repeat (LTR) or a constitutive promoter such as actin, accompanied by an antibiotic resistant gene, may be used to stably transfect suitable cells. The FGF cDNA may contain additional sequences from pre-pro regions of secreted growth molecules such as nerve growth factor to enhance the extracellular secretion of the FGF.

The present invention demonstrates that supplying the grafted cells with additional glycosyaminoglycans that selectively couple FGF-2 with FGFR1 greatly increases the viability of such grafts. Furthermore, supplying GAGB to any such neural transplant expressing FGF will greatly increase the benefits of such grafts, both anatomically and behaviourally. In addition, the grafts may involve non-neural cells such as fibroblasts to carry the FGF-2 gene into the damaged brain tissue. This will be a useful strategy for enhancing the clinical effectiveness of dopaminergic treatments based on FGF-2 neurotrophic activity. Those skilled in the art will appreciate that the invention described herein is susceptible to variations and modifications other than those specifically described. It is to be understood that the invention includes all such variations and modifications. The invention also includes all of the steps, features, compositions and compounds referred to or indicated in this specification, individually or collectively, and any and all combinations of any two or more of said steps or features.

REFERENCES

la. Murphy, Drago, Bartlett, Neurosci. Res.2S, 463 (1990);

lb. Bartlett, Reid, Bailey, Bernard, Proc. Natl. Acad. Sci U.S.A. 85, 3255 (1988).

2a. Kjellen and Lindahl, Annu. Rev. Biochem. 60, 443 (1991);

2b. Ruoslahti, Annu. Rev. Cell Biol. 4, 229 (1988).

3a. Esko, Curr. Opin. Cell Biol 3, 805 (1991);

3b. Jackson, Busch, Cardin, Physiol. Rev.. 71, 481 (1991).

4. Flaumenhaft and Rifkin, Curr. Opin. in Cell Biol. 3, 817 (1991).

5a. Yayon, Klagsbrun, Esko, Leder, Ornitz, Cell 64, 841 (1991);

5b. Rapraeger, Krufka, Olwin, Science 252, 1705;

5c. Ruoslahti, Yamaguchi, Cell 64, 867 (1991);

5d. Klagsbrun and Baird, ibid. 67, 229 (1991).

6. Edgar, Timpl, Thoenen, J. Cell Biol. 106, 1299 (1988).

7. Marynen, Zhang, Cassiman, Van den Berghe, David, J. Biol. Chem. 264, 7017 (1989).

8. Cole and Burg. Exp. Cell Res. 182: 44-60 (1989).

9. Kojima et al. J. Biol. Chem. 267: 4859-4868 (1992). 10. Nurcombe et al. Anatomical Record 231: 416-424 (1991).

11. Gundersen et al. APMIS 92: 857-881 (1988).

12. Oppenheim et al. Science 240: 919-922 (1988).

13. Methods in Neurosciences: Volume 2. Cell Culture P. Michael Conn, ed. Academic Press. San Diego, 1990.

14. Turnbull et al. J. Biol. Chem. 267: 10337-10341 (1992).

15. Kan et fl/. Science 259: 1918-1921, (1993).

16. Shively and Conrad, Biochem. 15: 3932-3939, (1976).

17. Takiyama et al. Nature Medicine 1: 53-58. (1995).

18. Janson and Moller, Neuroscience 57: 931-941, (1993).

SEQUENCE LISTING

(1) GENERAL INFORMATION:

(i) APPLICANT: THE UNIVERSITY OF MELBOURNE and

THE WALTER AND ELIZA HALL INSTITUTE OF MEDICAL RESEARCH

(ii) TITLE OF INVENTION: "A THERAPEUTIC MOLECULE"

(iii) NUMBER OF SEQUENCES: 2

(iv) CORRESPONDENCE ADDRESS:

(A) ADDRESSEE: DA VIES COLLISON CAVE

(B) STREET: 1 LITTLE COLLINS STREET

(C) CITY: MELBOURNE

(D) STATE: VICTORIA

(E) COUNTRY: AUSTRALIA

(F) ZIP: 3000

(v) COMPUTER READABLE FORM:

(A) MEDIUM TYPE: Floppy disk

(B) COMPUTER: IBM PC compatible

(C) OPERATING SYSTEM: PC-DOS/MS-DOS

(D) SOFTWARE: Patentln Release #1.0, Version #1.25

(vi) CURRENT APPLICAΗON DATA:

(A) APPLICATION NUMBER: PCT INTERNATIONAL

(B) FILING DATE: 25-JAN-1996

(vii) PRIOR APPLICATION DATA:

(A) APPLICAΗON NUMBER: PN0784

(B) FILING DATE: 27-JAN-1995

(A) APPLICAΗON NUMBER: PN3560

(B) FILING DATE: 16-JUN-1995

(viii) ATTORNEY/AGENT INFORMATION: (A) NAME: HUGHES DR, E JOHN L

(C) REFERENCE/DOCKET NUMBER: EJH/EK

(ix) TELECOMMUNICATION INFORMATION:

(A) TELEPHONE: +61 3 9254 2777

(B) TELEFAX: +61 3 9254 2770 (2) INFORMATION FOR SEQ ID NO:1 :

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 22 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS : single

(D) TOPOLOGY: linear (ii) MOLECULE TYPE: protein

(xi) SEQUENCE DESCRIPTION: SEQ ID NO: 1 :

Gly Ala Ser Cys Glu Asp Cys Gin Thr Phe Tyr Tyr Gly Asp Ala Gin 1 5 10 15

Arg Gly Thr Pro Gin Asp 20

(2) INFORMATION FOR SEQ ID NO:2:

(i) SEQUENCE CHARACTERISTICS:

(A) LENGTH: 21 amino acids

(B) TYPE: amino acid

(C) STRANDEDNESS: single

(D) TOPOLOGY: linear (ii) MOLECULE TYPE: protein

(xi) SEQUENCE DESCRIPTION: SEQ ID NO:2:

Gly Thr Pro Gin Asp Cys Gin Pro Cys Pro Cys Tyr Gly Ala Pro Arg 1 5 10 15

Arg Thr Thr Pro Ala 20

Claims

CLAIMS:

1. An isolated glycosaminoglycan (GAG) polymer or derivative thereof from a heparan sulfate proteoglycan (HSPG) wherein said GAG polymer or is derivative is capable of interaction with a cytokine.

2. An isolated GAG polymer or derivative thereof according to claim 1 wherein the GAG polymer or its derivative is capable of interaction separately with either Fibroblast Growth Factor (FGF)-l or FGF-2 but not both.

3. An isolated GAG polymer or derivative thereof according to claim 2 wherein said GAG polymer or its derivative is capable of interaction with FGF-2 and wherein said HSPG is obtainable from murine cells at embryonic day from about 8 to about 10.

4. An isolated GAG polymer or derivative thereof according to claim 2 wherein said GAG polymer or its derivative is capable of interaction with FGF-1 and wherein said HSPG is obtainable from murine cells at embryonic day from about 11 to about 13.

5. An isolated GAG polymer or derivative thereof according to claim 3 comprising the following disaccharides in percentage amounts given in parentheses: iduronic acid N-acetylated glucosamine [UA-GlcNAc] (55.4%), iduronic acid N-sulfated glucosamine [UA-GlcNSO₃] (22.2%), iduronic acid N-acetylated glucosamine 6-sulfate [UA-GlcNAc (6S)] (3.2%), iduronic acid 2-sulfate N-acetylated glucosamine [UA-(2S)- GlcNAc] (1.8%), iduronic acid N-sulfated glucosmine 6-sulfate [UA-GlcNSO₃(6S)] (2.5%), iduronic acid 2-sulfate N-sulfated glucosamine [UA-(2S)-GlcNSO₃] (9.0%) and iduronic acid 2-sulfate N-sulfated glucosamine 6-sulfate [UA-(2S)-GlcNSO₃(6S)] (5.1 %).

6. An isolated GAG polymer or derivative thereof according to claim 4 comprising the following disaccharides in percentage amounts given in parentheses: iduronic acid N-acetylated glucosamine [UA-GlcNAc] (50.7%), iduronic acid N-sulfated glucosamine [UA-GlcNSO₃] (19.1%), iduronic acid N-acetylated glucosamine 6-sulfate [UA-GlcNAc (6S)] (4.7%), iduronic acid 2-sulfate N-acetylated glucosamine [UA-(2S)- GlcNAc] (2.6%), iduronic acid N-sulfated glucosmine 6-sulfate [UA-GlcNSO₃(6S)] (2.8%), iduronic acid 2-sulfate N-sulfated glucosamine [UA-(2S)-GlcNSO₃] (9.1%) and iduronic acid 2-sulfate N-sulfated glucosamine 6-sulfate [UA-(2S)-GlcNSO₃(6S)] (5.8%).

7. An isolated GAG polymer or derivative thereof according to claim 3 wherein the murine cells are embryonic day 8-10 neuroepithelial cells transformed with an oncogene in a retroviral vector.

8. An isolated GAG polymer or derivative thereof according to claim 4 wherein the murine cells are embryonic day 11-13 neuroepithelial cells transformed with an oncogene on a retroviral vector.

9. An isolated GAG polymer or derivative thereof according to claim 3 or 4 wherein the murine cells are 2.3D cells deposited at the PHLS Center for Applied Microbiology and Research, European Collection of Animal Cell Cultures (ECACC), Division of Biologies, Porton Down, Salisbury, Wiltshire, SP4 OJG on 16 May, 1995 under Accession Number 95051601.

10. An isolated GAG polymer or derivative thereof according to claim 3 or 9 wherein the murine cells are dividing freely in tissue culture.

11. An isolated GAG polymer or derivative thereof according to claim 4 or 9 wherein the murine cells are grown in culture under contact inhibiting conditions.

12. An isolated molecule comprising:

(i) a repeating disaccharide structure (X-Y)_n wherein:

X is hexuronic acid;

Y is glucosamine; and n is 2 to 200; (ii) an ability to bind either FGF-1 or FGF-2 but not both; and (iii) being isolatable from an HSPG which in one form comprises GAG polymers capable of binding FGF-1 and in another form comprises GAG polymers capable of binding FGF-2.

13. An isolated molecule according to claim 12 wherein X-Y are α,β-linked glucosamine and hexuronic acid in linkage sequence [(l→4)α-D-glucosaminyl-(l-»4)β- D-hexuronosyl]_n.

14. An isolated molecule according to claim 12 or 13 wherein the molecule binds to FGF-2 and n is from about 2 to about 20.

15. An isolated molecule according to claim 14 wherein n is from about 8 to about 12.

16. An isolated molecule according to claim 15 wherein n is about 9.

17. A non-full length derivative of a GAG polymer according to claim 1 or 3 or 10 or 12 comprising from about 5 to about 9 disaccharides in length, said derivative capable of binding to FGF-2 or of inhibiting formation of an activating FGF-2-heparan sulfate- FGF-2 receptor tertiary complex.

18. A non-full length derivative of a GAG polymer according to claim 17 comprising at least 9 disaccharide units and capable of interaction with FGF-2.

19. A non-full length derivative of a GAG polymer according to claim 17 comprising at least about 5 disaccharide units but less than 9 disaccharide units and capable of blocking formation of an activating FGF-2-heparan sulfate-FGF-2 receptor tertiary complex.

20. A method of purifying a GAG polymer or a derivative thereof capable of binding either FGF-1 or FGF-2, said method comprising generating a neuroepithelial cell line expressing an oncogene and growing and/or maintaining the cell line for a time and under conditions sufficient for said cell line to secrete HSPG molecules into the conditioned medium; collecting the HSPG at predetermined time intervals and subjecting same to HSPG isolating means; subjecting isolated HSPG to GAG polymer purification means.

21. A method according to claim 20 wherein the purification means comprises treating HSPG with a proteolytic enzyme and isolating the GAG polymers.

22. A method according to claim 21 wherein the proteolytic enzyme is pronase.

23. A method according to claim 20 wherein the cell line is the 2.3D cell line deposited at the PHLS Center for Applied Microbiology and Research, European Collection of Animal Cell Cultures (ECACC), Division of Biologies, Porton Down, Salisbury, Wiltshire, SP4 OJG on 16 May, 1995 under Accession Number 95051601.

24. A method according to claim 23 wherein the cell line is grown to from about 50% to about 90% confluency.

25. A method of promoting, stimulating and/or enhancing interaction between a particular cytokine and a target site on a cell in an animal, said method comprising administering to said animal a GAG polymer or derivative thereof which preferentially binds to said cytokine, for a time and under conditions sufficient for said GAG polymer or its derivatives to promote binding of said cytokine with said target sequence.

26. A method according to claim 25 wherein the cytokine is FGF-1 or FGF-2.

27. A method for promoting, stimulating and/or enhancing cell proliferation, migration and/or differentiation of any tissue which bears the appropriate FGF receptors in an animal said method comprising the administration of a GAG polymer or derivative thereof wherein said GAG polymer or its derivative interacts with FGF-1 or FGF-2 but not both.

28. A method of promoting or facilitating maintenance and survival of neuronal cells in an animal, said method comprising the administration of a GAG polymer or derivative thereof wherein said GAG polymer or its derivative interacts with FGF-1 or FGF-2 bui not both.

29. A method according to any one of claims 25 to 28 wherein the cells are motor neurons and the effect of the GAG polymer or its derivative in combination with FGF-1 or FGF-2 is to rescue motor neurons during the period of cell death.

30. A method according to claim 25 or 26 wherein the GAG polymer maintains cells in a viable state.

31. A method according to claim 25 or 26 wherein the GAG polymer prevents or delays cell death.

32. A method for rescuing neurons during the period of cell death in a mammal, said method comprising administering to said mammal an effective amount of a GAG polymer or a derivative thereof from an HSPG obtainable from embryomc day 8-10 cells and wherein said GAG polymers or derivative is capable of interaction with FGF-2 but not FGF-1.

33. A method for rescuing neurons during the period of cell death in a mammal, said method comprising administering to said mammal an effective amount of a GAG polymer or a derivative thereof from an HSPG obtainable from embryonic day 11-13 cells and wherein said GAG polymers or derivative is capable of interaction with FGF-1 but not FGF-2.

34. A method for promoting the viability of cells carrying an FGF receptor type RlIIIc receptor for FGF-2 in a mammal, said method comprising administering to said mammal an effective amount of a GAG polymer or a derivative thereof from an HSPG obtainable from embryonic day 8-10 cells and wherein said GAG polymers or derivative is capable of interaction with FGF-2 but not FGF-1.

35. A method for promoting the viability of cells carrying an FGF receptor type RlIIIc receptor for FGF-2 in a mammal, said method comprising administering to said mammal an effective amount of a GAG polymer or a derivative thereof from an HSPG obtainable from embryonic day 11-13 cells and wherein said GAG polymers or derivative is capable of interaction with FGF-1 but not FGF-2.

36. A method according to claim 32 or 34 wherein the mammal is a human.

37. A method according to claim 32 or 34 wherein the cells are 2.3D cells deposited at the PHLS Center for Applied Microbiology and Research, European Collection of Animal Cell Cultures (ECACC), Division of Biologies, Porton Down, Salisbury, Wiltshire, SP4 OJG on 16 May, 1995 under Accession Number 95051601.

38. A method of ameliorating the effects or symptoms of neurological disorders in a mammal, said method comprising transplanting cells capable of synthesizing FGF-2 to a neurological environment and contacting said transplanted cells with an effective amount of a GAG polymer capable of interaction with FGF-2 but not FGF-1 and obtainable from an HSPG from murine cells at approximately embryonic day 8-10.

39. A method of ameliorating the effects or symptoms of neurological disorders in a mammal, said method comprising transplanting cells capable of synthesizing FGF- 1 to a neurological environment and contacting said transplanted cells with an effective amount of a GAG polymer capable of interaction with FGF-1 but not FGF-2 and obtainable from an HSPG from murine cells at approximately embryonic day 11-13.

40. A method of ameliorating the effects or symptoms of neurological disorders in a mammal, said method comprising transplanting cells genetically engineered to express FGF-2 to a neurological environment and contacting said transplanted cells with an effective amount of a GAG polymer capable of interaction with FGF-2 but not FGF-1 and obtainable from an HSPG from murine cells at approximately embryonic day 8-10.

41. A method of ameliorating the effects or symptoms of neurological disorders in a mammal, said method comprising transplanting cells genetically engineered to express FGF-1 to a neurological environment and contacting said transplanted cells with an effective amount of a GAG polymer capable of interaction with FGF-1 but not FGF-2 and obtainable from an HSPG from murine cells at approximately embryonic day 11-13.

42. A method according to claim 38 or 39 or 40 or 41 wherein the murine cells are 2.3D cells deposited at the PHLS Center for Applied Microbiology and Research, European Collection of Animal Cell Cultures (ECACC), Division of Biologies, Porton Down, Salisbury, Wiltshire, SP4 OJG on 16 May, 1995 under Accession Number 95051601.

43. A method according to claim 38 or 39 or 40 or 41 wherein the mammal is a human.

44. An antagonist of FGF-2-receptor interaction comprising a fragment of a GAG polymer obtained from an HSPG from murine cells at approximately embryonic day 8- 10.

45. An antagonist according to claim 44 comprising a fragment of GAGB.

46. An antagonist according to claim 45 wherein the fragment is less than 9 disaccharides in length.

47. An antagonist of FGF-1 -receptor interaction comprising a fragment of a GAG polymer obtained from an HSPG from murine cells at approximately embryonic day 11- 13.

48. An antagonist according to claim 44 comprising a fragment of GAGA.

49. An antagonist according to claim 45 wherein the fragment is less than 9 disaccharides in length.