WO1993018064A1 - Endothelial cell-derived differentiation modulating factors, their preparation and use - Google Patents

Abstract

The invention provides a basic secreted polypeptide factor which has, if obtained from bovine aortic endothelial cells, an observed molecular weight, of from about 29 kD to about 34 kD on SDS-polyacrylamide gel electrophoresis using the following molecular weight standards: Phosphorylase b 97,400; Bovine serum albumin 60,000; Ovalbumin 45,000; Carbonic anhydrase 31,000; Soybean Trypsin Inhibitor 21,000; Lysozyme 14,000, which factor has the ability to promote the differentiation of 0-2A progenitor cells into type-2 astrocytes when applied to 0-2A progenitors growing in serum-free medium, and which promotion of astrocytic differentiation is more rapid when said factor is applied to 0-2A progenitors in the presence of CNTF or LIF. The invention includes formulations encompassing this factor and its prophylactic and therapeutic use.

Description

Endothelial Cell-Derived Differentiation Modulating Factors, Their Preparation and Use.

This invention relates to new polypeptides characteristic of vertebrate species, which polypeptides are capable of promoting particular patterns of differentiation in cultured glial precursor cells of the central nervous system and also are capable of enhancing the effectiveness of other differentiation modulating factors and/or also are capable of altering the effect of other differentiation modulating factors. The invention is also concerned, inter alia, with the isolation process capable of producing such factors, and the therapeutic application of such factors.

The precise control of differentiation of vertebrate precursor cells into particular cell types is essential to the normal development of the organism, to maintenance of healthy tissue and to the promotion of repair. The differentiation of many cell types can be modulated by the presence of exogenous factors which can cause precursor cells to differentiate along one or another of their possible pathways. An increasing number of factors capable of modulating cellular differentiation have been identified in the past decade, and some of these factors have been found to be of considerable clinical usefulness. It has also begun to be appreciated that the action of particular modulators of differentiation may require the additional presence of other factors in order for effects to be observed. These additional factors may function to enhance the activity of a particular molecule, and may also function to cause particular effects to be elicited from a cell by application of exogenous factors.

Thus, Lillien et al, in The Journal of Cell Biology (Lillien, L. E., Sendtner, M., and Raff, M. C. (1990) J. Cell Biol. 111 , 635-644), describe the finding that extracellular matrix produced by cultures of bovine aortic endothelial cells or rat meningeal cells will cooperate with ciliary neurotrophic factor (CNTF) to cause oligodendrocyte-type-2 astrocyte (O-2A) progenitor cells (Raff, M. C, Miller, R.H. and Noble, M.

(1983a) Nature 303, 390-396) isolated from optic nerves of perinatal rats to differentiate into type-2 astrocytes, a particular type of glial cell (Raff, M. C, Abney, E. A., Cohen, J., Lindsay, R., and Noble, M. (1983b) J. Neurosci 3, 1289-1300). Application of CNTF on its own produced a transient expression of glial fibrillary acidic protein (GFAP), a marker of astrocytic differentiation. Cells grown in chemically-defined medium, and exposed to CNTF, did not however remain GFAP⁺. Instead, after 3 days, the O-2A progenitors had turned off GFAP expression and had differentiated into oligodendrocytes (Hughes, S.M. and Raff (1987) Development 101, 157-167; Lillien, L. E., Sendtner, M., Rohrer, H., Hughes, S. M. and Raff, M. C. (1988) Neuron 1 , 485-494; Lillien, L. E., Sendtner, M., and Raff, M. C. (1990) J. Cell Biol. 111, 635-644). The achieving of full astrocytic differentiation required the additional presence of a factor, or factors, contained within the extracellular matrix produced by heterogeneous cultures of CNS cells. Examination of the extracellular matrix material produced by individual cell types of the sort which might be expected to be present in such cultures demonstrated that the biological activity was only present in extracellular matrix prepared by endothelial cells (in this case cultures of bovine aortic endothelial cells) and cultures of meningeal cells. Beyond the demonstration that the biologically active material(s) could be removed from the extracellular matrix by washing matrix preparations with 0.5M NaCl, no characterization of the biologically active material was described (Lillien, L. E., Sendtner, M., and Raff, M. C. (1990) J. Cell Biol. 111 , 635-644). This invention is based upon the discovery that endothelial cells prepared from a variety of different sources secrete a soluble activity which promotes the astrocytic differentiation of O-2A progenitors in vitro. This activity is secreted, inter alia, by rat cerebral capillary endothelial cells, mouse cerebral capillary endothelial cells, bovine adrenal gland capillary endothelial cells and bovine aortic endothelial cells.

Thus, in one aspect, the invention provides a basic polypeptide composition of matter which, if obtained from bovine aortic endothelial cells, is a secreted material with an observed molecular weight of about 29-34 kD. This molecular weight can be observed by SDS-polyacrylamide gel electrophoresis, using the following molecular weight standards:

Phosphorylase b 97,400

Bovine serum albumin 60,000

Ovalbumin 45,000

Carbonic anhydrase 31 ,000

Soybean Trypsin Inhibitor 21 ,000

Lysozyme 14,000

or by chromatography on a Superose 12 column (Pharmacia), using the following molecular weight standards:

which polypeptide composition of matter has the ability to promote the differentiation of O-2A progenitor cells into type-2 astrocytes when applied to O-2A progenitors growing in serum-free chemically-defined medium [DMEM-BS, as defined in Section (A)], which promotion of astroctyic differentiation is more rapid when the factor of the invention is applied to O-2A progenitors in the additional presence of CNTF or LIF (leukemia inhibitory factor; Gough et al., 1988; Proc. Natl. Acad. Sci. USA 85:2623; Moreau et al., 1988, Nature 336: 690), which together are members of a broad biological-family of modulators of differentiation; (Bazan 1990, Immunol. Today, 1 1 :350; Bazan, 1990, Proc. Natl. Acad. Sci. U.S.A. 87:6934; Bazan, 1991 , Neuron 7: 1991 ; Patterson, 1992, Curr. Opinion Neurobiol. 2:94).

It will be appreciated that the molecular weight of 29-34 kD is not exact but is subject to variation depending, inter alia, upon the conditions under which the sample is prepared and analyzed. A variation of, say, about 10% would not, for example, be impossible for material prepared differently or analyzed differently. Nonethelesss, with the source material used thus far, all measurements to date produce molecular weight values of 31-33 kD for the factor of the invention.

This polypeptide composition of matter is distinct from the extracellular matrix material described by Lillien et al. (Lillien, L. E., Sendtner, M., and Raff, M. C. (1990) J. Cell Biol. 111 , 635-644) in several respects: First the activity is not purified from extracellular matrix but is instead collected as a secreted protein. Second, the activity is not produced by cultures of rat meningeal cells. Third, the activity acts by itself to promote astrocytic differentiation, although the rapidity of differentiation is enhanced by the co-application of CNTF. Fourth, this factor has the further effects of altering the effects of CNTF application from that of promoting oligodendrocyte generation, maturation and survival to that of promoting astrocytic differentiation.

The factor of the invention can be purified from medium conditioned by bovine aortic endothelial cells as a secreted protein. This protein is basic, and of apparent molecular weight of 29-34 kD when analyzed by SDS gel electrophoresis or when analyzed by Superose 12 column chromatography. The protein is stable over a pH range of 2.5-9.5 and can be stored in a solution of 0.1 % trifluoracetic acid (TFA): 60% acetonitrile for at least 6 months without apparent loss of activity. Activity is also heat resistant and is not destroyed by heating to 90°C for 5 minutes or by extensive purification at room temperature. Activity is however eliminated by exposure to reducing agents. The invention also provides a process for the purification of this factor, and methods for defining the activity of this and other factors of similar activity. Therapeutic application of the factor of the invention in the prophylaxis or treatment of any pathophysiological condition in which a factor-sensitive or factor-responsive cell type is involved is a further significant aspect of the invention.

The invention includes any modifications of the endothelial-cell derived factor which do not exhibit a significantly reduced activity. For example, modifications in which amino acid content or sequence is altered without substantially adversely affecting activity are included. By way of illustration, in EP-A-109748. muteins of native proteins are disclosed in which the possibility of unwanted S-S bonding is avoided by replacing any cysteine in the native sequence not necessary for biological activity with a neutral amino acid. Other modifications allowing production of a biologically active material are also envisaged, with the critical limitation being that the modified polypeptide composition of matter retains as a minimum characteristic the structural features which enable binding of the factor of the invention to cells sensitive to the activity of the factor. The statements of effect and use contained herein are therefore to be construed accordingly, with such uses and effects employing modified or equivalent factors as aforesaid being part of the invention.

None of the factors described in the art has the combination of characteristics possessed by the new polypeptide factor.

A further aspect of the invention is a process for the preparation of a polypeptide composition of matter as defined above comprising collecting medium conditioned by endothelial cells to obtain protein, subjecting the resulting material to chromatographic purification comprising concentration by ultrafiltration followed by sequential purification using DEAE-cellulose (or any reasonable alternative), Mono-Q-FPLC (or any reasonable alternative) SDS-polyacrylamide gel electrophoresis and/or Reverse Phase Chromatography and collecting a fraction which has an apparent molecular weight of 29-34 kD when analyzed by SDS-polyacrylamide gel electrophoresis using the following molecular weight standards:

Phosphorylase b 97,400

Bovine serum albumin 60,000

Ovalbumin 45,000

Carbonic anhydrase 31 ,000

Soybean Trypsin Inhibitor 21 ,000

Lysozyme 14,000

which polypeptide composition of matter has the ability to promote the differentiation of O-2A progenitor cells into type-2 astrocytes when applied to O-2A progenitors growing in serum-free chemically-defined medium [DMEM-BS, as defined in Section (A)], which promotion of astroctyic differentiation is more rapid when the factor of the invention is applied to O-2A progenitors in the additional presence of CNTF or LIF (leukemia inhibitory factor; Gough et al., 1988; Proc. Natl. Acad. Sci. USA 85:2623; Moreau et al., 1988, Nature 336: 690), which together are members of a broad biological family of modulators of differentiation; (Bazan 1990, Immunol. Today, 1 1 :350; Bazan, 1990, Proc. Natl. Acad. Sci. U.S.A. 87:6934; Bazan, 1991 , Neuron 7: 1991 ).

Preferably, the above process starts by collecting medium conditioned by endothelial cells. This conditioned medium may, for convenience, be concentrated by ultrafiltration and buffer exchange. It is also preferred that isolation of the described factor begins by isolating a relevant fraction obtained by DEAE-cellulose chromatography (or any reasonable alternative) of endothelial cell-conditioned medium. Alternatively, conditioned medium can be passed directly onto a Q-Sepharose column. It is further preferred after DEAE-cellulose chromatography (or any reasonable alternative), that mono-Q-FPLC (or any reasonable alternative) and/or reversed phase chromatography be sequentially employed prior to SDS-gel electrophoresis or Superose 12 chromatography. Alternatively, it is possible to use SDS-gel electrophoresis, followed by removal of SDS by standard techniques, before proceeding to reversed phase chromatography with the active material obtained from SDS-gel electorphoresis. At each stage in the process, the fractions containing the enriched biological activity may be determined using induction of astrocytic differentiation in O-2A progenitor cells, preferably derived from appropriate tissues of the rat, as a measure in an assay in which O-2A progenitor cells present in heterogeneous cultures prepared from optic nerve or other regions of the CNS, are exposed to the factor of interest for lengths of time between 1 day to 8 days after which cells are labeled with antibodies which recognize GFAP. Alternatively, activity may be assayed on purified populations of O-2A progenitor cells. All assays are carried out in serum-free chemically-defined medium (DMEM-BS, for example).

It will be appreciated that the protocol of purification described is not restrictive, and that other protocols leading to purification of the same polypeptide composition of matter can be developed.

Another aspect of the invention is a pharmaceutical or veterinary formulation comprising a factor as defined above formulated for pharmaceutical or veterinary use, respectively, optionally together with an acceptable diluent, carrier or excipient and/or in unit dosage form. In using the factors of the invention, conventional pharmaceutical or veterinary practice may be employed to provide suitable formulations or compositions.

Thus, the formulations of this invention can be applied to parenteral administration, for example, intravenous, subcutaneous, intramuscular, intraorbital, ophthalmic, intraventricular, intracranial, intracapsular, intraspinal, intracisternal, intraperitoneal, topical, intranasal, aerosol, scarification, and also oral, buccal, rectal or vaginal administration.

Methods well known in the art for making formulations are to be found . in, for example, "Remington's Pharmaceutical Sciences'. Formulations for parenteral administration may, for example, contain as excipients sterile water or saline, polyalkylene glycols such as polyethylene glycol, oils of vegetable origin, or hydrogenated naphthalenes. Biocc patible, biodegradable lactide polymer, lactide/glycolide copolymer, or polyoxyethylene- polyoxypropylene copolymers may be used to control the release of the present factors. Other potentially useful parenteral delivery systems for the factors include ethylene-vinyl acetate copolymer particles, osmotic pumps, implantable infusion systems and liposomes. Formulations for inhalation may contain as excipients, for example, lactose, or may be aqueous solutions containing, for example, polyoxyethylene-9- lauryl ether, glycocholate and deoxycholate, or may be oily solutions for adminstration in the form of nasal drops, or as a gel to be applied intranasally. Formulations for parenteral adminstration may also include glycocholate for buccal adminstration, methoxysalicylate for rectal administration, or citric acid for vaginal administration.

The present factor can be used as the sole active agent or can be used in combination with other active ingredients, e.g., other growth or differentiation modulating factors which could be used to promote cell division and/or survival and/or differentiation. Due to the indications (see above) that the factor of invention may interact beneficially with multiple members of a family of biologically active compounds (which themselves are known to effect a broad range of cell types) , no limitation is envisaged on the types of cells whose physiological properties might be beneficially altered by administration of the factor of the invention.

The concentration of the present factor in the formulations of the invention will vary depending upon a number of issues, including the dosage to be administered, and the route of administration.

In general terms, the factor of the invention may be provided in an aqueous physiological buffer solution containing about 0.1 to 10% w/v compound for parenteral administration. General dose ranges are from about 1//g/kg to about 1 g/kg of body weight per day; a preferred dose range is from about 0.01 mg/kg to 100 mg/kg of body weight per day. The preferred dosage to be administered is likely to depend upon the type and extent of progression of the pathophysiological condition being addressed, the overall health of the patient, the make up of the formulation, and the route of administration.

As indicated above, O-2A progenitor cells (the precursors of myelin forming oligodendrocytes of the central nervous system) are stimulated to undergo astrocytic differentiation in the presence of the factor of the invention. These progenitor cells are present throughout the central nervous system both during perinatal development and in adulthood, and are the only precursor cell thus far identified for oligodendrocytes. The appropriate generation of oligodendrocytes is essential for normal function of the central nervous system, as these cells are involved in creating the myelin sheath around individual nerve fibres which is important for proper conduction of electrical impulses along individual axons. The failure to generate oligodendrocytes properly is associated with severe clinical problems (as exampled by Canavan's disease or other leukodystrophyies with associated myelinopathies) and the failure to repair demyelinated damage is also associated with severe clinical problems (as exampled by the clinical deficits associated with myelin breakdown and failure of effective repair in multiple sclerosis patients).

-O-2A progenitors represent one of a small number of multipotential precursors cells whose differentiation is sufficiently controllable in vitro to allow these cells to be used as assay systems for the purification of novel biological activities capable of modulating cellular differentiation. However, it is unlikely for several reasons that these are the only cells whose differentiation will be effected by the factor of this invention. First, the activity examined is secreted by all endothelial populations examined thus far, and thus secretion of this activity appears to be a fundamental aspect of endothelial cell biology. As endothelial cells are present in all tissues of the body, starting with early stages of embryogenesis, it is likely that a differentiation-modulating factor secreted by these cells will have effects on a multitude of cell types. Second, the activity of the invention has been purified from medium conditioned by aortic endothelial cells, and this material is effective on 0-2A progenitor cells of the central nervous system. It is very unlikely that the normal target of action of the factor secreted by aortic endothelial cells is a specialized precursor cell from the central nervous system. Third, the activity of the invention is purified from medium conditioned by bovine aortic endothelial cells by virtue of its activity on rodent 0-2A progenitor cells; thus, this activity is readily effective across species barriers, indicating conservation of function across species. Such conservation of function is generally characteristic of factors with a wide range of biological effects. Fourth, 0- 2A progenitor cells are not the only cells responsive to the effects of CNTF (for example, see Patterson, 1992, Curr. Opinion Biology, 2:94-97) and the ability of the factor of the invention to modify the effects of CNTF on 0-2A progenitors may also apply to other cells in which differentiation is modulated by application of CNTF. Fifth, CNTF is itself a member of a broad family of molecules with profound effects on differentiation, and the factor of the invention may also modulate the effects of other members of this family. This family thus far includes LIF, growth hormone, prolactin, oncostatin M> myelomonocytic growth factor, interleukin 6, interferons α and β, interleukin 5, interleukin 7, interleukin 10, erythropoietin and granulocyte-colony stimulating factor (G-CSF) (Bazan 1990, Immunol. Today, 1 1 :350; Bazan 1990, Proc. Natl. Acad. Sci. U.S.A. 87:6934; Bazan, 1991 , Neuron 7:197). In addition, it is well known that factors which have the ability to both have direct effects and to modify the effects of other factors are able to influence the activity of broad ranges of factors. Finally, the factor of interest also enhances the killing of cells by tumor necrosis factor, thus indicating further the breadth of activity of this biological material and demonstrating that its range of activity extends to synergistic interactions with other factors besides those contained within the structural family which includes CNTF, LIF et al.

The invention envisages application of the factor of the invention in situations where benefit would be gained by modifying the activity of other factors used to promote the division, differentiation or survival of particular cell types. For example, it has been demonstrated that application of CNTF can prevent the in vivo death of motor neurons damaged by axotomy (Sendtner et al.. Nature 345:440); it is thus envisaged that application of the factor of the invention could be used to enhance the activity of CNTF in such applications. It has also been demonstated that application of LIF (a member of the CNTF family which also shows cooperative interactions with the factor of this invention in promoting astrocytic differentiation of O-2A progenitors in vitro) can be used to promote the differentiation of lymphoid tumours (Metcalf, 1990, Phil Trans. Roy. Soc. Lond. B 327:99-109); it is envisaged that application of the factor of the invention could be used in vivo to enhance the efficacy of LIF in promotion of lymphoid tumour cell differentiation in vivo and may be more generally useful in helping to promote the differentiation of other tumours. Due to the structural similarities between CNTF, LIF and, inter alia, oncostatin M, interleukin-6 and granulocyte-colony stimulating factor, it is further envisaged that application of the factor of the invention could be used in vivo to enhance the efficacy of any member of this family of differentiation-modulating factors in promotion of lymphoid tumour cell differentiation in vivo and may be more generally useful in helping to promote the differentiation of other tumours. In addition, application of the factor of the invention, either alone or in combination with other differentiation-inducing agents, may be of general usefulness in promoting the differentiation of glial tumours of the central nervous system, which may be derived from glial precursor cells. A still further example of the possible usefulness of this factor would be to augment the activity of compounds such as Granulocyte-Macrophage colony stimulating factor (GM-CSF) or Granulocyte Colony Stimulating Factor (G-CSF) when applied in vivo for the treatment of syndromes involving an insufficiency of lymphoid cells, (e.g., Moore, 1991 , Blood 78:1 ; Moore, 1991 , Ann. Rev. Immunol. 9: 159-191 ; Waddick et al., 1991 , Blood 77:2364-2371 ). In addition, application of the factor may be used in situations where it is desirable to enhance the activity of TNF.

The invention also specifically includes a method for the prevention of glial scar formation in the central nervous system, which comprises administering an effective amount of a substance which inhibits the binding of a factor as defined above to a receptor therefor. One of the possible conditions in which astrocytic differentiation of O-2A progenitors in vivo would, specifically be undesirable would be in association with lesions in the central nervous system. It has been shown that lesions in the white matter tracts (the specific site of residence of O-2A progenitors) of the central nervous system are associated with a greater production of chondroitin-sulphate proteoglycan (a component of glial scar tissue in the central nervous system) than lesions in the grey matter tracts (McKeon et al., 1991 , J. Neurosci. 1 1 :3398-341 1 ), and it has also been shown in vitro that chondroitin-sulphate proteoglycan is preferentially expressed by type-2 astrocytes (the astrocytic progeny of O-2A progenitors) as compared with type-1 astrocytes (Gallo and Bertolucci, 1990, Exp. Cell Res., 187:211-223), thus indicating that glial scars in white matter tracts may contain type-2 astrocytes. O-2A progenitors are also present in the adult CNS, and these cells have thus far been found to be responsive to the same physiological stimuli which are active on O-2A progenitors derived from the perinatal CNS (Wolswijk, G. and M. Noble, Development, 105 (1989) 387-400; Wolswijk, G., P. Riddle and M. Noble, Development, 109 (1990) 691-698; Wolswijk, G., P. Riddle and M. Noble, Glia, 4 (1991 ) 495-503; Wren, D., G. Wolswijk and M. Noble, J. Cell Biol., 1 16 (1992) 167-176.). The possible contribution of type-2 astrocytes to glial scars in the adult CNS, where they would be expected to be derived from O-2A progenitors of the adult CNS, is of particular interest in light of observations that oligodendrocytes are eventually lost from the lesions found in brains and spinal cords of multiple sclerosis patients, and that these lesions come to be composed of astrocytic scars. Thus, application of an antagonist of the factor of the invention is included in the invention as a means of preventing glial scar formation following injury in the CNS either as result or traumatic injury or a due to a disease process. Application of an antagonist of the factor could also be used in situations, such as cachexia, septic shock, AIDS, multiple sclerosis, acute injury or stroke (as non-limiting examples) where it would be desirable to reduce the activity of TNF in causing tissue damage.

In general, the invention includes the use of the present polypeptide factor in the prophylaxis or treatment of any pathophysiological condition in which a factor-sensitive or factor-responsive cell type is involved.

The polypeptide factor of the invention can also be used as a immunogen for making antibodies such as monoclonal antibodie following standard techniques. These antibodies can, in turn, be use for diagnostic purposes. Thus, conditions perhaps associated wit abnormal levels of the factor may be tracked by using such antibodies In vitro techniques can be used, employing assays on isolated sample using standard methods. Imaging methods can also be employed i which the antibodies are, for example, tagged with radioactive isotope which can be imaged outside the body using techniques employed in th art of, for example, tumour imaging.

The invention also includes the general use of the present facto as a modulator of cell differentiation in vivo or in vitro, and the factor fo such use. One specific embodiment is thus a method for producing o enhancing a differentiation event in a vertebrate by administering a effective amount of the factor of the invention. A preferred embodimen is such a method in the treatment or prophylaxis of a pathophysiologic condition where it would be advantageous to promote the differentiatio of a cell type responsive to the factor of the invention.

A further general aspect of the invention is the use of the facto of the invention in the manufacture of a medicament, preferably for th prophylaxis or treatment of a a pathophysiological condition where i would be advantageous to promote the differentiation of a cell typ responsive to the factor of the invention.

Also included in the invention is the use of the factor of th invention in competitive assays to identify or quantify molecules havin receptor binding characteristics corresponding to those of said polypeptides. The polypeptides may be labelled, optionally with a radioisotope. A competitive assay can identify both antagonists and agonists of the relevant receptor.

In another aspect, the invention provides the use of the factor of the invention in an affinity isolation process, optionally affinity chromatography, for the separation of a respective corresponding receptor. Such processes for the isolation of receptors corresponding to particular proteins are known in the art, and a number of techniques are available and can be applied to the factor of the present invention. For example, in relation to IL-6 and IFN-gamma, the reader is referred to Novick, D. et al., J. Chromatogr.; 1990, June 27; 510. 331-7, in relation to gonadotropin releasing hormone reference is made to Hazum, E. , J. Chromatogr.; 1990, June 27; 510, 233-8, in relation to vasoactive intestinal peptide reference is made to Couvineau, A. et al., J. Biol. Che .; 1990, Aug 5; 265(22) , 13386-90, in relation to IL-2 reference is made to Smart, J.E. et al. , J. Invest. Der atol.; 1990, June: .94 (6 Suppl.), 158S-163S, and in relation to human IFN-gamma reference is made to Stefanos, S., et al., J. Interferon Res. ; 1989, Dec, £(6), 719-730.

In the following description, techniques for obtaining and characterizing the endothelial-cell derived protein are described as an Example of the invention. In the accompanying drawings. Figures 1-4, 10 A and B describe the biological activity of the factor of the invention. Figures 11 and 12 demonstrate that application of CNTF in the absence of the factor of the invention promotes oligodendrocytic differentiation and Figures 5-9 are purification activity profiles obtained as described within the Example. Figure 1 demonstrates that endothelial cells from differ sources secrete biological activity which promotes astroc differentiation of O-2A progenitors in heterogeneous cultures of o nerve cells derived from 7 day old rats.

In these experiments, postnatal day 7 optic nerve cells w cultured on coverslips (7500 cells/coverslip) for 3 days in serum-f DMEM-BS medium ( -d~), and in a 30% solution of DMEM-BS p conditioned media from either endothelial cells of the bovine aorta (°) from bovine adrenal capillary (-'❖ , prepared as described in (F) bel The cells were double labelled with NSP4 and anti-GFAP antibodies, described in (I) below. The proportion of NSP4⁺ process-bearing c that were GFAP⁺ was determined. Each point is the mean +_ S.E.M. three experiments in triplicate and at least 200 cells were scored culture.

Figure 2 demonstrates that activity which promotes astroc differentiation is secreted only by the endothelial cell colonies deri by isolation of rat brain capillaries.

In these experiments, postnatal day 7 optic nerve cells w cultured on coverslips (7500 cells/coverslip) for 3 days in DMEM- medium, plus conditioned media (30%) from cloned cortical capill endothelial cells (•■) and a heterogenous population of non-endoth cells *=*) isolated in a similar manner (as described in Section E). O- lineage cells were double labelled with NSP4 and anti-GFAP antibodi and the proportion of NSP4 process-bearing cells that were GFAP⁺ determined. Each bar is the mean +_ S.E.M. of at least th experiments in triplicate, with at least 200 cells scored per culture. Figure 3 demonstrates that biological activity promoting astrocyti differentiation of O-2A progenitors is riot secreted by rat skin fibroblasts meningeal cells or purified cortical astrocytes.

In these experiments, postnatal day 7 optic nerve cells wer cultured on coverslips (7500 cells/coverslip) for 3 days in DMEM-B medium (-*^a"), plus conditioned media (30%) from various cell types bovine capillary endothelial cells (-c⁵"), rat meningeal cells (-*-), rat ski fibroblasts (-X-), purified cortical rat brain type-1 astrocytes ( ^"-*^" ), wit all conditioned media prepared as described in Sections (F, G and H). Th cells were double labelled with NSP4 and anti-GFAP antibodies, and th proportion of NSP4 process-bearing cells that were GFAP⁺ wa determined. Each point is the mean _+ S.E.M. of three experiments i triplicate and at least 200 cells were scored per coverslip.

Figure 4 demonstrates that the induction of astrocyti differentiation of O-2A progenitors contained within heterogeneou cultures of optic nerve cells requires that cultures be derived fro animals of an appropriate age and that the cell density in the cultur must be above a certain threshold.

In these experiments, optic nerve cells were cultured o coverslips at various densities in DMEM-BS containing 30% EndoCM After 3 days, the cells were stained and counted. EndoCM did not induc astrocytic differentiation of O-2A^'progenitor cells in cultures derived fro one day old animals which cells were grown for 3 days at a platin density of 1-10,000 cells/coverslip. In contrast, in cultures derived fro 7 day old rats and grown at densities >4000 cells/coverslip in th presence of EndoCM, 87% of the O-2A lineage cells differentiated int type-2 astrocytes within 3 days. However, induction of astrocyti differentiation at the 3-day time point was not as effective in cultu plated at densities of < 2000 cells/coverslips. Each point represents mean _+. S.E.M. of three experiments in triplicate and at least 200 c were scored per coverslip.

Figure 5 is the profile for product from DEAE-cellul chromatography.

In these experiments, 2.5 I of concentrated, protease inhibi treated, bovine aortic endothelial cell conditioned media was applied a column of DEAE-cellulose, which was washed until the absorbance the eluate, monitored at 280 nm, fell to zero. The figure shows elution profile of protein concentration when the column was develo with a gradient of 0-0.5M NaCl. The solid horizontal bar represents fractions containing the GFAP inducing activity, which were pooled the next purification step.

Figure 6 is the profile for product from Mono-Q chromatograp

In these experiments, fractions from the DEAE-cellulose acti were dialysed, freeze-dried, reconstituted and applied onto a Mon column. The column was washed until the absorbance of the elua monitored at 280 nm, fell to almost zero.The figure shows the elut profile of protein concentration, when the column was developed a gradient of 0-1 M NaCl. The solid horizontal bar represents the fracti containing the GFAP inducing activity, which were pooled for the n purification step.

Figure 7 is the profile for product from reverse-ph chromatography on a ProRPC column.

The active fractions from the Mono-Q column (with 0.1 % TFA were injected onto a Pro-RPC reverse phase column. All the biologic activity bound to the column, which was washed and developed with gradient of 0-90% acetonitrile (0.1 % TFA). The active fraction wa eluted at an acetonitrile concentration of approximately 35%. The yiel from the Pro-RPC column was approximately 60% with a purification o 60 fold over the Mono-Q column.

Figure 8 is the profile for product from FPLC-size exclusio chromatography on a Superose 12 column.

In these experiments, EndoCM/ProR containing 290 μg tot protein was seperated into 0.2 ml fractions and 25 μ\ of each was adde to cultures of 7 day old optic nerve cells grown for 3 days in DMEM-BS The solid bar indicates the fraction of highest GFAP inducing activity The relative molecular weight M_r, was calculated as 33 kd. Arrow indicate relative mobility of molecular weight standards: thyroglobuli (670 kd), gamma globulin (158 kd), ovalbumin (44 kd), myoglobin (1 kd) and Vitamin B12 (1.35 kd). The solid vertical bars incdicat fractions expressing astrocyte-differentiation inducing activity.

Figure 9 is the profile for product from SDS-gel electrophoresi

The active fraction from the Pro-RPC column was checked f purity by running on SDS-PAGE followed by silver staining. A parallel g was run to determine the position, and hence the apparent relativ molecular weight (M_r) of the inducing activity. Proteins were eluted fro sections of the gel and bioassayed for GFAP-inducing activity. The M_r o the inducing activity from the gel was ^"calculated as approximately 31 kd. The arrow indicates the location of the active protein band and th asterisks denote the lane in which biological activity was present in th assays for astrocyte-inducing activity.

Figures 10 A and B demonstrate that the factor of the inventio works together with CNTF to promote astrocytic differentiation of O-2 progenitor cells but was also able, over a slightly longer time period, t induce astrocytic differentiation of these cells when applied in th absence of CNTF.

In these experiments, O-2A progenitors from P7 rat optic nerv were purified using the panning technique [as described in Section (B)] 2500 cells/coverslip were grown in DMEM-BS for one to three days i the presence of EndoCM (1//g/ml), CNTF (200pg ml^"1) and EndoCM/Pro + CNTF. Cells cultured with CNTF alone all differentiated int oligodendrocytes, as was the case for control cells grown in DMEM-BS In contrast, cells grown for 1 to 3 days in the presence of EndoCM/Pro alone did not differentiate into oligodendrocytes, but continued t express the A2B5⁺ GFAP^' GalC^" antigenic phenotype [for definition o antibodies, see Section (I)], characteristic of O-2A progenitors. Exposur to the combination of EndoCM/ProR and CNTF promoted the astrocyti differentiation of almost 90% of the O-2A progenitors into type- astrocytes within 3 days of in vitro growth (Figure 10A, in which eac bar represents the calculated average _+. S.E.M. of the various cell type on 15 coverslips from five individual experiments in which ail the cell were counted). However, growth of purified O-2A progenitors in th presence of EndoCM/ProR in the absence of CNTF for 6 days wa associated with astrocytic differentiation of 50% of the O-2 progenitors and growth in these conditions for 8 days was associated with astrocytic differentiation of virtually all O-2A progenitors (Figure 10B).

Figure 1 1 . CNTF applied to O-2A progenitors in the absence of the factor of the invention promotes oligodendrocytic differentiation. I. Almost 90% of the O-2A progenitors in cultures of cells purified by antibody-mediated cell capture are inhibited from differentiating into oligodendrocytes by growth in the presence of basic fibroblast growth factor (bFGF) for 3 days. In contrast, the additional presence of CNTF in the culture medium causes a 3-fold increase in the number of cells which become oligodendrocytes.

Figure 12. CNTF applied to O-2A progenitors in the absence of the factor of the invention promotes oligodendrocytic differentiation. II. The ability of CNTF to promote oligodendrocyte differentiation is further indicated by the increase in the numbers of oligodendrocytes which express myelin basic protein when culture medium contains CNTF. This increase is seen for cultures grown in DMEM-BS, and a similar degree of difference is even seen in cultures for which bFGF was included in the tissue culture medium.

EXAMPLES

Methods

(A) Preparation of optic nerve cultures

O-2A lineage cells that were used for the initial GFAP induction studies and in factor purification assays were isolated from optic nerves of 7-day-old Sprague-Dawley rats, and were dissociated into single cell suspensions using collagenase, trypsin and EDTA as described previousl (Raff, M. C, Miller, R.H. and Noble, M. (1983) Nature 303, 390-396; Noble, M. and Murray, K. (1984) EMBO J. 3, 2243-2247). Approximately 7500 cells were plated on 13 mm diameter glas r-^verslips coated with poly-L-lysine (PLL; Sigma; M_r 175,000; 20 g ml^"1) in 0.5 ml Dulbecco's modified Eagle's medium (DMEM) supplemente with 25 μg ml^"1 gentamicin, 2 mM-glutamine, 0.234 I.U.ml^"1 bovin pancreas insulin (Sigma), 100 g ml^"1 human transferrin (Sigma), 0.0286% (v/v) bovine serum albumin-pathocyte (Miles Laboratories, Inc) 0.2 μM-progesterone (Sigma), 0.10 μM-putrescine (Sigma), 0.45//M L-thyroxine (Sigma), 0.0224 μM-selenium (Sigma) and 0.49 /M 3,3',5-triiodo-L-thyroxine (Sigma); modified from Bottenstein and Sato (Bottenstein, J. E. and Sato, G. H. (1979) Proc. Natl. Acad. Sci. US 76, 514-517). This medium formulation is abbreviated as DMEM-BS.

(B) Purification of O-2A progenitor cells

The specific antibody phenotype of perinatal O-2A progenitor (A2B5⁺/GalC) allows purification of these cells from the whole cel population by using a specific antibody-capture assay (Harlow, E.an Lane, D. (1988) Antibodies. Cold Spring Harb. Labs. U.S.A.). This assa was adapted to the O-2A lineage by using a negative selection with th Ran-2 antibody (Bartlett, P. F., Noble, M., Pruss, R. M., Raff, M. C Rattray, S. and Williams, C. A. (1981 ) Brain Res. 204, 339-351 ) t eliminate type-1 astrocytes, followed by anti-GalC antibody treatment t remove oligodendrocytes (Ranscht, B., Clapshaw, P. A., Price, J., Noble M. and Seifert, W. (1982) Proc. Natl. Acad. Sci. USA 79, 2709-2713) The remaining cell suspension was plated on an A2B5 antibod (Eisenbarth, G. S., Walsh, F. S. and Nirenberg, M. (1979) Proc. Natl Acad. Sci. USA 76, 4913-4917) coated dish, to allow binding to th plate of all A2B5⁺ cells (which all appear to be 0-2A progenitors in opti nerve cultures prepared from postnatal rats; Raff, M. C, Miller, R.H. and Noble, M. (1983) Nature 303, 390-396). The supernatant was removed and the plate was washed with DMEM-BS. The bound cells were trypsinized and re-plated into a 24 well plate in a 25 μ\ drop of DMEM-BS (2500 cells/well). After the cells had attached to the coverslip (30min-1 h), 300 μ\ of DMEM-BS were added. This procedure yields 2x10⁵ O-2A progenitor cells from an initial 2x10^β mixed cells from rat optic nerve. In the final culture the number of contaminating A2B5^" cells (i.e.,non-O-2A lineage cells) represented <0.5% of the total cells. The specific protocol for application of this method is as follows:

1. Cell suspension

dissect optic nerve from 7 day old rats, cut in pieces and incubate

1 hr in Collagenase (1.33% = 2000U/ml) spin 5min at lOOOrpm, 1 min at 3000rpm add papain solution incubate 60 min at 37 ^*C pour contents into 15ml blue-top tube allow pieces to settle, discard supernata t add 1 ml papain inhibitor solution titurate with 23x 1 and 27x1/2 gauge needles

2. panning

2.1. preparation of panning dishes incubate 100mm Petri dishes with antilgG for RAN2/GalC or antilgM for A2B5 in 50mM Tris (pH9.5) for 24hrs at 4° C wash off solution 3x with MEM add 9ml primary AB solution, block with 1 ml 1 %BSA incubate for 60' at 37 ^βC wash 3x, leave MEM 2.2 panning procedure spin down cells and resuspend in 7ml L15 place cells on 100mm Ran-2 dish for 20' transfer cells to 100mm GalC dish for 20' transfer cells to 100mm A2B5 dish for 30' remove cells from dish with trypsin: dilute 2ml trypsin in 2ml MEM add trypsin solution to dish and incubate 10' in 37 ^*C, then wash cells off by flushing with medium add 2ml trypsin inhibitor solution spin cells 5min at 1000rpm/1 min at 3000rpm resuspend cell pellet in DMEM-BS and plate at appropiate density on coverslips

activate with 1 mg L-cysteine /20ul 1 N NaOH filter (0.22μm) and add immediately

Papain inhibitor solution

15mg ovomucoid

8mg BSA

100ul 0.4%Dnase in total of 10ml L15 adjust solution to pH 7.4 with 1 N NaOH aliquot and freeze

(C) Preparation of collagen coated dishes

Falcon 96 well tissue culture dishes were flooded with collagen (Vitrogen, Flow Labs) and allowed to stand at 37°C overnight. One hour prior to use, the collagen was aspirated and the dishes were allowed to dry in the tissue culture hood. The dishes were then exposed to ammonia vapour (35% ammonium solution in a closed box) for 5 min. The dishes were allowed to stand for 15 min to allow the ammonia vapour to dissipate, after which the dishes were washed twice with medium to remove traces of ammonia and to hydrate the collagen pricr to plating of cells.

(D) Preparation of gelatin coated flasks

Tissue culture flasks (Nuπc 25 or 80 cm²' were flooded with 2% (w/v) gelatin (Difco) made up in sterile tissue grade water. The flasks were allowed to stand at 37°C overnight. Just prior to, use, the gelatin was aspirated and the flasks were washed with medium.

(E) Preparation of rat brain capillary endothelial cell cultures

Rat brain capillary endothelial cells were prepared by modifications of the method of Hughes, C. C. W., Male, D. K. and Lantos, P. L. [(1988) Immunology 64, 677-681 )]. Four to six adult Sprague-Dawley rats (150 g), were decapitated under CO₂coma. Brains were removed, washed in Leibowitz L-15 medium (supplemented with 25 mg ml^"1 gentamicin) and then placed into a 30 mm petri dish containing a few mis of L-15. The cerebellum, corpus callosum and the optic bulb were dissected and the meningeal sheath was removed. The remaining grey matter was chopped finely with a sterile scapel blade and then forced through a 19 gauge needle once and incubated in 15 mi of 0.1 % collagenase:dispase (Boehringer Mannheim) in L-15 for 60 min at 37°C. The tissue was centrifuged at 1000g for 10 min at 4°C and the supernatant was discarded. To the pellet, 20 ml of 25% bovine serum albumin (BSA) in L-15 was added and mixed thoroughly, but withou frothing. Tissue was then spun at 2000g for 10 min. The floating laye of myelinated tissue together with the supernatant was removed with care, avoiding disturbance of the small pellet. The supernatant and the tissue were mixed and spun again at 2000g for 20 min. The tissue laye and supernatant were discarded and the two pellets were pooled and suspended in 10ml 0.5% BSA in L-15 and spun at 1000g for 10 min a 4°C, to wash the pellets. The pellets were resuspended in 0.1 % collagenase:dispase (in L-15) and incubated at 37°C for two hours. Afte incubation, DNAse was added to a final concentration of 10 mg ml^"1 fo 10 min and the suspension was spun at 1000g for10 min at 4°C. Th supernatant was discarded and the pellet washed with 5 ml Ca^{+ +}Mg⁺ free DMEM by gentle mixing, followed by centrifugation at 1000g for1 min at 4°C. The pellet was again suspended gently in 1 ml Ca^{+ +} Mg⁺ free DMEM and layered onto a 10 ml Percoll gradient and spun at 1000 for 10 min at 4°C. [A linear gradient of 50% Percoll (Pharmacia) i Ca^{+ +}Mg^{+ +} free phosphate buffered saline (PBS) was prepared i advance by mixing 5 parts isotonic Percoll {9 parts Percoll with 1 par 10X Ca^{+ +}Mg^{+ +} free PBS} with 5 parts 1 X PBS and centrifuged a 26,000g for one hr.]. The top half of the tube contained cellular debri and single cells. The bottom half contained red blood cells, seen as a re ring, and just above this ring were the intact capillaries. This layer wa removed carefully and suspended in 15 ml L-15 and spun at 1000g fo 20 min at 4°C. The supernatant was discarded and the capillarie suspended gently in growth media [DMEM 4.5 g L^"1 glucos supplemented with 20 mM glutamine, 20% plasma derived serum a described in(Vogel, A., Raines, E., Kariya, B., Rivest, M. J. and Ross, R. (1978) Proc. Natl. Acad. Sci. 75, 2810-2814) 10 I.U.ml^*1 Heparin (Sigma), 5 ng ml^"1 recombinant-bFGF (Boehringer-Mannheim), 0.5 μg ml"¹ ascorbic acid (Sigma), 25 μg ml^"1 gentamicin (Flow laboratories)]. The capillaries were plated onto Vitrogen coated 96 well plates at 50% well occupancy (approximately 48 capillary fragments per plate at 100μl of medium per well) and incubated at 37°C in 7.5% CO₂. The medium was changed after three days and every two days thereafter. Wells with single capillaries were tagged at day three and followed to confluence. Endothelial cells arising from these capillaries grew as colonies with tight boundaries. Wells which gave rise to colonies from single cells and which did not express the characteristic cobble-stone morphology of endothelial cells were considered as non-endothelial. Only cells which were considered by morphology to be endothelial cells (in comparison to capillary endothelial cells from rat brain (Hughes, C. C. W., Male, D. . and Lantos, P. L. (1988) Immunology 64, 677-681 ) labelled weakly with rabbit antisera directed against von Willebrand factor (Dako-Patts, 1 :1000). Once colonies were well-established within individual wells, serum-containing medium was washed off with two washes of DMEM and then and replaced with 50//I DMEM-BS. Medium was allowed to condition for 2 days and was then collected and applied to cultures of 7 day old rat optic nerve cells at a 1 :10 dilution. Cultures were allowed to grow for 3 days, after which time they were labeled with antibodies to examine astrocytic differentiation of the O-2A progenitors.

(F) Growth of bovine endothelial cell cultures

Bovine adrenal capillary and bovine aortic endothelial cells were both obtained from J. Folkman (Children's Hospital, Boston). Fcr small scale cultures, bovine adrenal capillary endothelial cells were grown on gelatin coated flasks (25 cm²¹ in 20% donor calf serum (DCS) in DMEM supplemented with 2mM glutamine and 5 ng ml^'1 bFGF. Bovin aortic endothelial cells were cultivated on either gelatin or PLL coate tissue culture flasks (80 cm²¹ in the presence of 20% DCS in DME supplemented with glutamine alone. To prepare conditioned media fro these cells the serum was washed out gradually (50% dilution wit DMEM-BS on a daily basis for a week) and then cells were washe thoroughly with DMEM-BS to remove all residual serum traces. The cell were then grown in DMEM-BS for 24 h and washed again in DMEM-B and incubated for a further 48 h with DMEM-BS. For large scal production of conditioned media for the purification procedure a syste of microcarrier beads in stirring flasks was used (see below).

(G) Preparation of purified cortical astrocytes and conditione medium:

Medium conditioned by purifed cortical astrocytes wa prepared by modifications of previously described methods (Noble, and Murray, . (1984) EMBO J. 3, 2243-2247). Cortices dissected fro 1-2 day old Wistar rat pups were minced and incubated at 37°C in 0.2 ml 2,000 I.U. ml^"1 collagenase (Worthington Biochemical Corp., USA) i L-15, 2.5 ml 30,000 I.U. ml^"1 bovine pancreas trypsin type III (Sigma) i DMEM- CMF (Ca^{+ +}Mg^{+ +}-free DMEM) and 5.0 ml sodiu ethylenediaminetetra-acetic acid (2.5 mg ml^"1; EDTA: Sigma) i DMEM-CMF for 30 min. Enzymatic digestion was stopped by addition 5.0 ml SBTI-DNAse [DMEM supplemented with 5,200 I.U. ml^"1 soybea trypsin inhibitor (Sigma), 74 I.U. ml^'1 bovine pancrease DNAse I (Sigma and 3.0 mg ml^"1 BSA (fraction V; Sigma)]. The suspension was the spun for 2 min at 4,000 rpm. The pellet was resuspended in DMEM 10% FCS followed by trituration of the tissue through a 5 ml blow o pipette and through 25G and 27G hypodermic needles. The c suspension was seeded into PLL coated Nunc 80 cm² tissue cultu flasks at a density of 2 cortices per flask. After 1-2 h at 37°C, the medium was replaced by fresh DMEM + 10% FCS. Cultures were re-fed the following day and on the third and sixth day of culture. After 7 days of in vitro growth, the cultures consisted of a monolayer of flat cells (predominantly type-1 astrocytes and fibroblast-like cells) and a top layer of process-bearing cells (mostly O-2A lineage cells, neurones, and macrophages). Top cells were removed by shaking the culture overnight on a rotating platform (100 revolutions min^"1). The medium was replaced by fresh DMEM + 10% FCS and the cultures grown for a further 24 h before being treated twice with 10 μM cytosine arabinoside (Sigma) for a total period of 4 days. This procedure yielded cultures which consisted of 95-98% GFAP⁺ astrocytes with the antigenic phenotype of type-1 astrocytes (Raff, M. C, Abney, E. A., Cohen, J., Lindsay, R., and Noble, M. (1983b) J. Neurosci 3, 1289-1300; Noble, M., Fok-Seang, J. and Cohen, J. (1984) J. Neurosci 4, 1892-1903 ). Before being used to collect astrocyte-conditioned DMEM-BS, cultures growing in DMEM + 10% FCS were rinsed in DMEM-BS and then grown for 1 day in DMEM-BS. Fresh DMEM-BS was then added, and the medium collected over the next 24 hours was used as astrocyte-conditioned medium (Astro-CM).

(H) Preparation of rat meningeal cells and rat skin fibroblasts conditioned medium

Meningeal sheaths from cortices and fibroblasts from finely chopped rat skin (0.25 cm²) were obtained from 7 day old animals, as described previously (Noble, M., Fok-Seang, J. and Cohen, J. (1984) J. Neurosci 4, 1892-1903 ). The tissues were incubated for one hour in collagenase at 37°C, followed by centrifugation at 1000g for 10 min. The pellet was resuspended in Ca^{+ +}Mg^{+ +}- free DMEM and trypsin at 37°C for 25 min. After addition of DNAse for 5 mins, the tissue was centrifuged at 1000g for 10 min. Cell pellets were resuspended in 10% FCS in DMEM and the tissue was triturated through 25 G and 27 needles. The cells were then washed and grown to confluence i DMEM + 10% FCS supplemented with glutamine and gentamicin Cultures of astrocyte-free meningeal cells were generated by splittin flasks of meningeal cells 1 :5, and growing cells to confluence in fres flasks in DMEM + 10%FCS. Conditioned medium was collected fro these cultures as for cultures of purified cortical astrocytes.

(I) Immunofluorescence analysis

Four antibodies were used to identify the differentiation stat of O-2A lineage cells in the assays. The NSP4 monoclonal antibod (Rougon, G., Hirsh, M. R., Hirn, M., Guenet, J. L. and Gordis, C. (1983 Neurosci. 10, 51 1 -520), (hybridoma supernatant diluted 1 : 1 wit staining medium) reacts with a carbohydrate moiety which, in culture derived from optic nerves of perinatal rats, is expressed specifically b O-2A progenitors and type-2 astrocytes, but which disappears durin oligodendrocytic differentiation (ffrench-Constant, C. and Raff, M (1986) Nature Lond 323, 335-338). This antibody reacts with the sam population of cells that is labeled by the A2B5 monoclonal antibod (Eisenbarth, G. S., Walsh, F. S. and Nirenberg, M. (1979) Proc. Natl Acad. Sci. USA 76, 4913-4917), which alternatively was used as marker in assays using purified populations of O-2A progenitors. Both o these antibodies are IgMs. Monoclonal anti-galactocerebroside (Gal antibody (Ranscht, B., Clapshaw, P. A., Price, J., Noble, M. and Seifer W. (1982) Proc. Natl. Acad. Sci. USA 79, 2709-2713), (hybridom supernatant diluted 1 :10 with staining medium) binds to myelin-specific galactolipid specifically expressed by oligodendrocytes i the central nervous system (Raff, M. C, Mirsky, R., Fields, K. L., Lisa R. P. et al (1978) Nature Lond. 274, 813-816). The anti-GalC antibod is an lgG₃, and can thus be distinguished from A2B5 or NSP4 by use appropriate subclass-specific second antibodies. Rabbit anti-GFAP antiserum (Dako-Patts, 1 :100) is a specific marker of astrocytes (Bignami, A., Eng, L.F., Dahl, D. and Uyeda, C. T. (1972) Brain Res. 43, 429-443) or cells in which astrocytic differentiation has been initiated (Hughes, S.M. and Raff (1987) Development 101 , 157-167). Anti-myelin basic protein antibody (Dako Ltd.) is an antibody to a later appearing marker of oligodendrocyte differentiation. NSP4 and anti-GalC antibodies were used to label living cells, while labeling with anti-GFAP antiserum and anti-MBP antibodies required prior fixation and permeabilization of the cell membrane. Antibodies were diluted in Hanks' balanced salt solution (HBSS; Imperial Laboratories) containing 5% heat-inactivated bovine donor calf serum (Imperial Laboratories) and buffered with 0.02 M-Hepes (Sigma). Live cells were incubated with NSP4 or anti-GalC antibodies for 30 min at room temperature, followed by incubation with rhodamine- or fluorescein-conjugated second antibodies (diluted 1 :100 in HBSS-5% DCS; all antibodies from Southern Biotechnology). After every incubation with antibody, cells were washed several times with HBSS-5% DCS. Following surface labeling, cells were fixed by flooding coverslips with cold (-70°C) methanol and incubating for 10 min at -20°C. After washing off the methanol with HBSS-5% DCS, coverslips were labelled with rabbit anti-GFAP antiserum, followed by sheep anti-rabbit-lg-fluorescein (1 :100, Southern Biotechnology Associates, Inc). After immunolabelling, coverslips were washed, mounted in a drop of glycerol containing 22mM-1 ,4-diazobicyclo[2,2,2] octane (Sigma) to prevent fading (Johnson, G. D., Davidson, R. S., McNamee, K. C, Russell, G. et al. (1982) J. Immunol. Methods, 55: 231 -242), sealed with transparent nail varnish and examined on a Zeiss Universal microscope equipped with phasecontrast, epifluorescence, and rhodamine and fluroescein optics and equipped for photography. At least 200 immunolabelled cells were scored per coverslip in each experiment. (J) Bioassay of chromatographic test fractions

Aliquots of test fractions were bioassayed using postnatal da 7 optic nerve cells cultured on 13 mm glass coverslips coated with PL in 24 well plates, as described above. The cells were grown for 3 day in the presence of the test fractions from the protein purificatio columns. No sample preparation prior to bioassay was required at an stage of the purification protocol. It was found that the 10 mM tris-HC buffer was tolerated well by the optic nerve cells at dilutions abov 1 :20. This concentration was well above that required for bioassay, fo which dilutions were usually in the range of 1 :50 to 1 :250. There wer also no toxicity effects of acetonitrile/TFA factions from th reverse-phase columns on optic nerve cells at dilutions above 1 :50 probably due to rapid vaporisation of the acetonitrile/TFA at 37°C. Thi concentration was well above that required for the bioassay, for whi dilutions were usually >_ 1 :250. The cells were then immunostaine with GFAP as described above.

Due to the very large number of fractions generated by th purification protocol, we devised a simplified means of scoring th promotion of type-2 astrocyte differentiation. The percentage of positiv cells for each condition was estimated using the scoring system give in Table 1. This qualitative method, which was used to localise the pea of activity from the column fractions, was scored independently by tw individuals.

Table 1

+ Above background

+ + 10 -25% of O-2A lineage cells GFAP⁺ + + + approximately 50 % of O-2A lineage cells GFAP⁺ + + + + approximately 75% of 0-2A lineage cells GFAP⁺ + + + + + Maximal activity

(K) Preparation of conditioned medium from endothelial ceils in microcarrier culture.

Bovine aortic endothelial cells were grown on collagen coated microcarrier beads (Cytodex 3, Pharmacia U.K.) in 20% donor calf serum in DMEM supplemented with glutamine. The Techne stirrer system was used to cultivate these cells to confluence. Thereafter the serum levels were reduced to nil by a 50% dilution of the media on a daily basis over a week. The cells were then stirred in DMEM alone for 30 mins to remove residual serum. These cells were then incubated in DMEM alone for 48-72 hours prior to collection of medium for subsequent purifications. To inhibit endogenous proteases the following inhibitors were added to the conditioned medium immediately after collection: Aprotinin (Sigma) 1 mg ml^"1; phenylmethanesulphonyl fluoride (Sigma) 5 mg ml^'1; Benzamadene (Sigma) 1 mg ml^*1 and 1 mM EDTA (Sigma). After collection, the conditioned media was stored at -70°C. After thawing, the endothelial cell-conditioned medium was concentrated 10-fold using a Pellicon Ultrafiltration System equipped with M_r 10,000 cut-off filter package and exchanged into 10 mM Tris-HCI, 1 mM EDTA pH7.8. The large scale purification described in this paper used 2.5 I of concentrated medium which was derived from an initial volume of 22 I.

(L) Ion exchange chromatography:

A test-tube method was first used to determine the ionic-nature of the protein. To two series of 10 X 15ml test-tubes 0.1g of either CM-cellulose or O.lg of DEAE-cellulose (Whatman) were added. The gel in each tube was equilibrated to a different pH by washing 10 times with 10 ml of 0.5M buffer (sodium acetate buffer pH 5.0 and 5.5; phosphate buffer pH 6.0 and 6.5; tris-HCl pH 7.0, 7.5, 8.0 and 8.5). The gel in each tube was equilibrated at a lower ionic strength of lOmM by washing 5 times with 10 ml of buffer of the same pH but lower ionic strength. 200μl of EndoCm made up to 2ml with different buffers at each pH unit was added to each tube and mixed with the gel for 10 min. The tubes were centrifuged at 1000 g for 5 mins. The supernatant from each tube was assayed. As *^" "cribed in Results, these experiments demonstrate binding to £..-..._-_--cellulose.

Two ion-exchange steps were used in the purification of the factor of the invention. The first of these was carried out at 4°C, in contrast to all the other steps which were carried out at room temperature. Concentrated endothelial conditioned media, exchanged into 10 mM Tris-HCl, 1 mM EDTA pH 7.8 buffer, was pumped at a flow rate of 8.3 ml min^" onto a column (15 cm diameter x 20 cm; Pharmacia, Bioprocess Column) of DEAE-cellulose (Whatman) equilibrated in the same buffer. The column was washed with 10 mM tris-HCl, 1 mM EDTA pH 7.8 until the absorbance of the effluent (monitored at 280 nm) fe 1 to zero. The active protein was then eluted with a gradient of 0-0.5 M NaCl in 10 mM tris-HCl, 1 mM EDTA pH 7.8. The flow rate was 7.5 ml min^"1 and the gradient volume was 3.6 1. Fractions of 50 ml were collected.

Active fractions from DEAE ion-exchange chromatography were pooled and dialysed against 20 volumes of 10 mM tris- HCl, 1 mM EDTA pH 7.8 at 4°C. The dialysate was then freeze- dried. This freeze-dried material was taken up in a small volume of 10 mM tris-HCl, 1 mM EDTA pH 7.8. The conductivity was calculated and adjustments

TE SHEET were made to maintain an ionic concentration of less than 50 mM NaCl. This material was then pumped at 3 ml min^"1 onto a Mono Q column HR 10/10 connected to a FPLC system (Pharmacia, U.K.) equilibrated in 10 mM tris-HCl, 1 mM EDTA pH 7.8. The column was then washed with 10mM tris-HCl, 1 mM EDTA pH 7.8 until the absorbance of the effluent fell to zero. The active protein was then eluted with a gradient of 0-1.0 M NaCl in 10 mM tris-HCl, 1 mM EDTA pH 7.8. The flow rate was 2 ml min^"1. Fractions of 6 mis were collected during the loading and wash steps, while fractions of 4 mis were collected as the column developed. A gradient volume of 160 ml was used. Multiple runs were carried out to maintain an optimum protein loading for the column.

(M) Reverse-phase FPLC chromatography:

Reverse-phase chromatography was carried out on a Pro-RPC HR 5/10 column (Pharmacia, UK) connected to a FPLC system (Pharmacia, UK). The column was equlibrated with solvent-A (0.1 % triflouroacetic acid-TFA in water). Active fractions from Mono-Q ion-exchange chromatography were pooled and TFA added to a final concentration of 0.1 %. This material was then pumped at a flow rate of 0.25 ml min^"1 and equilibrated in solvent-A until the absorbance fell to zero. The active fraction was eluted with 0-90% gradient of solvent-B (0.1 % TFA in acetonitrile). The flow rate was 0.25 ml min^"1 with a gradient volume of 40 ml. Fractions of 0.5 ml were collected. Multiple runs were carried out to maintain an optimum protein loading capacity of the column.

(N) SDS-polyacrylamfde gel electrophoresis

• The active protein was recovered from mini-gels by elution into a BSA solution as follows: an aliquot of the active fraction from the ProRPC reverse phase chromatography run was incubated with the sample buffer containing 0.9% SDS for 45 min at 65°C, loaded in 30 μ\ onto a 0.75 mm thick, 12.5 % poiyacrylamide mini-vertical slab gels (BIO-Rad, Mini-PROTEAN II system) and electrophoresed under nonreducin conditions using the buffer system of Laemmli (Laemmli, U. K. (1970) Nature 227, 680). The gel was cut into 2.5 mm slices, each of whic was washed thoroughly in 1 ml solution of 10 mg ml^"1 BSA in PB (Ca^{+ +}Mg^{+ +} free) in 1.5 ml eppendorf tubes (room temperature, 1 hr) using a rotor wheel. The supernatant was aspirated gently and the slic washed for an additional 5 min in 0.5 ml of 0.5 mg ml^'1 BSA in PBS. Th supernatant was discarded and the gel slices were ground individuall using an eppendorf pestle. To these tubes 50 μ\ of 0.22 mm filtered 5 μg ml^"1 BSA in PBS. were added and mixed on the rotor wheel overnigh at 4°C. The tubes were centrifuged at 12,000 rpm for 15 min at 4°C an 35 μ\ of the supernatant was used in the GFAP inducing bioassay Parallel tracks on the gels were silver stained using the method o Ansorge (Laemmli, U. K. (1970) Nature 227, 680), which is capable o detecting less than 1 ng of some protein species.

RESULTS

(O) Promotion of type-2 astrocyte differentiation by endothelial cells.

Conditioned medium derived from either bovine adrena capillary endothelial cells or bovine aortic endothelial cells promote differentiation of O-2A progenitors into type-2 astrocytes (Figure 1 ) When optic nerve cells derived from 7 day old rats were grown i cultures containing 30% conditioned medium: 70% DMEM-BS expression of GFAP was first seen within 24h of culture in 24.8% +_ 5. of NSP4 cells. With continued growth in culture, the proportion of th O-2A lineage cells which differentiated into GFAP⁺ type-2 astrocyte increased still further (Figure 1 ). After 4 days of in vitro growth 75%-85% of O-2A progenitors exposed to endothelial cell conditioned medium expressed the typical NSP4⁺ GFAP⁺ phenotype and stellate morphology of type-2 astrocytes. Only a small proportion (8.3%) of O-2A lineage cells were GaIC⁺ oligodendrocytes in these cultures. It should be noted that staining of cultures with only the anti-GFAP antibody and secondary fluorochrome, followed by analysis for the presence of stellate astrocytes, offers a suitably sensitive method for assaying the activity of the factor of the invention.

We next compared the effect of medium conditioned by cloned rat brain capillary endothelial cells with the effects of medium conditioned by other (non-endothelial) cells isolated by identical methods (prepared as in Section G). In these experiments, only medium conditioned by the endothelial cells was effective at inducing the differentiation of O-2A progenitors into type-2 astrocytes (Figure 2). Conditioned medium derived from meningeal cells, fibroblasts or type-1 astrocytes did not promote type-2 astrocyte differentiation in cell cultures prepared from optic nerves of 7d old rats (Figure 3). As all endothelial cell preparations were equally effective at inducing astrocytic differentiation of O-2A progenitors, aortic endothelial cells (which were the easiest to grow in large numbers) were used as a source of conditioned medium in all further experiments. This material is from now on referred to as EndoCM.

(P) Promotion of type-2 astrocyte differentiation by EndoCM is dependent on age of the animals from which eels were derived and on the plating density of the optic nerve cells.

The ability of EndoCM to promote astrocytic differentiation of O-2A progenitors required that the optic nerve cultures were derived from appropriately aged animals and plated at an appropriate density (Figure 4). For example, EndoCM did not induce astrocytic differentiatio of O-2A progenitors in cultures which were prepared from optic nerve of 1 day old rats and grown for 3 days at plating densities of 1 -10,00 cells per 13mm PLL coated glass coverslip. In contrast, in culture derived from 7d old rats and grown at densities of _>4000 cells/coversli in the presence of EndoCM, 87% of the O-2A lineage cells differentiate into type-2 astrocytes within 3 days. However, induction of astrocyti differentiation was markedly less effective if cultures prepared from opti nerve of 7 day old rats were plated at densities of _<. 200 cells/coverslip. These results suggested that the action of EndoCM ove a relatively short time frame required the additional presence of a factor or factors, present in the higher density optic nerve cultures, althoug the factor of the invention itself is sufficient to induce astrocyti differentiation over a longer time period. Therefore, all assays o purification of the endothelial-cell derived activity were carried out usin optic nerve cells derived from P 7 rats and plated at a density of >_ 700 cells/coverslip.

(Q) Physical and chemical properties of EndoCM.

Prior to attempts to purify the activity contained in EndoCM the physical and chemical stability of the factor(s) was assessed t determine the purification protocol. pH stability: The activity contained in EndoCM was stable in overnigh incubation following titration of the pH of the medium (with either 0.1 HCI or 0.1 M NaOH) to a final pH in the range 2.5-9.5, spread over hal pH units. The conditioned media (with 1 mM EDTA and proteas inhibitors) were incubated at both room temperature or at 4°C overnigh after which they were neutralised to pH 7.0 and bioassayed. There wa no relative loss of activity compared with the untreated material eith at room temperature and at 4°C. Heat stability: 200 μ\ of EndoCM was heated at 90°C for 5 min in a microfuge tube. This material was then biossayed. There was a recovery of 90% of the initial activity when compared to the untreated material. Trypsin sensitivity: 25 /I of insofuble-trypsin bound to glass beads were washed twice in a microfuge tube by adding 500 μ\ of PBS, mixing and centifuging for one min. The liquid above the beads was removed and 100 μ\ of EndoCM (without protease inhibitors) was added. Incubation for 30 min at 37°C, with constant mixing (followed by addition of 10 //I of-SBTI-DNAse were added to stop the enzymatic reaction), followed by removal of beads, produced a total loss of bioactivity. Freeze-thawing: There was no loss of activity when compared to the control material when 200 μ\ of EndoCM was subjected to five repeated freeze-thaw cycles.

Ionic-nature: Following binding to different chromatography beads in a test-tube assay, the supernatant from each tube was assayed. No binding of the activity was seen with CM-ceilulose at any pH point. Binding of the activity was seen with DEAE-cellulose at pH above 6.5. This experiment indicated that the activity of interest was anionic in nature and that DEAE-cellulose with a neutral buffer was the appropriate gel for purification.

Sensitivity to reducing agents: Overnight incubation (at 4°C) of EndoCM with either 2-mercaptoethanol (final concentration = 5 mM) or dithiothreitol (2mM) produced a total loss of activity.

(R) DEAE-cellulose chromatography

Ultrafiltration with the Pellicon system was used to reduce sample volume while at the same time allowing buffer exchange for subsequent chromatography. The initial step employed ion exchange chromatography on DEAE-cellulose. All detectable biological activity bound to the anion exchange column. When the column was developed with a gradient of 0-0.5M NaCl, the biological activity eluted at a concentration of approximately 0.25M (Figure 5). The peak of activity ran with the main protein peak, thus suggesting that the main use of this step was to further concentrate the activity.

(S) Mono-Q chromatography

Active fractions from the DEAE-cellulose chromatograph column were dialysed against 10mM tris-HCl, 1 mM EDTA pH 7.8 and freeze dried. The sample was then reconstituted in a small volume and the conductivity was adjusted to achieve a salt molarity of 50mM and applied to a Mono-Q column. All biological activity bound to the column, which was washed and developed with a gradient of 0-1.0 M NaCl in 10mM tris-HCl, 1 mM EDTA pH7.8. Multiple samples were applied to the column to maintain an optimum column protein loading on the column. Inducing activity eluted at approximately 0.45M. (Figure 6). 75% of th activity applied to the Mono-Q column was recovered, and the activit was enriched 15 fold between the DEAE-cellulose and Mono-Q columns.

(T) Reverse phase chromatography

The active fractions from the Mono-Q column (with 0.1 TFA) were injected onto a Pro-RPC reverse phase column (Figure 7). Al the biological activity bound to the column, which was washed an developed with a gradient of 0-90% acetonitrile (0.1 % TFA). The activ fraction was eluted at an acetonitrile concentration of approximatel 35%. The yield from the Pro-RPC column was approximately 60% wit a purification of 60 fold over the Mono-Q column.

The concentrated material from the Mono-Q column wa diluted 1 :10 in 60% acetonitrile:0.1 % TFA and left for 3 days at roo temperature and at 4°C. The material was then assayed and compare with control material from the Mono-Q column which was not stored in acetonitriie:TFA. A dilution curve indicated no loss of activity over this time period. Other active fractions isolated from the Pro-RPC column in acetonitrile:TFA was stored at -20°C and assayed for activity at a variety of time periods. No reduction of activity was observed over a period of at least 6 months.

(U) Gel filtration chromatography and SDS-gel electrophoresis

A sample from the active fractions of the Pro-RPC chromatography (dried and reconstituted with 0.15 M NaCl) was also applied to a gel filtration column (Superose 12, Pharamacia) and ran in 10 mM tris-HCl, 1 mM EDTA with 0.15 M NaCl. This procedure yielded an activity with an apparent M_r of approximately 33 Kd (Figure 8), in close agreement with the data obtained from SDS-PAGE analysis.

The active fraction from the Pro-RPC column was checked for purity by running on SDS-PAGE followed by silver staining (Figure 9). A parallel gel was run to determine the position, and hence the apparent relative molecular weight (M_r) of the inducing activity. Proteins were eluted from sections of the gel and bioassayed for GFAP-inducing activity. The M_r of the inducing activity from the gel was calculated as approximately 31 kd.

(V) Enhancement of the astrocytic differentiation of O-2A progenitors by co-application of CNTF and the factor of the invention.

We next examined the properties of material purified to the stage of elution from a Pro-RPC column (EndoCM/ProR) to determine whether this material cooperated with CNTF to promote astrocytic differentiation of O-2A progenitors, as has been described for the extracellular matrix produced by endothelial cells (Lillien, L. E., Sendtner, M., and Raff, M. C. (1990) J. Cell Biol. 111 , 635-644). In these studies we examined the induction of GFAP expression in O-2A progenitor derived from optic nerves of 7 day old rats and enriched to .>_99.5 purity by panning on antibody coated dishes. Purified populations wer used in this particular assay so as to eliminate effects of type-1 astrocytes, which produce CNTF in these cultures (Lillien, L. E. Sendtner, M., and Raff, M. C. (1990) J. Cell Biol. 111 , 635-644). I these particular experiments, the A2B5 monoclonal antibody was use both to purify O-2A progenitors and as a marker of O-2A lineage cell when cultures were assayed (Raff, M. C, Miller, R.H. and Noble, M (1983) Nature 303, 390-396).

As shown in Figure 10A, purified O-2A progenitors grow for 3 days in the presence of CNTF alone differentiated int oligodendrocytes, as was the case for control cells grown in DMEM-BS In contrast, cells grown for 3 days in the presence of EndoCM/Pro alone did not differentiate into oligodendrocytes, but instead continue to express the A2B5 ⁺ GFAP^'GalC antigenic phenotype characteristic o O-2A progenitors. Unlike the effects of exposure to either EndoCM/Pro or CNTF alone, exposure to both of these materials together promote the astrocytic differentiation of almost 90% of the O-2A progenitor within 3 days of in vitro growth. Although it is on occasion possible t see GFAP expression induced by the factor of the invention over a perio of 2-3 days, this expression is always synergistically enhanced by th presence of CNTF or LIF.

(W) Induction of astrocytic differentiation of O-2A progenitors by th activity of the factor of the invention.

As shown in Figure 10B, application of the factor of th invention to purified O-2A progenitors induced differentiation of thes cells into type-2 astrocytes over the course of 6 to 8 days of growth in vitro. Cells exposed to CNTF alone for this period of time showed little or no differentiation along the type-2 astrocyte pathway, while 50% of the cells exposed to material purified through the Pro-RPC stage of purification differentiated into A2B5 ⁺GFAP⁺ type-2 astrocytes in 6 days. I _R.Pjjμlations of O-2A progenitors exposed to EndoCM/ProR for 8 days in the absence of added CNTF exhibited almost complete differentiation into type-2 astrocytes.

(X) Promotion of oligodendrocytic differentiation by the CNTF when applied on its own to O-2A progenitor cells.

The ability of CNTF to promote oligodendrocyte differentiation when applied to O-2A progenitors in the absence of the factor of the invention was first demonstrated by growing purified O-2A progenitors in the presence of basic fibroblast growth factor (bFGF, which inhibits oligodendrocyte differentiation; McKinnon et al., 1990, Neuron 5:603) and either exposing or not exposing these cells to CNTF. In cultures grown in DMEM-BS containing bFGF for 6 days, 12% of the O-2A progenitors differentiated into oligodendrocytes (Figure 1 1 ). In contrast, the additional presence of CNTF was associated with a 3-fold increase in the production of oligodendrocytes over this time period. CNTF also promoted oligodendrocyte maturation, as demonstrated by the ability of this compound to induce more rapid differentatiation of oligodendrocytes to the stage of producing myelin basic protein (Figure 12). In cultures grown in DMEM-BS alone for 3 days 25% of the oligodendrocytes were myelin basic protein positive after 3 days, as compared with expression of myelin basic protein in 57% of the oligodendrocytes in cultures of purified O-2A progenitors grown in the presence of DMEM-BS + CNTF.

The above experiments demonstrate that the factor of the invention, along with the properties described thus far, has the furthe effect of altering the effects of CNTF application on primary O-2 progenitors isolated from optic nerves of perinatal rats from that o promoting oligodendrocyte survival to that of promoting astrocyti differentiation. The possibility that CNTF could promote oligodendrocyt survival or differentiation as well as causing a transient induction o GFAP expression in.O-2A progenitors has not been appreciated in the ar (Hughes, S.M. and Raff (1987) Development 101, 157-167; Lillien, L E., Sendtner, M., Rohrer, H., Hughes, S. M. and Raff, M. C. (1988 Neuron 1 , 485-494; Lillien, L. E., Sendtner, M., and Raff, M. C. (1990 J. Cell Biol. 111 , 635-644), due to the design of earlier experiments.

(Y) The factor of the invention interacts with LIF to promote astrocyti differentiation of O-2A progenitors. All experiments were carried out i the same manner as the experiments with CNTF on purified O-2 progenitors, except that instead of CNTF, LIF was applied at concentration of 2ng/ml. As shown below, at a 3-day time point, LI applied in the absence of the factor of the invention did not induc astrocytic differentiation, and the factor of the invention also did no induce astrocytic differentiation over this time period. In contrast, th presence of both LIF and the factor of the invention resulted in th differentiation of 50% of the O-2A lineage cells into type-2 astrocyte

TABLE 2 % of O-2A lineage cells which were type-2 astrocytes on Day 3

LIF + CNTF 0

CNTF 0

ECM/ProR 0

LIF 0

ECM/ProR + LIF 50% (Z) The factor of the invention co-operates with tumor necrosis factor-alpha (TNF) to kill oligodendrocytes. Purified O-2A progenitors were grown in DMEM-BS to 3 days in order to cause all cells to differentiate into oligodendrocytes. After this time cultures were switched to the conditions indicated in Table 3. As shown in this Table, TNF only killed oligodendrocytes when it was applied in the presence of the factor of the invention, applied in this experiment as a preparation purified through the Mono-Q stage of purification. That the killing was mediated through classical TNF pathways is shown by the ability of N-acetylcysteine (NAC, applied at 20 mM) to block killing of oligodendrocytes.

TABLE 3

Coπditions% dead oligodendrocytes

TNF (100 ng/ml) 36 4

Interferon gamma (IFN) 29 _+, 4

TNF + MonoQ 91 ± 3

TNF + MonoQ + IFN 96 _+ 2

TNF + Mono Q + NAC 26 _+ 4

TNF + Mono Q + IFN + NAC 29 + 2

Claims

CLAIMS :

1. A basic secreted polypeptide factor which has, if obtained from bovine aortic endothelial cells, an observed molecular weight, of from about 29 kD to about

34 kD on SDS-polyacrylamide gel electrophoresis using the following molecular weight standards:

Phosphorylase b 97,400 Bovine serum albumin 60,000

Ovalbumin 45,000

Carbonic anhydrase 31,000

Soybean Trypsin Inhibitor 21,000 Lysozyme 14,000;

which factor has the ability to promote the differentiation of 0-2A progenitor cells into type-2 astrocytes when applied to 0-2A progenitors growing in serum-free medium, and which promotion of astrocytic differentiation is more rapid when said factor is applied to 0-2A progenitors in the presence of CNTF or LIF.

2. A basic secreted polypeptide factor having an observed molecular weight of from about 29 kD to about 34 kD by chromatography on Superose 12 using the following molecular weight standards:

Thryoglobulin 670,000

Gamma Globulin 158,000 Ovalbumin 44,000

Myoglobin 17,000

Vitamin B12 1,350;

which factor has the ability to promote the differentiation of 0-2A progenitor cells into type-2 astrocytes when applied to 0-2A progenitors growing in serum-free medium, and which promotion of astrocytic differentiation is more rapid when said factor is applied to 0-2A progenitors in the presence of CNTF or LIF.

3. A actor as claimed in claim 1 or claim 2 and having an observed molecular weight as stated of from about 31 to 33 kD.

4. A process for the preparation of a polypeptide •defined in claim 1 or claim 2, comprising collecting medium conditioned by endothelial cells to obtain protein, subjecting the resulting protein material to chromatographic purification and collecting a fraction which has an apparent molecular weight of 29 kD to 34 kD either when analyzed by SDS-polyacrylamide gel • electrophoresis using the following molecular weight standards:

Phosphorylase b 97,400

Bovine serum albumin 60,000

Ovalbumin 45,000 Carbonic anhydrase 31,000

Soybean Trypsin Inhibitor 21,000

Lysozyme 14,000,

or when analyzed by chromatography on Superose 12 using the following molecular weight standards:

Thryoglobulin 670,000

Gamma Globulin 158,000

Ovalbumin 44,000 Myoglobin 17,000

Vitamin B12 1,350;

and which fraction has the ability to promote the differentiation of 0-2A progenitor cells into type-2 astrocytes when applied to 0-2A progenitors growing in serum-free medium.

5. A process as claimed in claim 4, wherein said conditioned medium is passed to Q-Sepharose chromatography.

6. A process as claimed in claim 4, wherein said purification includes SDS-gel electrophoresis, removal of SDS, and reversed phase chromatography.

A process as claimed in claim 4, wherein said conditione^r" medium is initially concentrated by ultrafiltr. ion and, if desired, buffer exchange.

8. A process as claimed in claim 4 or claim 7, wherein said chromatographic purification includes sequentially employed DL E-cellulose chromatography, mono-Q-FPLC, and/or reversed phase chromatography prior to SDS-gel electrophoresis or Superose 12 chromatography.

9. A process as claimed in any one of claims 4 to 8, wherein at each stage biological activity of material obtained is assessed using induction of astrocytic differentiation in 0-2A progenitor cells (optionally derived from appropriate tissues of the rat) as a measure in an assay.

10. A process as claimed in any one of claims 4 to 9, wherein the endothelial cells used to condition the starting medium are bovine aortic endothelial cells.

11. A pharmaceutical or veterinary formulation comprising a polypeptide as claimed in any one of claims 1 to 3 formulated for pharmaceutical or veterinary use, respectively, optionally together with an acceptable diluent, carrier or excipient and/or in unit dosage form.

12. The use of a polypeptide as defined in any one of claims 1 to 3 as a glial cell differentiation agent in vivo or in vitro.

13. A polypeptide as claimed in any one of claims 1 to 3 for use as a glial cell differentiation agent.

14. A method for producing a glial cell differentiation effect in a vertebrate by administering an effective amount of a polypeptide as defined in any one of claims 1 to 3.

15. The use of a polypeptide as defined in any one of claims 1 to 3 as a cooperating factor with a differentiation modulating factor, optionally CNTF or

^• LIF, in vivo or in vitro.

16. A polypeptide as defined in any one of claims 1 to 3 for use as a cooperating factor with a differentiation modulating factor, optionally CNTF or LIF.

17. A method for producing a glial cell differentiation effect in a vertebrate by administering an effective amount of a differentiation modulating factor, optionally CNTF or LIF, and a polypeptide as defined in any one of claims 1 to 3.

18. The use of a polypeptide as defined in any one of claims 1 to 3 in enhancing the activity of TNF.

19. A polypeptide as defined in any one of claims 1 to 3 for use as a TNF activity enhancer.

20. A method for producing an enhanced cell killing effect by TNF comprising administering an effective amount of a polypeptide as defined in any one of claims 1 to 3.

21. A method of treatment or prophylaxis in which a polypeptide as defined in any one of claims 1 to 3 is administered to enhance the effect of TNF.

22. A method as claimed in claim 14 or claim 17 when used in the treatment or prophylaxis of a disease or disorder.

23. A method for the prophylaxis or treatment of a pathophysiological condition in which a cell type is ;..-volved which is sensitive or responsive to a polypeptide as defined in any one of claims 1 to 3 or to the action of a factor, the action of which can be modified by a polypeptide as defined in any one of claims ^• 1 to 3, said method including administering an effective amount of said polypeptide.

24. A method as claimed in claim 23 in which tumour cell differentiation is promoted.

25. The use of a polypeptide as defined in any one of claims l to 3 in the manufacture of a medicament.

26. The use of a polypeptide as defined in any one of claims l to 3 as an immunogen to generate antibodies, optionally said antibodies being for diagnostic purposes.

27. The use of a polypeptide as defined in any one of claims 1 to 3 in a competitive assay to identify or quantify molecules having receptor binding characteristics corresponding to those of said polypeptide.

28. The use of claim 27, wherein said polypeptide is labelled, optionally with a radioisotope.

29. The use of a polypeptide as defined in any one of claims 1 to 3 in an affinity isolation process. optionally affinity chromatography for the separation of a corresponding receptor.

30. A method for the prophylaxis or treatment of glial scar formation, cachexia, septic shock, AIDS, multiple sclerosis, acute injury or stroke, which comprises administering an effective amount of a substance which inhibits the binding of a polypeptide as defined in any one of claims l to 3 to a receptor therefor.

31. An antibody to a polypeptide as defined in any one of claims 1 to 3.