CA2169945A1 - Factor for fibroblast colony forming unit - Google Patents

Abstract

A colony forming unit fibroblast (CFU-F) factor has been identified having the following characteristics: (a) it is a polypeptide present in P388 or 416B (ECACC provisional accession no. 94082401) cells and capable of extraction from the cell lysate or conditioned medium;
(b) when added to explanted bone marrow adherent cells which are below the level capable of forming fibroblast colonies, it is capable of inducing the formation of fibroblast colonies; (c) it binds heparin-agarose at physiological pH; (d) the heparin binding fraction, when subjected to reducing SDS-PAGE, produces bands at about (12, 15, 16 and 17kD); (e) its activity is not affected by heating at 100 °C for 10 minutes; (f) it can be purified from the cell lysate by fractionation by anion-exchange chromatography, eluting at 0.3M sodium chloride;
followed by size exclusion chromatography, optionally followed by further purification such as HPLC and/or SDS-PAGE.

Description

FACTOR FOR FIBROBLAST COLONY FORMING UNIT

This invention relates to colony-forming unit fibroblast (CFU-F), and the factor which is responsible for its activity.
The definition of the haemopoietic (blood-forming) tissues of the bone marrow is relatively thorough and many of the factors which influence their performance are known and have been produced in recombinant form over the last 5-6 years. In the intact animal, and in culture systems which mimic the whole animal tlong-term bone marrow cultures), the processes of blood cell formation are dependent not only upon the blood-forming cells themselves, but also upon the microenvironment (or stroma) within which they reside. One cellular component of this complex is a fibroblast cell, which for the last 20 years has been studied in primary culture systems in the assay known as the CFU-F assay. Cells derived from CFU-F may have multiple roles to play, including (but not limited to) bone formation. Some fibroblasts may also play a part in tissue homeostasis and regeneration. Factors influencing their growth may well also have a role to play not only in bone formation and remodelling but also in repair and tissue healing.
The CFU-F assay was first described by Alexander Friedenstein in the early 1970s and at that time was applicable only to Guinea pig bone marrow cells. In the following years the assay has been adapted to suit many species including rat, mouse, rabbit and human, and remains the only quantitative assay for any component of the haemopoietic microenvironment.
The CFU-F assay comprises simply the addition of freshly isolated bone marrow cells to a tissue culture flask containing tissue culture medium and (carefully pre-selected) foetal calf serum (FCS). The number of cells explanted shows, within certain limits, a linear correlation with the number of CFU-F colonies formed.

W095/0~63 PCT/GB94/01868 ~lGg~4~

Should the number of cells explanted fall below a certain critical level (and the exact level varies from one species to another) then no colonies are formed. To regain linearity the addition of excess sterilized (by irradiation) bone marrow cells was required. Under these conditions, which we refer to as "feeder supplemented"
cultures, the assay shows precise and reproducible - linearity.
A modification introduced by Prof. A.J. Friedenstein took this observation to its extreme. ("Bone marrow s~ colony formation requires stimulation by haemopoietic cells", Friedenstein, A.J., Latzinik, N.V., Gorshaya, Y.F., Luria, E.A. and Moskvina, I.L.; Bone and Mineral 18, 199-213, 1992 and "Haemopoiesis - A Practical Approach", Editors: Testa, N.G. and Molineux, G., Oxford University Press, 1992). Bone marrow target cells were explanted at low numbers, allowed to adhere to the flask for 2 hours, and then all other cells were removed by washing. This had the effect of;
- 20 (i) Allowing the adherence of 100% CFU-F.
(ii) No colonies formed since no feeder activity was present.
(iii) Complete colony formation could be regained by adding sufficient sterilised bone marrow cells as feeders.
The present invention describes the purification and characterisation of CFU-F factor.

Brief descriPtion of the drawinqs Figure 1: The graph shows the results of an experiment to investigate the effect of b-FGF on colony formation at low and high serum concentrations. Adherent cells obtained from aliquots of 3 X 105 rat femur bone marrow suspension cells were used as targets. The total culture volume was 5ml and b-FGF was added at concentrations of 5, 10, 15 and 20ng/ml. Flasks were incubated at 37C for 7 days and then ~ wo 9~,0~0-~ ~ 1 B ~ 9 4 5 PCT/GB94/01868 fixed, stained and colonies counted.

Figure 2: The graph shows the results of an experiment to investigate the effect of PDGF on colony formation at low and high serum concentrations. Adherent cells obtained from aliquots of 3 X 105 rat femur bone marrow suspension cells were used as targets. The total culture volume was Sml and PDGF was added at concentrations of 5, 10, 15 and 20ng/ml. Flasks were incubated at 37C for 7 days and then fixed, stained and colonies counted.

Figure 3: The graph shows the effect of adding heparin (l~g/ml and b-FGF (lOng/ml) to adherent cells in the presence of 10' irradiated (lSOOcGy,yl3'Cs) rat femur bone marrow suspension cells. Adherent cells obtained from aliquots of 105 rat femur bone marrow suspension cells were used as targets and total assay volume was Sml. Flasks were incubated at 37C for 7 days and then fixed, stained and colonies counted. The data are compared to the positive control (ADS+) i.e. the number of colonies formed from the adherent cells in the presence of irradiated feeders only. ADS-refers to the negative control i.e. the number of colonies formed from adherent cells with no feeders added.
Figure 4-11: These graphs depict the assay of cell lines and their conditioned medium (CM) for feeder activity.
Adherent cells obtained for aliquots of 105 rat femur bone marrow suspension cells were used as targets and total assay volume was 5ml. All feeder cells were irradiated with l500cGy (y,13'Cs) and cell numbers quoted on the graphs refer to total feeder cells added per assay. Conditioned medium was filtered through 0.2,um Acrodisc filters prior to use and was added to the assays to the final % quoted.
Flasks were incubated at 37C for 7 days and then fixed, stained and colonies counted. The data are compared to the positive control (ADS+) i.e. the number of colonies formed from the adherent cells in the presence of 10' irradiated rat femur bone marrow suspension cells. ADS-refers to the negative control i.e. the number of colonies formed from adherent cells with no feeders added. The * on certain graphs indicates that the initial concentration of irradiated feeders added to the assay was too high causing a decrease in pH of the growth medium, hence the concentration of feeders was lowered 10 fold on the second day of the assay.
Figure 12: This shows the Sephadex G-50 Elution Profile. A dialysed 416B cell lysate was fractionated by anion-exchange chromatography on DE-52. The fraction eluting at 0.3 M sodium chloride was subsequently applied to a 700 X 26mm column of G-50 pre-equilibrated with 50mM
HEPES, 150mM sodium chloride pH 7.2. The column was then eluted with the same buffer at 30ml/hour and 85 X 5ml fractions were collected. These fractions were filtered through 0.2~um Acrodisc filters and then assayed for feeder activity. Adherent cells obtained from aliquots of 105 rat femur bone marrow suspension cells were used as targets and total assay volume was 5ml. Flasks were incubated at 37C
for 7 days and then fixed, stained and colonies counted.
The data are compared to the positive control (ADS+) i.e.
the number of colonies formed from the adherent cells in the presence of irradiated feeders only. ADS-refers to the negative control i.e. the number of colonies formed from adherent cells with no feeders added. The column had been calibrated with molecular mass markers to allow an estimation of molecular mass of eluting components.

Figure 13: This shows the elution profile from heparin-agarose chromatography of cell lysate or conditioned medium (CM) obtained from 416B cells.
Figure 14: This shows the 15~ SDS-PAGE gel fractionation of the 416B cell conditioned medium after .~ WO 9J/~0~3 PCT/GB94/01868 ~1~994~

heparin-agarose fractionation, eluting with 0.4-0.8M NaCl.

Figure 15: This shows the results of the assay involved in eliminating TGF-B as the CFU-F factor.

Detailed Description of the invention The assay procedure used in our work was based on the above described procedure of Friedenstein, as follows:
l. Working quickly to avoid degeneration of the cells, obtain single cell suspension of rat bone marrow cells.
Using male OBW rats of any age sacrifice and aseptically remove the femur. Flush out the contents of the femur with sterile Iscoves medium using gentle passage through syringe and 21G needle and count the nucleated cells. Use cells within one hour to ensure a high percentage of viable cells.

2. Prepare 25cm2 tissue culture flasks (of high quality, Falcon 3013E or equivalent) containing 105 marrow cells (sufficient to give rise to approximately 50 colonies) in Iscove's medium (320 mOsM/kg, prepared using water of high purity i.e. MilliQ or better) supplemented with 15% FCS.
FCS of suitable quality is difficult to obtain. Around l in 15-20 batches works. To obtain a batch of suitable quality screen samples at between 10 and 25% v/v in CFU-F
cultures prepared using unseparated bone marrow cells.
Maintain a concentration of at least 106 bone marrow cells per ml in these cultures. Gas the flasks with 5% CO2, 5% 2 in N2, place at 37C for 2 hours and leave undisturbed.
This gas mixture is essential if high plating efficiency is to be attained. Flasks must also be tightly closed to maintain pH.

3. Pour off supernatant liquid and aspirate residuum (in definitive experiments wash the flasks gently at this stage, this is not necessary for routine assays).

4. Replace supernatant with fresh medium and add "feeder"
activity.

5. Normal experimental design includes WO95/06063 ~16 9 9 4 S PCT/GB94/01868 a) Adherent cells only (zero colonies) b) Adherent cells plus irradiated marrow cells as positive control (100% for each specific assay) c) Complete cells (i.e. non-adherent cells not removed) d) Complete cells plus irradiated marrow feeder cells (to confirm that no potential CFU-F colony formers are removed in the adherence procedure).
Following standardisation of the assay system as indicated above, we then set about looking for a suitable source of the feeder activity. We also had to eliminate known growth factors as the source of the activity.

Elimination of known qrowth factors We found that, no known growth factor including FGF, EGF, PDGF, Interleukins 1-7, G-CSF, GM-CSF, SCF, HGF, MK
and TGF-B show this type of activity.
For example; basic fibroblast growth factor (b-FGF) was tested at high and low serum concentrations and these data are shown in Figure l.
Platelet-derived growth factor (PDGF) is synthesised by both platelets and megakaryocytes and is the most active known growth stimulating factor for fibroblasts. PDGF was assayed at low (2%) and high (15%) FCS concentrations.
These data are depicted in Figure 2.
The graphs clearly show that neither growth factor at either serum co~cPntration can reproduce the activity observed in the positive controls.
However, it is possible that FGF requires the presence of heparin to exert its effects. It is known that the b; nA; ~g of heparin to FGF facilitates the steric presentation of the facto~ to the cell surface receptor.
This hypothesis was investigated under the conditions of our assay these data are presented in Figure 3. The graph shows that when heparin and b-FGF are added to adherent cells and feeders, the efficiency of colony formation is lowered, hence the formation of CFU-F under our conditions W09~,~C~ 3 PCT/GB94/01868 21~3915 is not an FGF/heparin dependent event.

Identification of a Cell Line as a Feeder Source To facilitate identification and purification of the factor it was necessary to find a simple and readily amplified source. Obviously the irradiated bone marrow, used as our positive control was far too complex a mixture to ascertain which cells were producing the factor.
The ideal solution to this problem was to find a cell line which could produce the factor, that was also easily maintained and could be relied upon as a constant renewable source.
A number of cell lines available to us were assayed as feeder sources.
Assays were set up as previously described, using targets derived from the aliquots of 105 rat bone marrow suspension, Iscove's medium containing 15~ FCS in Falcon T-25 flasks containing a total of 5ml. Cells were spun down at 800g for 8 minutes at 20C, resuspended at suitable concentrations in Iscove's medium and then irradiated at 1500 cGy (y, l3'Cs).
The conditioned medium from these cells was also assayed for feeder activity after filtering through 0.2~m Acrodisc filters.
Cell lines assayed: L1210, TK6, RAJI, KYM, K562, P388, 416B, HOC-8 (data not presented) These data are shown on Figures 4-11; the positive control was counted as 100~ and all other results as fractions thereof. It is clear from figures 4-8, that none of the irradiated cells or their conditioned medium were a positive source of the factor.
Figure 9 shows that irradiated P388 cells were found to be a good source of the factor, giving up to 50% of the activity of the positive control at a concentration of 2.5xlO5/ml and the CM from these cells a maximum of 4~
activity of the positive control when CM was added to the assay to a final concentration of 10%.

~ 1 6 99 ~ ~

The experiment using irradiated P388 cells (a publicly available cell line) was repeated and Figure 10 depicts these data. This suggests that a concentration of lxlOs/ml feeder cells was most efficient at stimulating CFU-F
formation and these data were found to be reproducible. In addition CM was concentrated 10 fold by Amicon Ultrafiltration, in an attempt to concentrate the low activity observed in the CM, this was unsuccessful.
P388 was easily cultured in Iscove's medium and 10%
horse serum (HS) and hence work began on further characterization of the factor. Microscopic observation of the irradiated feeder cells in the assay system suggested that cell-cell contact between targets and feeders may be necessary. Feeders were seen to be closely associated with the CFU-F.
It therefore seemed possible that the factor could be a peripheral or integral membrane protein. Various buffers were used to disrupt ionic interactions to remove peripheral proteins from the P388 cell surface and buffers containing detergent were used to release the integral membrane proteins from the lipid bilayer. Subsequent fractions were assayed, after dialysis to remove the buffers. Very little activity was observed with the peripheral proteins fractions, although there was some activity in the whole membrane and digested membrane fractions. However, our failure to completely remove the detergents from the fractions prior to assay lead to inconsistent data. Solubilized samples of P388 membranes were also subjected to ion-exchange chromatography (IEC) and size-exclusion chromato~raphy (SEC). These experiments were limited in success, as the sensitivity of the assay to the reagents involved was always a problem.

The cell line 416B was tested as a feeder source in our assay system, after irradiation at 1500 Rads. The data _ WO95/06063 ~ 16 ~ 9 ~ S PCT/GB94/01868 for this initial experiment are shown in Figure 11. The irradiated cells and CM were shown to be equally good sources, giving 32% of the activity of the positive control where irradiated cells were added at a concentration of 5 x 105/ml and 10% CM respectively. At a concentration of 5 x 106/ml, irradiated 416B cells were found to give up to 80%
of the activity of the positive control.
The 416B cell line was derived from long term bone marrow cultures (LTBMC) infected with friend leukaemia virus (FLV)(Dexter, T.M., Allen T.D. & Scott D., "Isolation and characterisation of a bipotential haematopoietic cell line", Nature (1979) 277, 471-474.) The cell line is growth factor independent and has an activated ras gene.
There is also some evidence for the cells having the ability to differentiate into megakaryocytes which is of interest, in view of the data indicating that platelets are a positive source of the factor. The 416B cell line has been deposited under the Budapest Treaty provisions with the ECACC, Porton Down, England, on 24 August 1994, under the provisional accession number 94082401.

Conditioned Medium Logistically speaking, it should be much easier to purify a factor from CM than from a complex mixture of components e.g. a cell lysate, however we had the complication of growing the feeder cells in Iscove's medium and 10% HS. A time course was set up to ~X~m; ne the production of factor over time in serum free conditions i.e. Iscove's medium alone and also in phosphate buffered saline (PBS) - the simplest culture conditions we could maintain the integrity of the cells in. Fortunately, the production of factor varied very little between the Iscove's medium and the PBS. The optimum time for production of factor was 24 hours. Native polyacrylamide gel electrophoresis (PAGE) indicated the presence of less than 25 proteins smaller than lOOkDa, suggesting there would be relatively few contaminating proteins.

3 ~ $

It was found possible to concentrate the CM by freeze drying:

Added to assay: lml 0.5ml 0.25ml PBS CM 50% 50% 13%
X2 conc. PBS CM
~ of positive control 50% 66% 100%
___________________________________________________________ However, freeze drying the PBS CM was found to be a rate limiting step in the purification procedure and hence further progress was made with 416B cell lysates.
Attempts `to stimulate the cells to synthesize more factor into their growth medium using the lectin;
phytohaemaglutinin (PHA) and the phorbol ester; 12-0 tetradecanoylphorbol-13-acetate (TPA) were unsuccessful, however the CM was still considered to be an important source of activity.

416B Cell LYsates Crude cell lysates prepared by osmotic shock were found to be a positive source of activity. However, the lysates were seen to display only very low levels of activity unless fractionated further, probably due to the presence of some kind of inhibitor.

Lysis Method i) Cells grown up in Iscove's medium + 10% HS
ii) Spun down at 800g 8 minutes 20C
iii) Cells washed twice in PBS to remove residual HS.
iv) Cells resuspended in lysis buffer A ; lOmM HEPES, 2mM PMSF pH 7.2 and incubated on ice for 1 hour.
v) Cell debris spun down at 4000g 15 minutes 4C.
vi) An equal volume of lysis buffer B: 250mM HEPES, 5mM DTT, 5mM PMSF, 250mM sodium chloride pH 7.2, was then added to the supernatant.
Cells may be checked for lysis using trypan blue.

WO9~lOGOC~ ~ 16 3 9 4 5 PCT/GB94101868 ChromatoqraPhy (i) Anion-Exchange (Whatman DE-52) To lower the salt concentration the cell lysate was dialysed for 2 hours against 5 litres deionised water, the water being replaced after the first hour. The dialysed extract was then pumped onto a pre-equilibrated column (150 x 16mm) of DE-52. The equilibration buffer was lOmM HEPES
pH 7.2. the column was then washed until all unbound material had been eluted (tested by measuring absorbance at 280mm). Initially the column was eluted with increasing levels of salt (O.lM, 0.2M, 0.3M, 0.4M, 0.5M, l.OM in the equilibration buffer). Assay of these fractions after dialysis (to reduce salt concentration) indicated the highest level of activity to be in the 0.3M sodium chloride wash.

Added to assay: lml 0.5ml 0.25ml ___________________________________________________________ 0.3M wash 50% 100% 73%
% of positive control ___________________________________________________________ Anion-exchange chromatography proved to a successful first purification step, as the matrix has a high binding capacity and large volumes of material can be pumped onto the column without any detrimental effect on the separation. This step also serves to concentrate activity, as the desired fraction can be eluted in a small volume, hence suitable for a second step such as SEC, where sample size is important and affects resolution of components.
To PX~i ne whether the activity of the fractions at this stage could be ~ oved, they were exposed to heat (100C for 10 mins) or trypsin (2% for 1 hour at 37C) treatment. Preliminary data at this stage suggest the factor to be a peptide.
(ii) SEC (Pharmacia, Sephadex G-50) Sephadex G-50 is a SE matrix which can separate WO95/06063 ~ 5 PCT/GB94/01868 molecules with masses between 1.5-30kDa i.e. the mass range that most growth factors fit into. The dialysed 0.3M wash from the DE-52 column was applied to a pre-equilibrated column of G-50 (700 x 26mm) and 80 x 5ml fractions were collected. The equilibration and elution buffer was 50mM
HEPES, 150mM, sodium chloride (to prevent ionic interactions) pH 7.2.
As this buffer was compatible with the assay system these fractions could be assayed directly after filter sterilisation using 0.2,um Acrodisc filters.
The elution profile for this experiment can be seen in Figure 12. Clearly, most of the activity eluted at approximately 245ml. This elution volume corresponds to a molecular mass of less than 15kDa, which is typical of many growth factors.

(iii) Heparin-agarose Conditioned medium (CM) was prepared by incubation of 5xlOs/ml 416B in phosphate buffered saline (PBS) for 24 hours at 37C. The CM was filtered (0.2~m) prior to use and stored at 4C. A cell lysate was prepared as previously described.
A 2ml column of heparin agarose was prepared from type I heparin immobilised on agarose (Sigma) and equilibrated with lOmM HEPES pH 7.2. A 50ml aliquot of the CM was then applied to the column which was then washed with the equilibrating buffer. The column was then eluted with increasing concentrations of NaCl in the HEPES buffer. The cell lysate was dialysed and chromatographed by the same method. The resultant fractions were then desalted by dialysis against dH20 for 2 hours, the water being replaced after the first hour. All the above procedures were performed at 4C. The dialysed fractions were then filter sterilised (0.21um).
The fractions were assayed for stimulatory activity in the CFU-F assay, using the adherent cell depletion assay described above. Adherent cells from 105 rat bone marrow WO 95/06063 ~ 1 6 3 9 4 ~ PCT/GB94/01868 ( BM ) cells were used as targets and the fractions tested were added to a final concentration of 20% ( i.e. lml in a total of 5ml). After 7 days incubation at 37C, the medium was removed from the flasks, the adherent cells were then washed with PBS, fixed with methanol and stained with 1%
crystal violet. Colonies of fibroblasts (>50cells) were then counted.
The heparin agarose elution profile is shown in Figure 13. The number of colonies counted in the positive control for the assay was counted as 100%.
For PBS CM, the active fractions eluted from the column between 0. 4M and 1. OM NaCl and 1. 6M. For cell lysate, the active fractions eluted from the column between 0.4M and l.OM NaCl, 1.2-1.4 and 1.8-2Ø
It appears therefore that there is at least one heparin binding molecule, secreted into the PBS CM by 416B, that is capable of stimulating CFU-F colony formation.

(iv) SDS-PAGE
The active fraction eluting off heparin-agarose between 0.4-0. 8M NaCl was freeze dried and desalted by dialysis for 2 hours against dH20, the water being replaced after the first hour. The freeze-dried sample was then redissolved in SDS sample buffer with the addition of ~-mercaptoethanol, as a reducing agent, to one half. The samples were then heated to 95C for 4 minutes. The samples were then applied to a 4% stacking, 15% resolving acrylamide gel. Rainbow markers (Amersham) were used as an estimation of molecular weight. The gel buffer, sample buffer and electrode buffers were as defined in Schagger &
von Jagow, Analytical Biochemistry, (1987), 166, 368-379.
The gel was then run at 30V for 60 minutes to allow stacking and then 30mA for separation. The gel was then fixed and stained as described by Schagger & von Jagow (supra).
As shown in Figure 14, four bands were observed in the active~fraction off heparin-agarose, with approximate sizes 216~945 of 12, 15, 16 and 17 kDa. The gel thus indicates that there are four low molecular weight peptide components in the heparin-binding fraction. In hindsight, it is possible that the reducing agent in lane 5 could have contaminated the non-reduced sample in lane 4. Therefore, it is impossible to say whether or not the peptides are naturally monomeric or subunits of a larger protein. These four bands are routinely observed in the heparin-binding fraction.
Heat StabilitY
PBS CM was prepared as described above and was filter sterilised (0.2~m) into glass universal bottles and then placed into a boiling water bath for 10 minutes.
CFU-F assays were performed as described above. The untreated and heat treated CM was added to the CFU-F assay at the percentages indicated.
%CM COLONY No. IN COLONY No. IN
ASSAYS WITH HEAT ASSAYS WITH
TREATED CM UNTREATED CM
10% 6.5 5 20% 5.0 5 50~ 2.0 3-5 refers to the amoun of CM added to the assay. Values in the table refer to colony numbers (~50 cells). The negative control for the assay had a mean number of 1.5 colonies and the positive control 27.5.
It can be concluded that the ability of the PBS CM to stimulate CFU-F colony formation is not affected by heat treatment.

Eliminatina Growth Factors Hepatocyte Growth Factor (HGF) HGF is a heparin-binding growth factor, eluting at approximately 0.7-0.8M NaCl. It was necessary therefore to eliminate it as a component of the 416B PBS CM. The cell line MDCK is sensitive to HGF and in its absence grow as _ W095/06063 ~ 1 6 3 9 ~ ~i. ` . PCT/GB94/01868 clumps and in its presence are seen to be much more dispersed (scattered). The easiest way to eliminate HGF
was to see if the PBS CM could induce this scatter effect on the MDCK cell line.
In 24 well plates, in a total volume of lml the following experiments were set up. The growth medium used was Iscoves and 10% horse serum.
(a) 103 MDCK Alone (b) 103 MDCK HGF lng/ml 0 ( C ) 103 MDCK 10% PBS CM
(d) 103 MDCK 40~ PBS CM
(e) 103 MDCK 50% PBS CM
The only experiment resulting in the scatter effect was (b), from which it can be concluded that the PBS CM
does not appear to contain HGF.

Midkine (MK) MK protein is a pluripotential embryonic stem cell-derived neuroregulatory factor. It was first described as a retinoic acid-induced protein in HM-l embryonal carcinoma cells (Kadomatsu et al, Biochem Biophys Res Commun, (1988), 151, 1312-1318). MK has a 65~ homology with the 18kDa heparin-binding protein known as HB-GAM (Rauvala et al, EMBO J, (1989), 8, 2933-2941), pleiotropin (Li et al, Science, (1990), 250, 1690-1694) and heparin-binding neurotrophic factor (Bohlen et al, Growth Factors, (1990), 41, 97-107). It too elutes from heparin-agarose at similar salt concentrations to the active fraction from PBS CM and has a similar molecular weight. The MK cDNA has been cloned into the eucaryotic expression vector PXMT2 and this plasmid was supplied to us (J Heath). Purified recombinant factor was found to be inactive and therefore the factor must be assayed in the presence of the producer cells.
The plasmid DNA was electroporated into Cos-7 cells which were then incubated for 48 hours at 37C. Efficiency was monitored by a positive control where RSV LUC was electroporated into the cells and a negative control where W095/06063 ~ 16 9 9 4 3 PCT/GB94/01868 PGL 2 was used. Luciferase levels were estimated after 48 hours. The Cos-7 cells were lysed and 105 rat bone marrow suspension cells added to the flasks. It was not possible to transfer adherent cells to the flasks after trypsin treatment. Iscove's medium containing 15~ foetal calf serum was then added, the flasks gassed and incubated for 7 days at 37C.
There was no stimulation of CFU-F colony formation above background level in the presence of MK. We therefore conclude that MK is not responsible for the stimulation of CFU-F colony formation.

TGF-~l, TGF-~2 and TGF-~3 TGF-~ is a member of a family of related multifunctional cytokines abundant in platelets and bone marrow. TGF-~ is a potent inhibitor of many haemopoietic cells, with a decreasing effect as cells mature. However, its presence in both platelets and bone marrow necessitate its elimination as stimulatory molecule in the CFU-F assay.
The three growth factors were added to adherent cells in the ansence of irradiated feeders at the following concentrations; 25, 250, 2500, 25000 pg/ml. Assay conditions and times were as described above.
The results are shown in Figure 15. In the assay the negative control had an unusually high number of colonies, owing to a higher number of cells being added to each flask; however this was fortuitous as the negative effect on the colony number could be seen. The dose response seen was bell-shaped but a consistent inhibition of colony formation was observed with all three factors at 250pg/ml.

Concludinq comments The heparin-agarose separation procedure proved to be reliable and reproducible on CM or cell lysate. Active fractions from heparin-agarose chromatography can be further separated by SEC, as previously noted, and if desired further separated by HPLC and/or SDS-PAGE.

W095/06063 2 ~ 6 9 3 ~ ~ PCT/GB94/01868 Thus, the two-stage purification (anion exchange/SEC
or heparin-agarose/SEC) can be applied to CM or cell lysate to provide sufficient quantities of factor for further purification if desired, e.g. using HPLC and/or SDS-PAGE.
The SEC step is optional, but may be preferred in order to provide an additional purification step to the initial two-step procedure or to reduce the number of contaminating proteins going into later purification steps such as SDS-PAGE or HPLC.
The isolated polypeptides thus obtained can be further analysed, for example amino acid sequencing. This can be accomplished in conventional manner, eg after blotting the SDS-PAGE bands onto nitrocellulose.
The amino acid sequences obtained from the bands can be compared with databases of protein sequences with a view to eliminating any irrelevant proteins before investigating the CFUF properties associated with the protein bands.
The latter investigation may typically involve the cloning and expression of DNA encoding the polypeptides.
A cDNA library can be prepared from 416B for use in cloning and expressing DNA encoding the factor.

INDICATIONS REIATING TO A DEPOSITED MICROORGANISM
(PCI' Rule 13bis) A. The i~ a~ionC made below relate to the microorganism referred ~o in the description on page 9 , line 17 -- ~n 1~. IDENTIFICATION OF DEPOSIT Further deposits are identified on an a~ inn~l sheet O
Name of depositary ti~ - an EUROPEAN COLLECTION OF ANIMAL CELL CULTURES
Address of depositary ;~ vtin~ (inclulin~ postal code and col~ntry) E~ ed,- Collection of Animal Cell Cultures Centre for Applied Microbiolo~y ~ Research S~lisbl-ry Wiltshire GB
Date of deposit Accession Number 24 August 1994 416B 9408Z401 C. ADDITIONAL lNDICATIONS (leavt blank if not applicable) This infnrm~jon is continued on an ~ it; - I sheet O

D. DESIGNATED STATES FOR WHICH INDICATIONS ARE MADE (if~heindicationsarcnotforaadesigna~cdStatcs) ALL DESIGNATED STATES

E. SEPARATE FURNISIIING OF INDICATIONS (leave blank if not applicable) Ibe; ~ir-~;n~clistedbelowwillbes~bmit~Pdtothelntema~ionalBureaulater(specifythegcncral,,alurcofthcin~ca~ionscg.~Acccssion Nvm~r of Dcposit') For receiving Offce use only For International Bureau use only E~i lbis sheet was received with tbe international application [~ n~iS sheet was received by the International Bureau on:

Autbonzed officer Authonzed offcer NATHAN GJNTER
ROOM:- ~'3 E)(T~
Fomm PCIIRO/134 (luly 1992)

Claims (5)

Claims

1. A colony forming unit fibroblast (CFU-F) factor having the following characteristics:
(a) it is a polypeptide present in P388 or 416B (ECACC
provisional accession no. 94082401) cells and capable of extraction from the cell lysate or conditioned medium;
(b) when added to explanted bone marrow adherent cells which are below the level capable of forming fibroblast colonies, it is capable of inducing the formation of fibroblast colonies;
(c) it binds heparin-agarose at physiological pH;
(d) the heparin binding fraction, when subjected to reducing SDS-PAGE, produces bands at about 12, 15, 16 and 17 kD;
(e) its activity is not affected by heating at 100 deg C for 10 minutes;
(f) it can be purified from the cell lysate by fractionation by anion-exhange chromatography, eluting at 0.3M sodium chloride, or by heparin-agarose chromatography at physiological pH; optionally followed by size exclusion chromatography and/or further purification such as HPLC and/or SDS-PAGE.

2. A method of producing a CFU-F factor from a cell line which, when irradiated and added to explanted bone marrow adherent cells which are below the level capable of forming fibroblast colonies, is capable of inducing the formation of fibroblast colonies; said method comprising:
(a) culturing said cells and producing therefrom serum-free conditioned medium (CM) or cell lysate;
and subjecting said CM or cell lysate to (b) heparin-agarose chromatography at physiological pH
or to anion exhange chromatography eluting with 0.3M
NaCl, and collecting an active fraction therefrom;
optionally followed by (c) size exclusion chromatography, to obtain proteins having a molecular weight below about 30kD, preferably below about 20kD;
optionally followed by (d) further purification such as HPLC and/or SDS-PAGE
to separate the proteins, and identifying therefrom a component having CFU-F
activity.

3. A method according to claim 2, wherein the cell line is 416B (ECACC provisional accession no. 94082401).

4. A protein having CFU-F factor activity as obtainable by the method according to claim 2 or claim 3.

5. A protein according to claim 4:
(i) which binds heparin at physiological pH;
(ii) which produces a band at about 12, 15, 16 or 17 kD when subjected to reducing SDS-PAGE; and (iii) whose CFU-F activity is not affected by heating at 100 deg C for 10 minutes.