Angiogenesis, the development of new blood vessels from an existing vascular bed, is a complex multistep process that involves the degradation of components of the extracellular matrix and then the migration, proliferation and differentiation of endothelial cells to form tubules and eventually new vessels. Angiogenesis is important in normal physiological processes including, by example and not by way of limitation, embryo implantation; embryogenesis and development; and wound healing.

Angiogenesis is also involved in pathological conditions such as tumour cell growth and non-cancerous conditions such as ophthalmological conditions for example, neovascular glaucoma; diabetic retinopathy; age-related macular degeneration, pterygium ; retinopathy of prematurity; choroidal and other intraocular disorders Angiogenesis is also involved in pathological conditions such as atherosclerosis ; haemangioma; haemangioendothelioma; warts; hair growth; Kapos's sarcoma ; scar keloids; allergic oedema; dysfunctional uterine bleeding; follicular cysts; ovarian hyperstimulation,; endometriosis; peritoneal sclerosis, adhesion formation; obesity; osteomyelitis; pannus growth; osteophyte formation ; inflammatory and infectious processes (eg hepatitis, pneumonia, glomerulonephritis), asthma, nasal polyps, transplantation, liver regeneration; thyroiditis, thyroid enlargement ; and lymphoproliferative disorders.

The vascular endothelium is normally quiescent. However upon activation endothelial cells proliferate and migrate to form microtubules which will ultimately form a capillary bed to supply blood to developing tissues and, of course, a growing tumour. A number of growth factors have been identified which promote/activate endothelial cells to undergo angiogenesis. These include, by example and not by way

of limitation; vascular endothelial growth factor (VEGF) ; transforming growth factor (TGFb) ; acidic and basic fibroblast growth factor (aFGF and bFGF) ; and platelet derived growth factor (PDGF) (1,2).

VEGF is a an endothelial cell-specific growth factor which has a very specific site of action, namely the promotion of endothelial cell proliferation, migration and differentiation. VEGF is a dimeric complex comprising two identical 23 kD polypeptides. The monomeric form of VEGF can exist as four distinct polypeptides of different molecular weight, each being derived from an alternatively spliced mRNA. Of the four monomeric forms, two exist as membrane bound VEGF and two are soluble. VEGF is expressed by a wide variety of cell/tissue types including embryonal tissues ; proliferating keratinocytes; macrophages ; tumour cells. Studies (2) have shown VEGF is highly expressed in many tumour cell-lines including glioma and AIDS associated Karposi's sarcoma. VEGF activity is mediated through VEGF specific receptors expressed by endothelial cells and tumour cells. Indeed the VEGF receptor is up-regulated in endothelial cells which infiltrate tumours thereby promoting tumour cell growth. bFGF is a growth factor which functions to stimulate the proliferation of fibroblasts and endothelial cells. bFGF is a single polypeptide chain with a molecular weight of 16. 5Kd. Several molecular forms of bFGF have been discovered which differ in the length at their amino terminal region. However the biological function of the various molecular forms appears to be the same. bFGF is produced by the pituitary gland and is encoded by a single gene located on human chromosome 4.

A number of endogenous inhibitors of angiogenesis have been discovered, examples of which are angiostatin and endostatin, which are formed by the proteolytic cleavage of plasminogen and collagen XVIII respectively. Both of these factors have been shown to suppress the activity of pro-angiogenic growth factors such as vascular VEGF and bFGF. Both of these factors suppress endothelial cell responses to VEGF

and bFGF in vitro, and reduce the vascularisation and growth of experimental tumours in animal models.

We have discovered a potent, new inhibitor of angiogenesis which is a proteolytic fragment of fibrinogen.

Fibrinogen, the soluble circulating precursor of fibrin, is a dimeric molecule containing pairs of non-identical chains, (ie the oc-, P- and 7-chains). These are arranged as three discrete domains, the two outer D-domains and the central E- domain (4). Fibrinogen can be digested either by plasmin or thrombin.

The first step in plasmin cleavage of fibrinogen is the cleavage of the a chain C- terminal domain. Plasmin then cleaves the two D domains from the one E domain (consisting of the NH2 terminal regions of the a-, (3-and y-chains held together by disulphide bonds) and numerous smaller fragments including a small peptide, betal- 42 (amino terminal of the B-chain (5). Thrombin, on the other hand, produces a fibrin monomer and two copies of fibrinopeptides A and B (see figure 2) (4). Fibrinogen has been shown to accumulate around leaky blood vessels in solid tumours (5), Fibrinogen has also been shown to polymerise at host-tumour interface to form fibrin networks that promote tumour angiogenesis by supporting the adhesion, migration, proliferation and differentiation of endothelial cells (7).

The fibrin E-fragment (FnE-fragment), produced by the proteolytic cleavage of fibrin, stimulates angiogenesis in the chorioallantoic membrane assay (8). Furthermore, the amount of this protein present in invasive breast carcinomas positively corrolates with the degree of tumour vascularity (5).

W099/45135 describes polypeptides having, amongst other things, angiogenic activity. The polypeptides are referred to as Fibrinogen Domain Related (FDRG) because of a conserved carboxyl terminal region found in a number of polypeptides

(eg fibrinogen, angiopoietin, colin). This family of proteins have both a conserved structure and function. The polypeptide members of the FDRG family are distinguished from one another by unique variable amino-terminal regions. FDRG family members are implicated in a number of cellular processes including; modulation of angiogenesis, modulation of haematopoiesis, modulation of the proliferation, development or differentiation of adipocytes, modulation of insulin sensitivity and/or insulin responsiveness. However little, if any confirmatory experimental evidence is presented to fully corroborate these functions of the FDRG family of proteins. In addition the FDRG fragments disclosed in W099/45135 are located in the carboxy-terminal region of. the y chain and therefore not part of the E fragment.

STATEMENTS OF INVENTION According to a first aspect of the invention there is provided a nucleic acid molecule comprising DNA sequences selected from: i) a fragment of the DNA sequence encoding amino acids 1-78 of the a-chain of fibrinogen ; amino acids 43-122 of the (3-chain of fibrinogen ; and amino acids 1-62 of the y-chain of fibrinogen as represented in Figure 1 ii) DNA sequences which hybridise to the sequences presented in (i) which encode fibrinogen E which has anti-angiogenic activity ; and iii) DNA sequences which are degenerate as a result of the genetic code to the DNA sequences defined in (i) and (ii).

According to a further aspect of the invention there is provided a nucleic acid molecule comprising DNA sequences selected from : i) the DNA sequences as represented in Figure 6 ii) DNA sequences which hybridise to the sequences presented in Figure 6 which encode a polypeptide having anti-angiogenic activity ; and iii) DNA sequences which are degenerate as a result of the genetic code to the DNA sequences defined in (i) and (ii).

In a preferred embodiment of, the invention there is provided an isolated nucleic acid molecule which anneals under stringent hybridisation conditions to the sequences described in (i), (ii) and (iii) above.

Stringent hybridisation/washing conditions are well known in the art. For example, nucleic acid hybrids that are stable after washing in 0. 1xSSC, 0. 1% SDS at 60°C. It is well known in the art that optimal hybridisation conditions can be calculated if the sequence of the nucleic acid is known.

The DNA sequence of fibrinogen is known and, can be found in the NCBI website at http ://ncbi. nlm. nih. gov. using appropriate search terms.

According to a second aspect of the invention there is provided a polypeptide encoded by the nucleic acid according to the invention.

In a preferred embodiment of the invention said polypeptide is fibrinogen E.

In a further preferred embodiment of the invention fibrinogen E comprises the NH2 domains of the ot, ß, and y polypeptides.

In a still further preferred embodiment of the invention said fibrinogen E comprises amino acids 1 to 78 of the a-chain and amino acids 43 to 122 of the (3-chain ; and amino acids 1 to 62 of the y-chain, as represented in Figure 1.

In yet still a further preferred embodiment of ; the invention polypeptides comprising fibrinogen E is/are modified by deletion, addition or substitution of at least one amino acid residue. Ideally said modification enhances the antagonistic effects of fibrinogen E with respect to the inhibition of angiogenesis.

It will be apparent to one skilled in the art that modification to the amino acid sequence of polypeptides comprising fibrinogen E could enhance the binding and/or stability of the fibrinogen E with respect to its target sequence (e. g. VEGF and/or bFGF). In addition, modification of fibrinogen E may also increase the in vivo stability of the fragment thereby reducing the effective amount of fragment necessary to inhibit angiogenesis. This would advantageously reduce undesirable side effects which may result in vivo.

Alternatively, or preferably, said modification includes the use of modified amino acids in the production of recombinant or synthetic forms of fibrinogen E.

It will be apparent to one skilled in the art that modified amino acids include, by way of example and not by way of limitation ; 4-hydroxyproline, 5-hydroxylysine, N6- acetyllysine, N-methyllysine, N6, N6-dimethyllysine, N6, N6, N6-trimethyllysine, cyclohexyalanine, D-amino acids, ornithine. The incorporation of modified amino acids will confer advantageous properties on fibrinogen E fragments. For example, the incorporation of modified amino acids may increase the affinity of the fragment for its binding site, or the modified amino acids may confer increased in vivo stability on the fragment thus allowing a decrease in the effective amount of therapeutic fragment administered to a patient.

According to a further aspect the invention there is provided a therapeutic composition comprising fibrinogen E. In a preferred embodiment of the invention said therapeutic composition modulates angiogenesis. Preferably said modulation is the inhibition of angiogenesis. Preferably said inhibition relates to endothelial cell stimulated angiogenesis.

A number of conditions would benefit from an inhibition of angiogenesis. For example, ophthalmological conditions such as, neovascular glaucoma; diabetic retinopathy; age-related macular degeneration, pterygium ; retinopathy of prematurity; choroidal and other intraocular disorders. Also, atherosclerosis; haemangioma;

haemangioendothelioma; warts ; Kaposi's sarcoma; scar keloids; allergic oedema; dysfunctional uterine bleeding; follicular cysts ; ovarian hyperstimulation,; endometriosis; peritoneal sclerosis, adhesion formation; obesity; osteomyelitis; pannus growth; osteophyte formation ; inflammatory and infectious processes (eg hepatitis, pneumonia, glomerulonephritis), asthma, nasal polyps, transplantation, liver regeneration; thyroiditis, thyroid enlargement; and lymphoproliferative disorders.

Alternatively, or preferably, said inhibition is the inhibition of macrophage and/or tumour cell stimulated angiogenesis.

In a further preferred embodiment of the invention said inhibition is mediated by the inhibition of pro-angiogenic factors. Ideally these are either intracellular or cell surface receptors.

More preferably still, said inhibition is mediated via inhibition of the activity of pro- angiogenic growth factors. Ideally said growth factors are selected from: VEGF, bFGF ; aFGF ; TGFß ; PDGF.

According to a further aspect of the invention there is provided the use of fibrinogen E in the manufacture of a medicament for use in the treatment of cancer.

Polypeptides which comprise fibrinogen E can be manufactured by in vitro peptide synthesis using standard peptide synthesis techniques. Alternatively, or preferably, fibrinogen and/or polypeptides which comprise fibrinogen E can be manufactured by recombinant techniques which are well known in the art.

According to a further aspect of the invention there is provided a vector, wherein said vector includes a nucleic acid molecule which encodes for fibrinogen E for use in the recombinant manufacture of fibrinogen E.

Alternatively, vector (s) which include nucleic acid encoding polypeptides which comprise fibrinogen E can be engineered for recombinant expression.

In a preferred embodiment of the invention said vector is an expression vector adapted for prokaryotic or eulcaryotic cell expression.

Typically said adaptation includes, by example and not by way of limitation, the i provision of transcription control sequences (promoter sequences) which mediate cell/tissue specific expression. These promoter sequences may be cell/tissue specific, inducible or constitutive.

Promoter is an art recognised term and, for the sake of clarity, includes the following features which are provided by example only, and not by way of limitation. Enhancer elements are cis acting nucleic acid sequences often found 5'to the transcription initiation site of a gene (enhancers can also be found 3'to a gene sequence or even located in intronic sequences and is therefore position independent). Enhances function to increase the rate of transcription of the gene to which the enhancer is linked. Enhancer activity is responsive to trans acting transcription factors (polypeptides) which have been shown to bind specifically to enhancer elements. The binding/activity of transcription factors (please see Eukaryotic Transcription Factors, by David S Latchman, Academic Press Ltd, San Diego) is responsive to a number of environmental cues which include, by example and not by way of limitation, intermediary metabolites (eg glucose, lipids), environmental effectors (eg light, heat,).

Promoter elements also include so called TATA box and RNA polymerase initiation selection (RIS) sequences which function to select a site of transcription initiation.

These sequences also bind polypeptides which-function, inter alia, to facilitate transcription initiation selection by RNA polymerase.

Adaptations also include the provision of selectable markers and autonomous replication sequences which both facilitate the maintenance of said vector in either the eukaryotic cell or prokaryotic host. Vectors which are maintained autonomously are referred to as episomal vectors. Episomal vectors are desirable since these molecules can incorporate large DNA fragments (30-50kb DNA).

Episomal vectors of this type are described in W098/07876.

Adaptations which facilitate the expression of vector encoded genes include the provision of transcription termination/polyadenylation sequences. This also includes the provision of internal ribosome entry sites (IRES) which function to maximise expression of vector encoded genes arranged in bicistronic or multi-cistronic expression cassettes.

These adaptations are well known in the art. There is a significant amount of published literature with respect to expression vector construction and recombinant DNA techniques in general. Please see, Sambrook et al (1989) Molecular Cloning: A Laboratory Manual, Cold Spring Harbour Laboratory, Cold Spring Harbour, NY and references therein; Marston, F (1987) DNA Cloning Techniques: A Practical Approach Vol III IRL Press, Oxford UK; DNA Cloning: F M Ausubel et al, Current Protocols in Molecular Biology, John Wiley & Sons, Inc. (1994).

Alternatively, or preferably, fibrinogen for use in the manufacture of fibrinogen E is isolated from natural sources using standard protein purification techniques well known in the art. Additionally, fibrinogen can be isolated from animal sources, other than human, for example pig.

According to a further aspect of the invention there is provided a method for the production of fibrinogen E comprising: i) purifying fibrinogen from an animal ;

ii) incubating said fibrinogen polypeptide with an effective amount of a protease capable of cleaving fibrinogen to provide at least fibrinogen E; and iii) purifying fibrinogen E from fibrinogen and/or fibrinogen proteolytic fragments.

In a preferred method of the invention said fibrinogen is of human origin.

In a further preferred method of the invention said protease is plasmin.

According to a further aspect of the invention there is provided a method for the recombinant production of polypeptides comprising fibrinogen E including : i) providing a cell transformed/transfected with vector (s) including fibrinogen E nucleic acid; ii) providing conditions conducive to the manufacture of recombinant fibrinogen E polypeptides; and iii) purifying said fibrinogen E polypeptides from a cell, or a cells culture environment.

In a preferred method of the invention fibrinogen E polypeptides are provided with a signal sequence which facilitates the secretion of said polypeptides from said cell.

According to yet a further aspect of the invention there is provided a method to assemble fibrinogen E comprising: i) providing quantities of polypeptides which from fibrinogen E ; ii) providing conditions conducive to the assembly of at least fibrinogen E; and iii). purifying at least assembled fibrinogen E from the unassembled polypeptides.

According to yet still a further aspect of the invention there is provided a cell line transformed/transfected with at least one vector according to the invention wherein

said vector includes nucleic acid molecules which encode polypeptides which comprise fibrinogen and/or fibrinogen E.

It will be apparent to one skilled in the art that the recombinant production of fibrinogen E will be facilitated by providing cell expression systems adapted to produce and assemble fibrinogen and/or fibrinogen E.

According to yet still a further aspect of the invention there is provided a non-human, transgenic animal characterised in that said animal incorporates at least one fibrinogen gene into its genome wherein the expression of said fibrinogen transgene is facilitated.

It will be apparent to one skilled in the art that the provision of non-human transgenic animals genetically modified by the provision of a fibrinogen transgene (s) is an alternative source of fibrinogen. It is well known in the art that transgenic animals can be used to make various therapeutic polypeptides.

In a preferred embodiment of the invention said fibrinogen transgene is of human origin.

In a further aspect of the invention there is provided a method to treat an animal which would benefit from inhibition of angiogenesis comprising: i) administering an effective amount of an agent comprising fibrinogen E to an animal to be treated; ii). monitoring the effects of said fibrinogen E on the inhibition of angiogenesis.

In a preferred method of the invention said treatment is the inhibition of tumour development.

In an alternative method of treatment, polypeptides according to the invention are additionally conjugated, associated or crosslinked to an agent which augments the anti-angiogenic effect.

For example, a gene therapy vector includes a nucleic acid encoding a polypeptide according to the invention and a further nucleic acid encoding an anti-angiogenic agent.

Typically the agent could be a cytotoxic agent, another anti-angiogenic agent, a prodrug activating enzyme, a chemotherapeutic agent, a pro-coagulant agent or immunomodulatory factor. Examples of these are well known in the art, for example, and not by way of limitation cytotoxins, such as ricin A-chain or diphtheria toxin ; antagonists of the key pro-angiogenic factors in tumours (eg VEGF, bFGF, TNF alpha, PDGF) would include neutralising antibodies or receptors for these factors, or tyrosine kinase inhibitors for their receptors (eg. SU5416 for the VEGF receptor, Fllc- 1/KDR) ; prodrug activating enzymes such as, human simplex virus-thymidine kinase HSV-TK, which activates the prodrug, ganciclovir when it is then adminsitered sytemically ; chemotherapeutic agents, such as neocarzinostatin.

In addition, or alternatively, the cell surface domain of human tissue factor (this truncated form of tissue factor (tTF) could also be associated with polypeptides according to the invention. Truncated TF has limited anti-endothelial activity when free in the circulation, but becomes an effective and selective thrombogen (ie it causes extensive thrombosis and coagulation in blood vessels) when targeted to the surface of tumor endothelial cells.

An example of an immunomodulatory factor is the Fc effector domain of human IgGl. This binds natural killer (NK) cells and also the Clq protein that initiates the complement cascade. NK cells and complement then activate a powerful cytolytic response against the targeted endothelial cells.

It will be apparent that the above combinations of polypeptides and therapeutic agents will also have benefit with respect to the treatment of other conditions/diseases which are dependent on angiogenesis as hereindisclosed.

According to a yet further aspect of the invention there is provided an imaging agent comprising a polypeptide according to the invention.

It will be apparent to the skilled artisan that polypeptides according to the invention can be used to target imaging agents to, for example, tumours, to identify developing tumours or to monitor the effects of treatments to inhibit tumour growth. It will also be apparent that the combined therapeutic compositions which comprise both polypeptides according to the invention and a further anti-angiogenic agent may be further associated with an imaging agent to monitor the distribution of the combined therapeutic composition and/or to monitor the efficacy of said combined composition.

Methods used to detect imaging agents are well known in the art and include, by example and not by way of limitation, positron emission tomographic detection of FI8 and C11 compounds.

An embodiment of the invention will now be described, by example only, and with reference to the following figures: Figure 1 represents the amino acid sequences of the a-p-y-polypeptides of fibrinogen E; Figure 2 represents a schematic illustration of the role of the enzymes, plasmin and thrombin, in the generation of the fibrin (ogen) breakdown products. Fibrinogen consists of three pairs of polypeptide chains oc, ß and, joined by disulphide bonds to form a symmetric dimeric structure. The NH2-terminal regions of all six chains form the central E-domain. This fibrinogen molecule, when cleaved by plasmin, releases two D-fragments (the COOH-terminal regions), one E-fragment and several smaller

fragments including a small peptide, beta 1-42 (the amino terminal of the 8 chain).

Cleavage by thrombin releases the two fibrinopeptides A and B (Fp A and B) from the NH2-termini of a-and p-chains respectively, exposing polymerisation sites, which form electrostatic bonds between the E-domain of one molecule and the D-domain of an adjacent one resulting in lateral polymerisation of fibrin monomers into a fibrin polymer. Factor XIIIa, a transglutaminase, then introduces y-glutamyl-s-amino-lysine isopeptide cross-links between D-domains of adjacent fibrin polymers stabilising the polymer into crosslinked fibrin which is more resistant to cleavage. This can then be broken down by plasmin cleavage in the three stranded coils found between the D and E-domains yielding D-dimer, D-fragment and FnE-fragment (which lacks the fibrinopeptides A and B) and smaller fragments (4); Figure 3 represents mean ( SEM) number of human dermal microvascular endothelial cells (HuDMEC) migrating across a collagen-coated filter in response to control medium (no VEGF) or medium containing 1 Ong/ml VEGF in the absence or presence of various concentrations of FgE-fragment (A) or endostatin (B).

Representative data from 1 experiment are given as similar results were obtained in a further two identical experiments, and when VEGF was replaced by lOng/ml bFGF (data not shown). *P<0. 001 compared to positive control (VEGF alone); ^P<O. 01 compared to negative control (no VEGF) ; Figure 4 represents, upper panel (A): Tubule formation in the growth factor (GF)- reduced Matrigel assay (x 40 objective) in the absence of exogenous factors (I), or the presence of lOOnM FgE-fragment (I1), lOng/ml VEGF (III), or lOOnM endostatin (IV). Lower panels: mean ( SEM) area of tubule formation in the absence (empty bars) or presence of various concentrations of FgE-fragment or endostatin (shaded bars). HuDMECs were grown on GF-reduced Matrigel in DMEM + 1% FCS with either VEGF (lOng/ml) (Panel B) or bFGF (lOng/ml) (Panel C). Representative data from 1 experiment are given as essentially similar results were obtained in 3 identical experiments. *P<0.04 compared to control group. P< 0.02 compared to same dose of fibrinogen E ;

Figure 5 effects of various fibrinogen breakdown products ; fibrin E-fragment (FnE), and whole fibrinogen, on HuDMEC migration (panel A) or tubule formation in the GF-reduced Matrigel assay assessed as area (panel B) in the absence or presence of lOng/ml VEGF. Data are provided as means ( SEM) and all doses cited are in nM.

Representative data from 1 experiment are given as similar results were obtained in a further two identical experiments (and a further set of 3 experiments in which VEGF was replaced by lOng/ml bFGF). *P<0. 001 compared to respective group (ie. either with or without fibrinogen or FnE-fragment) with no VEGF, AP<O. 01 compared to respective group (ie. either with our without VEGF) with no fibrinogen or FnE- fragment ; Figure 6 represents the DNA and protein sequence of the a-ß-Y-chains of the fibrinogen E-fragment; and Figure 7 illustrates the in vivo efficacy of fibrinogen E-fragment on tumour growth in mice.

Materials and Metltods Cell culture. Adult human dermal microvascular endothelial cells (HuDMECs) were obtained commercially (TCS Biologicals, Buckinghamshire, United Kingdom) and cultured in microvascular endothelial cell growth medium (EGM) containing heparin (lOng/ml), hydrocortisone, human epidermal growth factor (lOng/ml), human fibroblast growth factor (lOng/ml) (such endothelial growth factors are necessary for routine passaging of HuDMECs in culture) and dibutyryl cyclic AMP. This was supplememented with 5% heat-inactivated PCS, 50u, g/ml gentamicin and 50ng/ml amphotericin B (TCS Biologicals, United Kingdom). Cells were grown at 37°C in a 100% humidified incubator with a gas phase of 5% CO and routinely screened for Mycoplasma. Prior to their use in the assays indicated below, HuDMECs were grown to 80% confluency, incubated in DMEM + 1% FCS for 2h, then harvested

with 0.05% trypsin solution, washed twice and resuspended at the required cell density.

Proteins and peptides.

Commercial human fibrinogen (plasminogen/plasmin and thrombin free) was obtained from Enzyme Research Laboratories (Swansea, United Kingdom). The fibrinogen did not clot at any point during the experiments indicating that there was no activity within the preparation to change its conformation. Human fibrinogen E- fragment was purchased from Diagnostica Stage, Asnieres, France (produced by plasmin cleavage of fibrinogen and purified by electrophoresis, immuno- electrophoresis, ion exchange and gel filtration). To generate human fibrin E- fragment, fibrinogen E-fragment was digested with human thrombin (Sigma-Aldrich Co, Dorset, United Kingdom), as previously described. To control for the possible effects of trace amounts of thrombin in the Fn E fragment preparation on our assays, the same amount of thrombin (0.5 U/ml) was added to control media used in experiments using Fn E-fragment. HPLC-purified FpA was obtained commercially from Bachem Ltd, Saffron Walden, UK. This peptide was included in the study as the amino termini of the two a fragments are retained in the Fgn E-fragment, but missing in the fibrin E-fragment (ie as the FpA portion of this is missing). We, therefore, compared the effects of equimolar amounts of FpA and Fgn E fragment in the assays described below to ascertain whether effects induced by Fgn E-fragment were due to an active site located in the FpA part of the molecule. Human recombinant endostatin was obtained from Calbiochem, La Jolla, CA.

Mi¢rntion assay The Boyden chamber technique was adapted from (13) and used to evaluate HuDMEC migration across a porous membrane towards a concentration gradient of either VEGF (lOng/ml) or bFGF (lOng/ml). The Neuro Probe 48 well microchemotaxis chamber (Neuro Probe Inc, Cabin John, MD) was used with 8Lm pore size polycarbonate membranes (Neuro Probe Inc, Cabin John, MD) coated with lOOlg/ml collagen type IV. lOng/ml VEGF or bFGE alone or with various concentrations of fibrinogen, fibrinogen E-fragment, fibrin E-fragment, or

fibrinopeptide A were dissolved in DMEM + 1% FCS and placed in the lower wells.

The collagen-coated membrane was then placed over this and 501l1 of 25x104 HuDMECs/ml (in DMEM containing 1% FCS) added to the upper chamber. The chambers were then incubated at 37°C for 4.5h. The chamber was then dismantled, the membrane removed and non-migrated cells scraped off the upper surface.

Migrated cells on the lower surface were fixed with methanol, stained with Hema 'Gurr'rapid staining kit (Merck, Leics, United Kingdom) and counted using a light microscope (x 160 magnification) in 3 random fields per well. Each test condition was carried out in 3-6 replicate wells and each experiment repeated 3 times.

Tubule formation assav.

24 well plates were coated with 301well of growth factor-reduced (GF-reduced) Matrigel (Becton Dickinson Labware, Bedford, MA). Endothelial cells plated on this matrix migrate and differentiate into tubules within 6h of plating as described previously (14). HuDMECs were seeded at a density of 4x104 cells/ml and incubated for 6h in 500gel of either DMEM + 1% FCS alone (control), or this medium + lOng/ml VEGF or bFGF in the presence or absence of whole fibrinogen or one of the fibrin (ogen) degradation products. Assessment of tubule formation involved fixing the cell preparation in 70% ethanol at 4°C for 15 minutes, rinsing in PBS and staining with haematoxylin and eosin. Three random fields of view in 3 replicate wells for each test condition were visualised under low power (x40 magnification), and colour images captured using a Fuji digital camera linked to a Pentium III computer (containing a frame grabber board). Tubule formation was assessed by counting the number of tubule branches and the total area covered by tubules in each field of view using image analysis software supplied by Scion Image.

Proliferation assay.

The MTT (3- [4, 5-Dimethylthiazol-2-yl]-2, 5-diphenyltetrazolium bromide) assay was used as previously described (12) to assess HuDMEC proliferation induced by

VEGF or bFGF in the absence or presence of fibrinogen or a fibrin (ogen) breakdown product. HuDMEC were seeded at 3x103 cells/100. 1 in DMEM + 1% FCS + lOng/ml VEGF or bFGF in test solution into 96 well microtitre plate for 4.5 and 6h.

At these time points, a quarter volume of MTT solution (2mg MTT/ml PBS) was added to each well and each plate was incubated for 4h at 37°C resulting in an insoluble purple formazan product. The medium was aspirated and the precipitates dissolved in 100tl DMSO buffered at pH 10.5. The absorbance was then read at 540nm on a Dynex ELISA plate reader.

Cytotoxicity Assay.

HuDMECs were seeded at a density of 1-2xl05 cells per well in a 24 well-plate in the absence or presence of fibrinogen or a fibrin (ogen) degradation product. After 6h, both live (following removal by trypsinisation) and dead (floating) cells were harvested and cell viability of all cells present assessed using propidium iodide staining of 5000 cells in each of triplicate samples per treatment using a FACScan (Becton Dickinson) equipped with a blue laser excitation of 15mW at 488nm. The data was collected and analysed using Cell Quest software (Becton Dickinson).

In Vivo Eflicccv of fibrihoe E fi'agmefat Animnls Experiments were performed on six-week-old Balb/C mice weighing 15g, obtained from Sheffield Field Laboratories. All experiments were approved by the Home Office Project Licence Number PPL50/1414.

Tumour Cell Culture The CT26 cell line was maintained by in vitro passage in Dulbecco's Minimal Eagles Medium containing 10% foetal calf serum, and 1%. penicillin and streptomycin and maintained at 37°C in humidified. atmosphere of 5% C02 in air. The cell line was

routinely checked to ensure freedom from mycoplasma (Mycoplasma rapid detection system, Gena-Probe Incorporated, U. S. A.) SubcutaneousTuniour Implantation Animals were anaesthetised with an intraperitoneal injection of diazepam (O. 5mg/ml, Dumex Ltd.) and hypnorm (fentanyl citrate 0.0315mg/ml and fluanisone lmg/ml, Janssen Pharmaceutical Ltd.) in the ratio of 1: 1 at a volume of O. lml/200g body weight, with supplementation as required to maintain adequate anaesthesia. Naive Balb/c mice were immunised s. c into the right flank, following removal of the fur.

Tumour cells were injected at a concentration of 3x105 viable CT26 cells per animal suspended in lOOul serum free medium. Animals were then allowed to recover.

Tumour growth and animal weights were monitored daily.

Admtnistration of Fibrinoven E Fra¢ment Tumour growth was measured daily and when the majority of animals in the cohort had tumour volumes of >100mm3 but <350mm3 animals were divided into experimental and control groups. This occurred between 14 and 18 days following implantation of the tumour cell suspension. Animals then received an intraperitoneal (ip) injection of either active drug (fibrinogen E fragment lOOMm) or vehicle (phophate buffered saline"ul). Daily injections continued'until the tumour growth. in the control animals reached the maximum burden allowed by Home Office legislation.

Assessment oltuniour growth Tumour volumes were assessed by claiper measurments of the perpendicular diameters and vulumes estimated using the equation:- Volume x b)/2 where a is the smaller and b the larger diameter Animals were weighed on a daily basis and the general well being monitored.

Statistical Analysis All experiments were performed at least three times and data analysed using the Mann-Whitney U test, a non-parametric test that does not assume a Gaussian distribution in the data being analysed. P<0. 05 was taken as significant.

Results and Discussion HuDMECs were seen to migrate across collagen-coated filters in the chemotaxis assay and form tubules on GF-reduced Matrigel in the absence of exogenous stimuli (although it should be noted that a residual level of growth factors is present even in GF-reduced Matrigel). Both cell activities were significantly (P<0. 001) increased in the presence of lOng/ml VEGF (Figures 3: A&B, 4A: photographs I & III, 5 : A).

Exposure to fibrinogen E-fragment (FgE-fragment) significantly (P<0. 001) inhibited both VEGF-induced migration (Figure 3A) and tubule formation, as assessed by either total tubule area or the number of branches of HuDMECs in a dose-dependent manner (Figures 4A: photograph II and 4B). None of the doses of FgE-fragment tested in this study altered cell migration in the absence of VEGF (Figure 3A). The inhibitory effects of FgE-fragment were not due to a cytostatic or cytotoxic effect of this molecule at 10 and lOOnM, as neither concentration had any notable effect on HuDMEC proliferation or viability (in control medium or medium containing lOng/ml VEGF; data not shown) in our assay systems. However, the marked decrease in HuDMEC migration and tubule formation evident at luM FgE-fragment may have been due, at least in part, to a cytotoxic effect, as this dose resulted. in a marginal but significant (P<0. 05) reduction in the viability and proliferation of HuDMECs (data not shown).

Essentially similar results were obtained in these studies when VEGF was replaced by lOng/ml bFGF (Figure 4C), indicating that FgE-fragment inhibits HuDMEC activity at a post-VEGF receptor locus common to both VEGF and bFGF signalling in endothelial cells. The putative receptor (s) that bind FgE-fragment on endothelial

cells have yet to be defined, although Dejana et al (13) indicated that FgE-fragment may be capable of binding the fibrinogen receptor in vitro. However, RGD motifs in the D-domains of the fibrinogen molecule mediate binding of this protein to the fibrinogen receptor (14). These sites are absent in FgE-fragment so binding to the fibrinogen receptor would involve a novel, non-RGD region of this fragment. It is not known whether this receptor is involved in the inhibitory effects of FgE-fragment demonstrated here and a distinct receptor/signalling pathway may be involved. It is to be noted that the 130kD endothelial cell receptor binding site B P 15-42 (Erban and Wagner J Biol Chem 267,2451-2458, 1992 ; Bach et al J Biol Chem 273,30719, 1998) is absent from the fibrinogen E-domain and therefore not directly involved in the anti-angiogenic effect.

It could be argued that the inhibitory effects of FgE-fragment are due to an indirect rather than a direct effect on endothelial cells, as there is no effect seen on non- stimulated endothelial cells. For example, fibrinogen has recently been shown to be capable of binding to such pro-angiogenic factors as bFGF (15), and could thereby block the pro-angiogenic function (s) of such cytokines. It is not known, however, whether fibrinogen can also bind VEGF or whether FgE-fragment, like its parent molecule, can bind either growth factor. It was possible that FgE-fragment may bind non-specifically to the filter in the chemotaxis assay and/or constituents of the Matrigel matrix in the tubule formation assay, thereby reducing endothelial cell adhesion and function. As one or both of these could, in theory, be responsible, wholly or in part, for the inhibition of HuDMEC migration and tubule formation by FgE-fragment recorded in this study, we repeated these studies and pre-exposed endothelial cells to FgE-fragment prior to their use in these assays, Exposure of HuDMECs to 10 and lOOnM FgE for lh wasssufficient to cause virtually the same level of inhibition in VEGF/bFGF-induced migration and tubule formation as that seen when FgE-fragment was present throughout the assay (data not shown).

In order to assess the anti-angiogenic potential of FgE-fragment, the level of endothelial cell inhibition was compared with that elicited by the well-characterised

anti-angiogenic agent, endostatin. Others have reported that 700ng/ml (35nM) endostatin is highly effective in. blocking angiogenesis in vitro (16), so various concentrations in this range were used in the present study. FgE-fragment produced similar or greater levels of inhibition than seen by any concentration of endostatin (Figures 3B and 4A: photograph IV and 4B&C). This finding suggests that, whatever the mechanism subserving its effect, FgE-fragment is a potent, new antagonist of angiogenic growth factors in vitro.

It may be important to note that the effects of FgE-fragment are not confined to endothelial cells. This polypeptide is known to also inhibit the migratory activity of neutrophils (17), stimulate fibrinogen release by hepatoctyes (18), and enhance the release of IL-6 by macrophages (19). Further studies are required to see whether these and possibly other effects of FgE-fragment, as yet undefined, will result in limiting side effects during or after its administration in vivo.

The anti-angiogenic effects of FgE-fragment contrast with results obtained using equimolar amounts of fibrinogen, fibrin E-fragment (FnE) and fibrinopeptide A.

Both fibrinogen and fibrin E-fragment (FnE) significantly (P<0. 001) increased control and VEGF-induced migration of HuDMECs at doses of IOOnM (Figure 5A).

Furthermore, both lOOnM fibrin E-fragment and lOOnM and lLM fibrinogen significantly (P<0. 05) enhanced tubule formation (Figure 5B). This accords well with previous reports showing that fibrinogen stimulates endothelial cell migration (13). Fibrin E-fragment has also been shown to be pro-angiogenic, possibly due to conformational changes induced within the fragment by thrombin cleavage of fibrinopeptide A. lOnM and lOOnM fibrin E-fragment also appeared to increase the proliferation rate of HuDMECs. However, as with fibrinogen E-fragment, the highest dose (luM) of fibrin E-fragment was cytotoxic for HuDMECs and triggered a significant (P<0. 001) decrease in cell viability and proliferation (data not shown).

This in turn caused marked reductions in HuDMEC migration and tubule formation in our assays systems (Figure 5A and B). Similar results were obtained when VEGF was replaced by 1 Ong/ml bFGF in these assays.

Fibrinogen E-and fibrin E-fragments differ in that the latter is denuded of fibrinopeptide A by thrombin cleavage. We, therefore, investigated whether the anti- angiogenic function of fibrinogen E-fragment resides in this part of the molecule by testing the effects of equimolar amounts of fibrinopeptide A alone on HuDMEC migration and tubule formation. This was not seen to exert a significant effect on such endothelial cell activities, suggesting that the active moiety resides in the central E domain of the molecule, or requires part or all of the rest of the domain, or resides in the FpA part of the amino terminus of the a chain but is only held in the correct conformation for biological activity when it is attached to the rest of the fragment.

We have also investigated the in vivo effect of the fibrinogen E fragment.

Example 1 This initial pilot study investigated the effects of fibrinogen E fragment (lOOmM) or vehicle administered daily i. p. for 10 days in two groups of mice (n=3). Starting tumour volume was less than 100mm3 in both groups. Tumours in the control group continued to grow at a steady rate over the ten day study period and reached a final tumour volume of 590 l 120 mm3 when the animals were killed at 10 days after commencing the injections. In contrast, tumours in the experimental group continued to grow at a similar rate to the control tumours until Day 5 (300mm3) when the growth rate stabilised for the remaining period of the study.

Example 2 The initial pilot study was repeated with an experimental group (n=8) receiving 100mM Fibrinogen E fragment ip and control group (n=7) receiving vehicle daily for 12 days. Starting tumour volume was less than 350mm3 in both groups. Tumours in the control group continued to grow steadily over the 12 day period reaching a final tumour volume of 3072 : L255 mm. In contrast tumours in the experimental animals

had a reduced but steady rate of growth with a final tumour volume of 2052 i 414mm3 (pO. OOl).

This data, see Figure 7, therefore demonstrates the potential of Fibrinogen E fragment as an in vivo anti-angiogenic agent.

References 1. Folkman J Angiogenesis in cancer, vascular, rheumatoid and other disease.

Nature Medicine, 1 : 27-31, 1995.

2. Leek R, Harris AL, and Lewis CE Cytokine networks in solid human tumours: regulation of angiogenesis. J. Leuk. Biol., 56 : 423-35,1994.

3. Cao Y Endogenous angiogenesis inhibitors: angiostatin, endostatin, and other proteolytic fragments. Prog Mol Subcell Biol., 20: 161-76,1998.

4. Doolittle R Fibrinogen and Fibrin. ScietAific American, 245 : 92-101,1981.

5. Costantini V, Zacharski LR, Memoli VA, Kisiel W, Kudryk BJ, and Rousseau SM Fibrinogen deposition without thrombin generation in primary human breast cancer tissue. Cancer Res., 51 : 349-53, 1991.

6. Dvorak HF, Nagy JA, Feng D, Brown LF, and Dvorak AM Vascular permeability factor/vascular endothelial growth factor and the significance of microvascular hyperpermeability in angiogenesis. Curr Top Microbiol Immunol, 237 : 97-132,1999.

7. Thompson WD, Wnag JEH, Wilson SJ, and Ganesalingham N Angiogenesis and fibrin degradation in human breast cancer. Angiogenesis : Molecular Biology, Clinical Aspects, 245-251,1994.

8. Thompson WD, Smith EB, Stirk CM, Marshall FI, Stout AJ, and Kocchar A Angiogenic activity of fibrin degradation products is located in fibrin fragment E.

J. Pathol, 168 : 47-53,1992.

9. Malinda KM, Ponce L, Kleinman HK, Shackelton LM, and Millis AJ Gp38k, a protein synthesized by vascular smooth muscle cells, stimulates directional migration of human umbilical vein endothelial cells. Exp Cell Res 250 : 168-73, 1999.

10. Shen J, Ham RG, Karmiol S Expression of adhesion molecules in cultured human pulmonary microvascular endothelial cells. Microvasc Res., 50: 360-72, 1995.

11. Liu J, Kolath J, Anderson J, Kolar C, Lawson TA, Talmadge J, and Gmeiner WH Positive interaction between 5-FU and FdUMP [10] in the inhibition of human colorectal tumour cell proliferation. Antisense Nucleic Acid Drug Dev., 9 (5): 481-6, 1999.

12. Dejano E, Languino LR, Polentarutti N, Balconi G, Ryckewaert JJ, Larrieu MJ, Donati MB, Mantovani A, and Marguerie G Interaction between fibrinogen and cultured endothelial cells. J. Clin. Invest., 75 : 11-18, 1985.

13. Suehiro K, Gailit J, Plow EF Fibrinogen is a ligand for integrin alpha5betal on endothelial cells. J. Biol. Chem., 272: 5360-6, 1997.

14. Sahni A, Odrljin T, and Francis CW Binding of basic fibroblast growth factor to fibrinogen and fibrin. J Biol Chem., 273: 7554-9,1998.

15. Taddei L, Chiarugi P, Brogelli L, Cirri P, Magnelli L, Raugei G, Ziche M, Granger HJ, Chiarugi V, and Ramponi G Inhibitory effect of full-length human endostatin on in vitro angiogenesis. Biochem Biophys Res Commun., 263 : 340-5, 1999.

16. Kazura JW, Wenger JD, Salata RA, Budzynski AZ, and Goldsmith GH Modulation of polymorphonuclear leukocyte microbicidal activity and oxidative

metabolism by fibrinogen degradation products D and E. J Clin Invest., 83 : 1916-24, 1989.

17. Qureshi GD, Guzelian PS, Vennart RM, and Evans HJ Stimulation of fibrinogen synthesis in cultured rat hepatocytes by fìbrinogen firagment E. Biochim Biophys Acta, 844 : 288-95,1985.

18. Lee ME, Rhee KJ, and Nham SU Fragment E derived from both fibrin and fibrinogen stimulates interleukin-6 production in rat peritoneal macrophages. Mol Cells, 9: 7-13,1999.