Priority This application claims priority of U.S. provisional application number 61/865,164 filed on August 13, 2013, the contents of which are fully incorporated herein by reference.

Sequence listing This application contains sequence data provided on computer readable format and a paper version. The paper version of the sequence data is identical to the data provided on computer readable format.

Field of the invention

The invention relates to pharmacologically active molecules for use in inhibiting angiogenesis. More specifically the invention relates to pharmacologically active SB 101 molecules and their use. Background of the invention

Angiogenesis is a fundamental step in development of larger tumors and in the transition of tumors from benign stage to malignant. For this reason inhibitors of angiogenesis have been considered to have a potential to constitute a treatment for cancer diseases, as well as for a number of other angiogenesis related diseases.

One target for a drug to inhibit angiogenesis has been CD44. CD44 is a cell surface receptor for the large glycosaminoglycan of the extracellular matrix hyaluronic acid (HA) and can also interact with other ligands, such as matrix metalloproteinases (MMPs). CD44 plays a role in various cellular and physiological functions, including lymphocyte activation, recirculation and homing, hematopoiesis, and tumor metastasis.

Various publications indicate that CD44 may inhibit tumor growth by binding to Hyaluronic acid: W099/45942 discloses the use of HA-binding proteins and peptides including CD44 to inhibit cancer and angiogenesis related diseases. Ahrens et al. (Oncogene 2001 20: 3399-3408) discloses that soluble CD44 inhibits melanoma tumor growth by blocking the cell surface CD44 binding to hyaluronic acid. Bajorath 2000 (Proteins: Structure, Function and Genetics 39: 103-111) discloses CD44 and its binding to HA, cell adhesion and CD44-signaling.

Bartolazzi et al. 1994 discloses an experiment where mammalian cell expressed CD44HRg-molecule inhibits tumor in nude mice, but a mutant molecule CD44-R41A-Rg, expressed in mammalian cells does not mediate attachment to HA and does not have similar inhibiting effect on the tumor.

US 8,192,744 discloses that non-glycosylated bacterial cell expressed CD44-3MUT does not have the HA-binding capacity but that it does inhibit angiogenesis and endothelial cell proliferation. It was also shown that CD44-3MUT does not bind to VEGF receptors in vitro, as does VEGF and therefore, the anti-angiogenic effect of CD44-3 MUT is most probably not due to orthosterical or allosterical inhibition of VEGF binding to its receptor.

As is shown in US 8,192,744 bacterially expressed CD44-3MUT inhibits angiogenesis and endothelial cell proliferation. However, this ~12 kD bacterially expressed recombinant protein displays high serum clearance rate which consequently affects its in vivo efficacy.

Accordingly, there is a need for an inhibitor of angiogenesis and endothelial cell proliferation with an improved serum half-life time.

I

Summary of the invention

It is an object of this invention to provide a pharmacologically active pegylated SB 101 -molecule (SB 101 is used herein as a synonym for CD44 -3MUT) having an amino acid sequence substantially according to SEQ ID NO: 2 or SEQ ID NO: 6 for in vivo inhibition of angiogenesis and endothelial cell proliferation, wherein the molecule has vimentin dependent activity in suppressing cell- cell adhesion.

It is another object of the invention to provide a pharmacologically active SB 101- fusion protein expressed in mammalian cells for in vivo inhibiting angiogenesis and endothelial cell proliferation, wherein the molecule has activity in suppressing cell- cell adhesion.

Another object of the invention is to provide modified SB 101 -molecules having extended in vivo plasma half-life times.

It is yet another object of the invention to provide a method to inhibit angiogenesis in vivo by administering SBIOI-Fc -fusion protein having amino acid sequence substantially according to SEQ ID NO: 4.

Another object of the invention is to provide a method to block endothelial cell response to growth factor -stimulation by administering SB 101- Fc fusion protein having an amino acid sequence substantially according to SEQ ID NO: 4. Still another object of the invention is to provide a method to down-regulate or inhibit

VEGFR2 by administering SBIOI-Fc fusion protein having an amino acid sequence of SEQ ID NO: 4.

A further object of the invention is to provide a method to inhibit angiogenesis in vivo by administering bacterial cell expressed pegylated SB 101 having an amino acid sequence of SEQ ID NO:2 or of SEQ ID NO:6.

A yet another object of the invention is to provide a method to inhibit tumor growth in a mammal in vivo by administering SBIOI-Fc fusion protein.

It is an object of this invention to provide a method to inhibit tumor growth in a human in vivo by administering SBIOI-Fc fusion protein.

Still another object of the invention is to provide a method to block endothelial cell response to VEGF- stimulation by administering bacterial cell expressed pegylated SB 101 having an amino acid sequence of SEQ ID NO:2 or of SEQ ID NO:6. Yet another object of the invention is to provide an expression vector comprising a nucleotide sequence substantially according of SEQ ID NO: 3 encoding SBIOI-Fc fusion protein, together with control elements enabling the expression of said nucleotide sequences in a host cell.

It is also an object of the invention to provide a method to inhibit growth of a tumor by introducing expression of nucleotide sequence according to SEQ ID NO: 3 in tumor cells.

A further object of the invention is to provide a method to inhibit growth of a tumor by introducing amino acid sequence according to SEQ ED NO: 2, SEQ ID NO:4 or SEQ ID NO:6 into tumor cells.

To increase serum residence time of CD44-3MUT two different approaches were used - in one case bacterially expressed CD44-3MUT was modified by PEGylation and alternatively, CD44-3MUT was expressed and purified in mammalian system as human IgGl Fc region fusion protein. Both approaches surprisingly resulted in biologically active protein with significantly increased in vivo serum half-life. Altogether, CD44-3MUT direct PEGylation or its expression as IgGl Fc fusion protein significantly improves its pharmacokinetic properties.

Brief description of the drawings

Figure 1 illustrates expression of SB 101 and purification of inclusion bodies (IB). (A) Sequence alignment of amino acid residues 27-132 of human CD44 (SEQ ID NO:5) and SB 101 (SEQ ID NO:6) proteins. The positions of mutated residues in SB 101 (R41A, R78S and Y79S) are indicated by upward arrows. Secondary structure elements are shown on top of sequence alignment and are based on crystal structure of CD44 HABD, 1UUH (Teriete, P.S. et al., 2004, Mol Cell 13: 483-96). Right arrows, beta-structures. Helices, alpha-helices and 3 ₁₀ -helix.

(B, left panel) SDS-PAGE analysis of SB 101 expression. SB 101 was transformed into E.coli BL21-CodonPlus(OE3)-RP strain and expressed in MAI for 22 h at 37°C. SB 101 (indicated by arrow) expression was estimated by analyzing samples from pre-culture and final culture (total lysate) on SDS-PAGE.

(B, right panel) SDS-PAGE analysis of the course of IB purification. SB 101 expressing cells were lysed and soluble fraction was separated from inclusion bodies (IB) by centrifugation. After washing (IB washes), IBs were solubilized in 8 M urea containing buffer and centrifuged to separate insoluble IB fraction. SB 101 containing soluble IB lysate was further used for SB 101 purification.

Figure 2 illustrates purification of SB 101. (A) Ion exchange chromatogram of urea dissolved inclusion bodies (IB). Soluble IB lysate was loaded onto HiPrepTM 16/10 DEAE FF 20 ml column and bound proteins were eluted by stepwise and linear gradient of NaCl. Eluates from FT and from first and second peak (PI and P2, respectively) were collected throughout IEC and analyzed on SDS-PAGE (B). Eluates containing monomeric SB 101 (shown by arrow) were pooled (IEC pool). (C) GF chromatogram of IEC purified SB101. IEC pool containing SB101 (load) was applied onto HiLoadTM 26/60 SuperdexTM 200 preparation grade column. Proteins eluted in 3 peaks (PI, P2 and P3). Fractions from P3 were analyzed on SDS-PAGE (D) and those containing monomeric SB 101 were pooled (GF pool) for refolding. (E) SDS-PAGE analysis of purified SB 101. SDS-PAGE analysis was performed under reducing (+DTT) and nonreducing (-DTT) conditions. (F) GF analysis of purified SB 101 (250 Superdex 200 10/300 GL preparation grade column). (G) MALDI- TOF mass spectrum of purified SB 101.

Figure 3 illustrates PEGylation of SB 101. (A) SDS-PAGE analysis of SB 101 PEGylation reaction. SB 101 was conjugated with 20 kD methoxy-PEG-propionaldehyde. PEG, SB 101 and PEGylation reaction mixture (mixture) were analysed by SDS-PAGE. (B) SDS-PAGE analysis of purified PEG-SB 101. PEG-SB 101 was purified from reaction mixture by IEC (Mono Q 4.6/100 PE column). SDS-PAGE gels were co-stained with Coomassie brilliant blue R-250 and PEG specific Bal ₂ (A and B left panels) or Coomassie brilliant blue alone (A and B right panels). (C) Western blot analysis (WB) of SB101 and PEG-SB 101 detected with anti-PEG (aPEG, left panel) or anti-SB 101 (right panel) antibodies. Unmodified and mono-PEGylated SB 101 are indicated by arrows. Figure 4 illustrates plasma half-life of SB 101. Rats were injected intravenously with PEG-SBlOl, SBlOl or GST-SB 101. Protein plasma levels were measured by ELISA assay (see Materials and methods), which was carried out in triplicates. Data are represented as mean ± SE. As non-PEGylated proteins display very short plasma half-life, these are also plotted on the graph with shorter time scale (insert).

Figure 5 illustrates comparison of SB lOl and PEG-SB 101 functional activity by endocytosis and cell layer impedance assay.

(A) Immunofluorescence images of anti-SB 101 mAb lA2-stained MLEC, which were incubated with SB lOl or PEG-SB 101 for 10 min at 37°C followed by 20 min chase in fresh media. Scale bars, 20 μιη.

(B) Quantification of cell-bound SBlOl. Boxplot represents log-transformed per-cell mean intensities. Filled circles represent suspected outliers or outliers, empty diamonds represent mean.

(C) Immunofluorescence images of anti-beta-catenin stained wild-type MLEC treated 18 h with SB lOl or left untreated. Scale bars, 20 μπι.

(D) Dose-response effect of SBlOl treatment to cellular beta-catenin in wild-type (squares) or vimentin knockout (circles) MLEC. Solid line represents dose-response curve fitted to wild-type MLEC data.

(E) Real-time measurement of resistance of wild type or vimentin deficient MLEC layers seeded to ECIS electrodes at confluence. Cells were treated with SB lOl or PEG- SB 101. Arrow indicates media change and treatment replenishment. Error bars, standard error. Figure 6 A) illustrates Western-blot analysis of SB IOI-Fc expression from

CHOEBNALT85 supernatant (pQMCF-5) 48 h after transfection. Lane 4, SB IOI-Fc reduced; lane 5, prestained protein ladder; lane 9, SBIOI-Fc non-reduced. B), illustrates coomassie blue stained SDS-PAGE analysis of purified SBIOI-Fc. Lane 1, SB IOI-Fc gel-filtration fraction 1 reduced; lane 2 SB IOI-Fc gel-filtration fraction 2 reduced; lane 3, pre stained protein ladder; lane 4. SB IOI-Fc gel-filtration fraction 1 non-reduced; lane 5, SB IOI-Fc gel- filtration fraction 2 non-reduced. 5 μg of purified protein was loaded per lane. Figure 7 illustrates plasma half-life of SBIOI-Fc. Rats were injected intravenously with SB IOI-Fc. Protein plasma levels were measured by ELISA assay. Data represent mean of triplicate wells ± 95% CI. Figure 8 illustrates inhibition of in vivo angiogenesis in mice by SBIOI-Fc intraperitoneal administration at doses 0.5 to 25 mg/kg. A) illustrates experimental scheme, where basement membrane extract filled angioreactors containing premixed FGF2, VEGF and heparin or PBS alone for uninduced controls were implanted SC into flanks of nude mice. Next day after implantation, mice started to receive every second day via IP injections of SB IOI-Fc, control human IgGl-Fc or vehicle (PBS). After 14 days, angioreactors were resected and the population of endothelial cells within the angioreactor matrix was assessed by FITC-lectin staining. The number of fluorescent cells was quantitated by micro plate reader. Background subtracted raw readings from each experiment were scaled by division with their root mean square. B) illustrates summary of directed in vivo angiogenesis assay results, bargraphs show mean ± SD. N, number of independent experiments. C) adjusted P- values (upper values) and effect sizes (lower values) of pairwise comparisons of treatment groups where N > 2. P values were calculated from post-hoc pairwise comparisons using t tests with pooled SD. P value adjustment method: fdr. Effect sizes were calculated by Cohen's d formula (95% confidence intervals). Confidence intervals were calculated using bootstrapping. GF, growth factors (FGF2/VEGF).

Figure 9. A) illustrates growth curves of 2 to 251 ng/ml VEGF stimulated HUVEC after release from quiescence.N = 3 independent experiments.

B) illustrates SBIOI-Fc blocking endothelial cell response of VEGF stimulation. N = 3 independent experiments. HUVEC were seeded onto 96WE1 ECIS arrays (Applied Biophysics, Troy, NY, USA). Next day, cells were changed into 0.5% FBS containing starvation media. After overnight serum starvation, cells were incubated 1 h with indicated amounts of SBIOI-Fc or rhlgGl-Fc and thereafter released from serum starvation by addition of 25 ng/ml VEGF in presence of recombinant protein. Cell growth was continuously monitored by measuring impedance at seven different frequencies from 1000 Hz to 64000 Hz using ECIS ΖΘ instrument (Applied BioPhysics). Shown is impedance data collected at Zlk, Z4k, Z16k and Z64k, impedance at 1000 Hz, 4000 Hz, 16000 Hz and 64000 Hz, respectively. Error bars, standard deviation. Figure 10 illustrates SBIOI-Fc causing inhibition of VEGFR2 Tyrosine-1175 phosphorylation. A) illustrates western blotting with anti-pVEGFR2(Tyrll75) and anti- VEGFR2 antibodies. B) illustrates densitometric quantitation of pVEGFR2(Tyrll75) and total VEGFR2 western blots from three or five experiments, respectively. HUVEC were seeded to 6-well culture plate at density 80000 cells/well. After 24 hours cells were starved in Ml 99 media supplemented with 1% FBS over-night. After starving cells were treated with hlgG-Fc or SBIOI-Fc (0-150 μ§/ιη1) in 10% FBS containing HUVEC growth media for 1 hour at 37 °C and thereafter stimulated with 8 ng/ml VEGF for 10 min at 37°C. After stimulation cells were lysed and analysed by western blotting.

Figure 11 Angiogenic response in mouse strains from different genetic backgrounds. Angiogenesis assay and quantitation was performed essentially as described in Example 7. Error bars, standard deviation.

Detailed description of the invention

It has been a general understanding that any potential use of CD44 for inhibition of angiogenesis is dependent on CD44's capability to bind hyaluronic acid. In US 8,192,744 it was first time shown however, that non-glycosylated bacterial cell expressed CD44-3MUT (also called SB-101) does not have the HA-binding capacity but that it does inhibit angiogenesis and endothelial cell proliferation. It was also shown that CD44-3MUT does not bind to VEGF receptors in vitro, as does VEGF and therefore, the anti-angiogenic effect of CD44-3MUT is most probably not due to orthosteric or allosteric interference with VEGF binding to its receptor.

In this disclosure, the concept of non-hyaluronic binding mutant CD44's capability to inhibit angiogenesis and endothelial cell proliferation is further developed. Novel pharmacologically active compounds to inhibit angiogenesis and endothelial cell proliferation are introduced.

SBlOl-protein (synonymous to CD443MUT) contains amino acid residues 27-132 of human CD44 and carry amino acid substitutions R41A, R78S and Y79S. The sequence of SB 101 (SEQ ID NO: 6) is shown in Figure 1A in comparison with the sequence of residues 27-132 of CD44 (SEQ ID NO: 5). A longer variation of the protein (SEQ ID NO: 2) where the N-terminus has an additional MQIDLNI-sequence was also used. Both of truncated (SEQ ID NO: 6) and the non-truncated version (SEQ ID NO:2) are referred here as SB 101 -protein. The triple mutation in SB 101 abolishes its property to bind hyaluronic acid (Bajorath, J., B. Greenfield, S. Munro, et al. ₍ 1998, J Biol Chem 273: 338-343; Pall T., Gad A., Kasak L., et al., 2004, Oncogene 23: 7874-7881).

The initial concept was to provide pharmacologically active modified SB 101 molecules. Recombinant bacterially expressed SB 101 exhibits very short plasma half-life in vivo. Thus, the goal for this study was to improve the pharmacokinetic properties of recombinant SB101. Furthermore, the goal was to introduce mammalian cell expressed SB 101 -protein derivatives with an ability to inhibit angiogenesis. Such molecules would be novel and non-obvious, because all previously published information teaches that glycosylated non HA-binding CD44 mutants would not have the effect of inhibiting angiogenesis.

Pegylation is a process of attaching the strands of the polymer PEG to peptides or proteins. Protein pegylation increases the molecular weight of the molecule and can give significant pharmacological advantages over unmodified forms, such as increased drug , stability and improved drug solubility. Additionally, pegylation masks protein surface, and reduces protease degradation and immunogenicity.

SB 101 was cloned and expressed in E.coli cells. The recombinant SB 101 isolated was necessarily in non-glycosylated form. As disclosed in US patent US 8,192,744 the recombinant SB 101 did not have the HA binding capability, but it still is capable of inhibiting angiogenesis.

Bacterial cell expressed SB 101 (SEQ ID NO: 2 or SEQ ID NO:6) was pegylated and unexpectedly it showed similar biological activity as the non-glycosylated non modified form. As is shown below in Example 4 the pegylated SB 101 was capable of inhibit endothelial cells similarly to unmodified SB 101. Analysis of the bacterial cell expressed pegylated SB 101 molecule indicated not only that the pegylated protein was biologically active, but it also showed that the pharmacokinetics of this molecule were superior to the unmodified SB 101: The half life time of the pegylated SB 101 is over 70 times longer than the half life time of the unmodified molecule. Comparing the molecular weights of the unmodified molecule and the pegylated molecule it was recognized that increase in molecular weight of -4.6 times lead to increase in the half life time over 70 times.

It was earlier shown by Bartolazzi (1994) that recombinant CD44 protein with mutated HA-binding site expressed in mammalian cells was not capable of inhibiting tumor growth. Based to the finding that pegylated bacterial expressed SB 101 having increased molecular weight was capable of inhibiting endothelial cells, increased molecular weight mammalian cell expressed SB 101 was produced by creating a fusion protein. Against the expectations based on previous teachings of mammalian cell expressed SB 101 being unable to inhibit angiogenesis, the product of this application inhibited angiogenesis in vivo. However, the increase in molecular weight of SB 101 fusion protein did not correlate to the same degree with the increased plasma half life time as we observed in case of pegylated SB 101. Moreover, GST-SB 101 fusion protein did not show longer half life time than unmodified SB 101 protein. SBIOI-Fc showed about 11 times higher half life time than the unmodified SB 101 several times smaller half life time than the pegylated bacterial expressed SB101.

In one aspect of the invention pegylated bacterial cell express SB 101 is provided with a half-life time approximately 70 time higher than of unmodified SB 101.

In another aspect of the invention a SB 101 -fusion protein is provided with at least 10 times higher half life time than of unmodified SB 101.

In one aspect the invention, modified pegylated bacterial cell expressed SB 101 protein is used to treat tumors, cancers and other angiogenesis dependent diseases.

In one aspect of the invention, mammalian cell expressed fusion protein SBIOI-Fc is used to treat tumors, cancers and other angiogenesis dependent diseases. In one aspect of the invention, the modified SB 101 proteins can be used in the treatment of solid tumors, brain cancer, breast cancer, colorectal cancer, kidney cancer, liver cancer, lung cancer, ovarian cancer, prostate cancer, bladder cancer, malignant lymphoma, and pancreatic cancer.

In one aspect of the invention the modified SB 101 proteins are administered intraperitoneally to a mammal at doses between 1 and 50 mg kg, preferably between 5 and 25 mg kg.

In one aspect of the invention the modified SB 101 proteins are administered several times per week.

Another aspect of the invention a pharmaceutical formulation comprising a pharmaceutically acceptable compound of the present invention, which provides, upon administration to a subject (e.g., a human), a decrease in tumor burden. The formulation may be administered in any suitable means including gene therapy.

Yet another aspect of the invention is a method of treating ovarian cancer in a subject (e.g., a human) in need thereof by administering to the subject an effective amount of the compound or the pharmaceutical formulation of the present invention.

Yet another aspect of the invention is a method of treating colon cancer in a subject (e.g., a human) in need thereof by administering to the subject an effective amount of the compound or the pharmaceutical formulation of the present invention.

Yet another aspect of the invention is a method of treating breast cancer in a subject (e.g., a human) in need thereof by administering to the subject an effective amount of the pharmaceutical formulation of the present invention.

Yet another aspect of the invention is a method of treating colon cancer before or after surgical resection and/or radiation therapy, in a subject (e.g., a human) in need thereof by administering to the subject an effective amount of the compound or the pharmaceutical formulation of the present invention.

Yet another aspect of the invention is a method of treating cancer before or after surgical resection and or radiation therapy, in a subject (e.g., a human) in need thereof by administering to the subject an effective amount of the compound or the pharmaceutical formulation of the present invention, including adjunctive therapy with one or more additional therapeutic agents, or their pharmaceutically acceptable salts. The compounds of the invention may also be administered sequentially with known anticancer or cytotoxic agents when a combination formulation is inappropriate. In any combination treatment, the invention is not limited in the sequence of administration.

The amount and frequency of administration of the compounds of the invention will be regulated according to the judgment of the attending clinician considering such factors as age, condition and size of the patient as well as severity of the symptoms being treated.

The invention is now described with non-limiting examples.

Example 1. Purification of SB 101

SB 101 was cloned into Ndel/Hindlll site of pETllc vector (Novagen, Darmstadt, Germany). SB 101 was codon optimized by ProteinExpert (SEQ ID NO: 1. Grenoble, France). For protein expression, SB 101 -vector was transformed into E.coli BL21-CodonPlus(DE3)- RP strain. For preculture, 3-4 colonies of transformants were inoculated into 20 ml of autoinduction media (MAI; Studier FW, 2005, Protein Expr Purif 41(1): 207-34) containing 100 μg ml ampicillin and 34 μg/ml chloramphenicol. The preculture was grown at 37° C overnight (ON) at 200 rpm until OD600 reached -10. For large scale protein production, 0.5 ml of starter culture was added to 500 ml of ampicillin-supplemented MAI in 2 L baffled Erlenmeyer flask and grown at 37° C for 20 h at 220 rpm. After 20 h cells were harvested by centrifugation for 10 min at 5000 rpm at 4° C. Cells were washed once with ice-cold PBS and harvested again at 5000 rpm for 15 min at 4° C. Cell pellet was weighted, snap frozen in liquid nitrogen and stored at -80° C until protein purification. Average wet bulk yield of bacteria was 9-12 g/L.

SB 101 was mostly expressed in insoluble protein fraction, as there was no SB 101 present in soluble fraction (Figure IB). Therefore, we purified SB 101 from inclusion bodies (IB).

For protein purification the ON thawed cell pellet was resuspended in Buffer A, 50 mM Tris-HCl, pH 8.0, 0.1% Triton X-100, 2 mM MgCl ₂ , 150 mM NaCl, and 1 mM PMSF, by using a Dounce hand homogenizer. The buffer volume was 5 ml/g cells. Cell suspension was supplemented with 100 μg/ml lysozyme (Sigma) and 10 μg/ml DNase I (Sigma) and incubated with continuous stirring for 2 h at 4° C. Cells were lysed using French press 3 times at 18000 psi. After lysis, one volume of Buffer A was added to the lysate, and incubated for 2 h at 4° C on end-over-end shaker. Inclusion bodies (IB) were isolated by centrifugation of the lysate at 46000xg for 30 min at 4° C. For removal of contaminating bacterial debris, IBs were washed 5 times. During each wash, IBs were fully homogenized in precooled wash buffers I- V by hand using Dounce homogenizer, incubated one hour at 4° C with continuous stirring and pelleted at 46000xg for 30 min at 4° C. Wash buffer I: 50 mM Tris-HCl pH 8.0, 2 M NaCl, 0.1% Triton X-100, 1 mM PMSF. Wash buffer II: 50 mM Tris-HCl pH 8.0, 150 mM NaCl, 0.1% Triton X-100, 1 M urea, 1 mM PMSF. Wash buffer III, IV and V: 50 mM Tris- HC1 pH 8.0, 150 mM NaCl, 1 M urea, 1 mM PMSF. The volume of IB wash and dissolving buffer was calculated as 12 ml per gram of starting cell mass. IBs were dissolved in Buffer B, 50 mM Tris-HCl, pH 8.0, 8 M urea, 25 mM ethylenediamine (EN), 1 mM PMSF, ON with continuous strirring at 4° C and insoluble debris was pelleted by centrifugation at 46000xg for 30 min at 4° C. After dissolving isolated IBs in 8 M urea-buffer and removal of insoluble material, soluble IB lysate contained considerable amount of SB 101 (Figure IB).

Cleared lysate was sterilized by filtration 0.22 μπι and left for 24 h at 4° C on an end- over-end shaker to equilibrate. Ion exchange chromatography (IEC) was applied as first step of the purification protocol. In order to prevent protein-carbamylation of urea dissolved IBs, buffer B was supplemented with ethylenediamine (EN). Urea dissolved IBs (up to 150 ml) were loaded onto a HiPrep™ 16/10 DEAE FF 20 ml column (GE Healthcare) equilibrated with five column volumes (CV) of Start Buffer 50 mM Tris-HCl pH 8.0, 8 M urea, 25mM EN and connected to AKTA HPLC system (AKTA Explorer 10, GE Healthcare) at flow rate 1.5 ml/min. The column was further washed with five CVs of Start Buffer and bound proteins were eluted at flow rate 4 ml/min by stepwise and linear gradient of NaCl 40 mM, 80 niM and 80-500 mM NaCl respectively in Elution Buffer 50 mM Tris-HCl pH 8.0, 8 M urea, 1 M NaCl and 25 mM EN. During EEC the flow through fractions were collected and analyzed by SDS-PAGE (Figure 2A). EN addition resulted in shift of SB 101 into flow through fraction (FT), as shown by SDS-PAGE analysis of eluted EEC fractions (Figure 2A and 2B). Most of the SB 101 protein was eluted in FT and peak 1 and peak 2 contained only little amount of SB 101. While SB 101 in FT appeared to be in its monomeric state, peak 1 and 2 contained mostly aggregated or oligomeric SB101 as assessed by SDS-PAGE analysis under nonreducing conditions (Figure 2B). For further purification we pooled SB 101 containing FT and termed it as EEC pool.

Fractions containing correct molecular weight protein were pooled for subsequent purification. To separate monomeric SB 101 from oligomeric forms and other remaining E. coli contaminants, gel filtration (GF) on a HiLoadTM 26/60 Superdex ^T ^ 200 preparation grade column (GE Healthcare) with AKTA HPLC system was used. The column was pre- equilibrated with three CVs of Start Buffer. 20-23 ml of IEC-purified protein solution was loaded onto GF-column and eluted at flow rate 2.5-3 ml/min. Through the run, fractions were collected and analyzed by SDS-PAGE. Peak 3 contained considerable amount of pure monomeric SB 101 as found by SDS-PAGE under reducing and non-reducing conditions (Figure 2C and 2D). Fractions containing purified monomeric SB 101 and used this pool in refolding steps were pooled. For refolding, monomeric protein containing fractions were pooled and diluted up to 0.1 mg/ml with GF buffer. Refolding was performed by stepwise dialysis from 8 M urea solution into PBS. Each dialysis step was performed twice, except last PBS step which was performed six times, for 1 h on a magnetic stirrer at 4°C. The sequentially used dialysis solutions were I: 6 M urea, 50 mM Tris-HCl pH 8.0, 18mM EN; II: 4 M urea, 50 mM Tris-HCl pH 8.0, 12 mM EN; III: 2 M urea, 50mM Tris-HCl pH 8.0, 150mM NaCl, 6mM EN; IV: 1 M urea, 50mM Tris-HCl pH 8.0, 150mM NaCl, 3mM EN; V: 0.5 M urea; 50mM Tris-HCl pH 8.0, 150mM NaCl, 1.5mM EN; VI: 50mM Tris-HCl pH 8.0, 150mM NaCl; VII PBS pH 7.4. All dialysis solutions were pre-cooled and pH was adjusted directly before use. Refolded protein solutions were centrifuged at 15000xg for 20 min at 4° C, sterilized using 0.22 μπι low protein-binding filter (Millipore) and concentrated using 10 kD MW cut-off centrifugal filter devices (Millipore). Analysis of refolded protein by SDS- PAGE showed that apparently all SB 101 was in monomeric form (Figure 2E). The monomeric form of SB 101 was further confirmed by GF analysis (Figure 2F). However, SDS-PAGE analysis under non-reducing conditions showed that SB 101 always migrated as two close bands, both at the approximate molecular size of monomeric SB 101 (Figures 2B, 2D and 2E). MALDI-TOF MS peptide fingerprinting showed that both of these bands contained SB 101. Therefore, it was assumed that these bands represent different conformations of SB 101. However, it was non possible to separate these putative differently folded forms neither by IEC nor GF. Calculated average mass of SB 101 is 11497.8 Da, assuming that the N-terminal methionine of the protein is excised. MALDI-TOF MS analysis of SB101 showed that the average molecular mass of the protein was 11500.4 Da (mass difference +2.6 Da; relative error 226 ppm; Figure 2G). To confirm the identity of purified protein, MALDI-TOF MS fingerprinting of Glu-C or trypsin digested SB 101 was performed. Fingerprinting results showed that Glu-C and trypsinolysis peptide spectra covered 100% of SB 101 sequence (Table 1). Glu-C digestion of SB 101 also confirmed the excision of N- terminal methionine.

Table 1. MALDI-TOF MS peptides of Glu-C or trypsin digested SBlOl

Enzyme/ Calculated Measured Difference Relative Peptide sequence SEQ ID Modificat peptide mass mass error, NO: _ion a residues ppm

Glu-C digestion

1-11 1322.631 1322.692 0.061 46 TCRFAGVFHVE 7 lxC

12-22 1225.617 1225.677 0.060 49 KNGAYSISRTE 8

23-41 2098.961 2099.024 0.063 30 AADLCKAFNSTLPTMAQME 9 lxC

42-49 864.483 864.528 0.045 52 KALSIGFE 10

50-57 900.377 900.402 0.025 28 TCSSGFIE 11 lxC

58-100 4825.277 NA GHVVIPPJHPNSICAANNTGVYILTYNTS 12 2xC

QYDTYCFNASAPPE

Trypsin digestion

4- 12 1033.547 1033.572 0.025 24 FAGVFHVEK 13

13-20 867.432 867.472 0.040 46 NGAYSISR 14

21-28 850.398 850.393 -0.005 -6 TEAADLCK 15

29-42 1568.745 1568.782 0.037 24 AFNSTLPTMAQMEK 16

43-64 2376.202 2376.238 0.036 15 ALSIGFETCSSGFIEGHVVIPR 17 l C

65-106 4704.048" 4704.829 0.780 166 IHPNSICAANNTGVYILTYNTSQYDTYCF 18 2xC

NASAPPEEDCTSV

65-106 4761.100 ^b 4761.538 0.438 92 IHPNSICAANNTGVYILTYNTSQYDTYCF 19 3xC

NASAPPEEDCTSV

, cysteine carbamidomethylation; average mass.

Example 2. PEGylation of SB101

Purified SB 101 of Example I (either according to SEQ ID NO:2 or SEQ ID NO:6) was conjugated with 20 kD methoxy-PEG-propionaldehyde (Jenkem Technology USA, Allen, TX, USA) in the presence of 200-fold molar excess of NaB¾CN as reducing agent. The reaction was performed in PBS pH 7.4 for 24 h. After 6 h and 22 h the reaction mixture was supplemented with additional dose of reducer in 100-fold and 50-fold molar excess, respectively. The reaction mixture was dialyzed against 10 mM phosphate buffer by three consecutive steps and then loaded onto a strong anion exchange column Mono Q 4.6/100 PE (GE Healthcare) equilibrated with lOmM phosphate buffer pH 7.4. After washing the column with 5 column volumes (CV) of lOmM phosphate buffer, PEGylated SB 101 was eluted by linear gradient of 0-1 M NaCl in 10 mM phosphate buffer. Reaction mixture was analysed by SDS-PAGE stained with Coomassie brilliant blue, or PEG-specific Bal ₂ (Figure 3A). Expected molecular weight of monoPEG-SB lOl was 31.6 kD. However, unexpectedly it was found that PEG and its conjugates moved about 18 kD higher (-50 kD) than expected on SDS-PAGE, as 20 kD PEG moved at -38 kD on gel. This could be due to enlarged hydrodynamic volume. It was concluded that -50 kD band represents mono-PEGylated form of SB 101 on gel. Analysis showed that preferentially mono-PEGylated SB 101 formed during reaction (Figure 3A). Nevertheless, also ligands with larger molecular weights were detected on the gel, which indicated that at some extent also di- and oligo-PEGylated products formed during reaction. Next, IEC was used with strong anion exchange column to isolate mono- PEGylated SB 101 from reaction mixture. Although mono-PEGylated SB 101 was sufficiently separated from oligo-PEGylated products by IEC, it was impossible to purify monoPEG- SB101 to homogeneity, as minor amounts of unmodified and di-PEGylated SB 101 co-eluted with the mono-PEGylated product in the same fraction (Figure 3B). The purity of PEGylated SB 101 was further analyzed by immunoblotting using PEG- and SBlOl-specific antibodies. Immunoblotting also detected that small quantities of unmodified and di-PEGylated SB 101 could be found in IEC purified PEG-SB 101 preparation (Figure 3C).

Example 3. Determination of SB101 and PEG-SB101 serum half life in rats

To measure SB 101 and SB101-PEG serum half-life 170-178 g female F344/NCrHsd rats (Harlan, Netherlands) were used. For blood sampling polyurethane round tipped jugular vein catheters were installed by Harlan Laboratories Surgical Services prior to shipment. Rats were housed in individual cages and had access to food and water ad libitum. Rats were given intraperitoneally anesthetic mix of ketamine 75 mg/kg (Farmak, Ukraine) with medetomidine 0.5 mg/kg (Domitor, Pfizer/Orion Pharma, Finland or Dorbene Vet, Pfizer/SYVA, Spain). 1 mg of SB 101 or SB101-PEG was administered via tail vein injection. Repeated blood samples were collected from rats using jugular vein catheter. Rats were held on pre heated gel pad under general anesthesia for short-time interval blood sampling. Anesthesia was reversed by 5 mg/kg atipamezole administered via subcutaneous injection (Antisedan, Pfizer/Orion Pharma, Finland) after 3 hours when point sampling was completed. Subsequent samples were collected under short- time anesthesia induced by diethyl ether inhalation or small doses of prescribed mix with quick recovery by atipamezole. Catheters were flushed between blood withdrawals with heparinized saline (Heparin sodium salt, Sigma- Aldrich, 25 IU/ml in 0.9% NaCl, B. Braun) and filled with heparinized glycerol (500 IU/ml) for >3 hour intervals between samplings. At the end of repeated sample collection animals were allowed to recover or sacrificed using carbon dioxide asphyxiation. Blood samples were held at +37°C for 30 min to allow clot formation and then centrifuged at 1300 g for 10 min. We also included GST-SB101 (Pall T., A. Gad, L. Kasak, et al., 2004; Oncogene 23:7874-81) to the study. GST-SB 101 was purified from soluble fraction of E.coli lysate. The plasma levels of test proteins were quantified by sandwich ELISA using anti-SB 101 mAb 1A2.H4 for GST-SB 101 and SB 101 capture and anti-PEG mAb for PEG-SB 101 capture. Anti-SB 101 pAb and anti- SB lOlmAb were used as detection antibodies, respectively. Vectastain ABC kit (Vector Laboratories, Burlingame, CA, USA) and tetramethylbenzidine substrate were used for colour development and absorbance was measured using ELISA plate reader (Tecan, Mannedorf, Swizerland). Plasma concentrations of tested proteins were interpolated from the standard curves generated by using serums with known concentrations of proteins. The obtained protein plasma concentration time curves are shown in Figure 4 (3D). Plasma concentration time-curves were fitted to exponential one-phase decay model using GraphPad Prism 5 software (version 5.02; Dec- 17, 2008; GraphPad Software, Inc.) and calculated pharmacokinetic properties are summarized in Table 2. Pharmacokinetic analysis showed that unmodified SB 101 proteins, regardless of their size and purification method, were rapidly cleared from rat circulation. In contrast, PEGylation of SB 101 significantly prolonged its half-life in rat circulation. PEG coupling to SB 101 extended its plasma half-life about 70- fold. In accordance with the increase in plasma residence time, PEGylation also considerably increased the systemic exposure (as measured by AUC) and respectively decreased the plasma clearance of SB 101 when compared to unmodified proteins. All tested proteins displayed similar volume of distribution values from 2.4-4.4% of total body weight, which indicates that all these proteins confined to plasma water. These data show that PEGylation clearly improves the pharmacokinetic properties of SB 101.

Example 4. PEG-SB 101 is endocytosed and suppresses cell growth comparably to

SB101 To assess PEGylated SB101 in vitro activity, we investigated its cellular uptake similarly to the method described in (Pall T., A. Pink, L. Kasak, et al., 2011; PLoS ONE 6: e29305). For internalization assay, mouse lung endothelial cells (MLEC) were seeded into 0.1% gelatin coated 8- well microscopy slides at density 20000 cells/well. MLECs were incubated 10 min with 0.11 μπιοΙ/L SB 101 or PEG-SB 101, followed by 20 min chase after changing to fresh media. Cell-bound SB101 was detected using 1A2.H4 mAb and anti- mouse-Alexa 488 conjugated secondary Ab (Molecular Probes). Hoechst staining was used to identify nuclei and anti-beta-catenin staining was used to identify cell borders and cytoplasm. For automated image analysis, confocal image stacks were converted to maximum intensity projections using Fiji package (http://pacific.mpi-cbg.de/wiki/index.php/Fiji). CellProfiler 2.0 (rl0415) software (Carpenter A.E., T.R. Jones, M.R. Lamprecht, et al., 2006; Genome Biol. 7: R100.) was used for image segmentation and quantization. Results show that both proteins were endocytosed (Figure 5A). Quantization of internalization results confirmed that PEG- SB 101 was endocytosed by MLEC similarly to SB 101 (Figure 5B). Immunofluorescence microscopy analysis showed that in endothelial cells 18 h incubation with SB 101 inhibited dose-dependently beta-catenin (Figure 5C and D). The effect is most probably mediated by vimentin, as vimentin deficient MLECs did not respond to the treatment. SB 101 inhibited beta-catenin in wild-type MLECs with logIC50 -6.8 mol L (95% CI -6.86 to -6.76, R square = 0.19, d.f. = 5227). Decreased beta-catenin levels in cell result in release of cell-cell contacts and inhibition of proliferation. These effects can be monitored by measuring of the cell-layer electrical resistance. Therefore, to test whether PEG-SB 101 has maintained its functional activity, we used impedance-based assay to measure its effect on endothelial cell growth. The measurement of cell layer resistance was performed using ECIS Z-theta and 96W10E+ arrays (Applied Biophysics, Troy, NY, USA). The arrays were cysteine pretreated according to manufacturer's instructions and coated with 1 μ^πιΐ fibronectin in PBS. Cells were seeded at 10000 cells/well, treatments were started 24 h later. Both, SB 101 or PEG- SB 101 treatment similarly reduced resistance of wild-type MLEC layer (Figure 5E, left panel). In agreement with beta-catenin immunofluorescence data, this function was lost in vimentin deficient cells (Figure 5E, right panel). Together, results of functional assays suggest that PEG-SB 101 has retained the same biological activity as SB 101.

Example 5. SBIOI-Fc fusion protein construction, expression and purification

DNA sequence encoding SB 101 was amended with sequence according to SEQ ID NO: 20 encoding N-terminal Cystatine-S signal peptide sequence (SEQ ID NO:21 ) and synthesized by Genewiz, Inc. The obtained synthetic DNA was inserted N-terminally into pBluescript vector containing human IgGl Fc region sequence (SEQ ID NO: 22) encoding IgGlFc region according to SEQ ID NO:23. The resulting SBIOI-Fc cDNA (SEQ ID NO 3) was cloned into pQMCF-5 expression vector containing RSV-LTR promoter (vector from Icosagen Cell Factory, Estonia). The resulting expression plasmid was transfected into CHOEBNALT85 cells. Protein expression was analyzed 48 h after transfection from cell media by Western Blot analysis (Figure 6A). Anti-human IgG-HRP conjugate (1:5000; LabAs, Estonia) was used for detection of protein expression. Expressed SB IOI-Fc protein (SEQ ID NO 4) was purified by Protein G sepharose, followed by Superdex 200 gel-filtration chromatography. Two different protein fractions were separated in gel filtration, collected and analyzed by coomassie blue stained SDS PAGE (Figure 6B). Concentration of purified protein was measured at A280 using extinction coefficient E=41.7 (Nanodrop, Thermo Scientific) and BCA Protein Analysis Kit (Pierce, Thermo Scientific). Endotoxin concentration of purified protein was measured using LAL Analysis Kit (Cambrex). Endotoxin concentration was 0.6-3 EU/mg.

Example 6. SBIOI-Fc serum half-life in rats The determination of SBIOI-Fc half-life was performed essentially as described in

Example 3 above for SB 101 and SB101-PEG, except that 250-300 g male F344/NCrHsd rats (Harlan, Netherlands) were used and 3 mg of SBIOI-Fc was administered via tail vein injection or into jugular vein catheter. Plasma concentration of SBIOI-Fc was measured using sandwich ELISA. 96- well plate (Maxisorp, Nunc) was coated with mouse anti-human IgGl antibody clone G17-1 (BD Biosciences) 4μg ml in 100 μΐ coating buffer (CB; 15mM Na ₂ C0 ₃ , 35mM NaHC0 ₃₎ pH 9) per well. Plates were blocked with 1.5% BSA in PBS 1 h at RT. SBIOI-Fc standards were serially diluted (40 μ^πιΐ - 0 μg/ml) into buffer containing 0.5% BSA and 2% rat serum in PBS. Serum samples collected from different time points were diluted 1:50 in 0.5% BSA in PBS. Biotin-conjugated mouse anti human IgG antibody clone G18-145 (BD Biosciences) 2 μ^πιΐ in 1% BSA in PBS was used for detection. After detection antibody incubation wells were washed 6 times with 200 μΐ 0.05% Tween-PBS and incubated with streptavidin-HRP (BD Biosciences) diluted 1: 1000 in 0.5% BSA in PBS 30 min at RT in dark. Incubation was terminated by washing wells 6 times with 200 μΐ 0.05% Tween-PBS. The colour reaction was developed with TMB solution (BD Biosciences) and stopped with 0.2 M H ₂ S0 ₄ . Optical density was measured by Tecan Microplate reader at 450 nm.

Plasma concentrations of SB IOI-Fc were interpolated from the standard curve generated by using rat serums with known concentrations of protein. The protein plasma concentration time curve obtained is shown in Figure 7. Plasma concentration time-curves were fitted to exponential one-phase decay model using GraphPad Prism 5 software (version 5.02; Dec- 17, 2008; GraphPad Software, Inc.) and calculated pharmacokinetic properties are summarized in Table 2.

Table 2. Pharmacokinetic properties of SB 101 proteins.

T

Protein Co g/ml) AUC ^g ^» h/ml) AUC (%ID'h/ml) CL (ml/h) V _d ( TBW) _1/2, (95% CI) _R 2 min

PEG-SB 101 133.7 812.5 81.3 1.2 4.3 171.2 (111 to 372) 0.8701

SB 101 237.4 13.7 1.4 73.0 2.4 2.4 (1.2 to 47) 0.9206

GST-

100.5 8.25 1.1 90.9 4.4 2.9 (2.2 to 4.1) 0.9942 SB 101

SBIOI-Fc 126 93.9 3.1 31.9 9.6 27.7 (22 to 36) 0.9932

Co is drug plasma concentration at time zero after iv administration. AUC is area under curve, expressed also as percent of initial dose (%ID). CL is clearance. Vd is volume of distribution, expressed as percent of total body weight (%TBW). T ₂ is half-life. CI is confidence interval. R2 is coefficient of determination.

Pharmacokinetic analysis showed that SBIOI-Fc exhibited intermediate plasma half- life value compared to SB101 and PEG-SB 101. This is surprising finding, given that SB101 Fc- fusion protein apparent molecular weight in its native dimeric form is ~ 130 kD (Figure 6), meaning that it's ~ 2.6 times higher than PEG-SB 101 (Figure 3B and C). Nevertheless, SBIOI-Fc plasma half-life showed at least 10-fold increase over SB 101 half-life. Surprisingly, SBIOI-Fc displayed approximately two times higher volume of distribution (Vd) compared to bacterially produced SB101 molecules. SBIOI-Fc Vd 9.6% (95% CI 9.1 - 10.2) of total body weight suggests that this protein binds also to erythrocytes in addition to plasma water residence. In accordance with the increase in plasma residence time, SBIOI-Fc showed increased systemic exposure (AUC) and respectively decreased plasma clearance compared to unmodified SB 101 (Table 2). IV- administration route data show that SBIOI-Fc displays clearly prolonged plasma half-life, but surprisingly less than expected based on its molecular size. SBIOI-Fc displays also an increased apparent volume of distribution.

Example 7. SBIOI-Fc inhibits in vivo angiogenesis In order to test in vivo anti- angiogenic effect of SB 101 and SBIOI-Fc, Directed in

Vivo Angiogenesis Assay kit (DIVAA™; Trevigen, USA) was used according to manufacturer's instructions. In this assay, angiogenesis is measured by blood vessel invasion into tumor extracellular matrix filled angioreactors. Blood vessel in growth is stimulated by combination of two tumor-related angiogenic factors bFGF and VEGF premixed into the angioreactor's matrix. For angiogenesis assay, 20 μΐ angioreactors (semi closed silicone cylinders) were filled with growth factor reduced basement membrane extract (BME) containing premixed bFGF, VEGF and heparin for the induction of angiogenic response. For uninduced controls, BME containing equal volume of PBS was used for filling angioreactors. Final concentrations of bFGF and VEGF in angioreactor's matrix were 1.4η§/μ1 and 0.47η^μ1, respectively. Angioreactors were implanted sc into dorsolateral flank of nine week old Hsd:Athymic Nude-Foxnl/nu female or male mice (Harlan, Netherlands) through 0.5-1 cm skin incisions. Implantation was performed on both flanks and two angioreactors were inserted per flank. Incisions were closed with absorbable 6-0 sutures (Monosyn; B Braun, Germany) and 7 mm skin clip (Autoclip; BD Diagnostics, USA). The surgical procedure was carried out under general anesthesia. Mice were given ip anesthetic mix of ketamine 75 mg/kg (Ketamin; Farmak, Ukraine) with medetomidine 1 mg/kg (Dorbene Vet; Pfizer/SYVA, Spain). Anesthesia was then reversed by 1 mg/kg atipamezole subcutaneous injection (Antisedan; Pfizer/Orion Pharma, Finland). All steps of angioreactor filling and implantation were performed under aseptic conditions and in accordance with Trevigen DIVAA™ protocol. Mice were also given 5 mg/kg of ketoprofen sc injection for postoperative pain relief (Anafen; Merial, France).

Mice were treated with SB IOI-Fc or vehicle (PBS) during two weeks starting from the next day post implantation (Figure 8A). In this experiment we used the protein according to SEQ ID NO:5 to exemplify the effect, but minor modifications on the protein sequence are in the scope of this invention. Preferably the SB IOI-Fc protein has a sequence at least 80% similar to SEQ ED NO:5 and more preferably about 95% similarity with SEQ ID NO:5. Drugs were administered by intraperitoneal injections every second day. At day 14 mice were sacrificed by carbon dioxide asphyxiation and angioreactors were dissected. Angioreactor contents were retrieved and processed for quantization of endothelial cell invasion according to manufacturer's protocol. BME was digested with metalloproteinase-based CellSperse™ reagent, endothelial cell receptors were then recovered by incubating cells in 10% FBS containing DMEM, then washed several times with wash buffer, repeatedly resuspending cells. Recovered cells were pelleted by centrifugation at 250 x g for 5 min at RT. FITC- Lectin staining was used for fluorometric quantitation of angioreactor-invaded endothelial cells. Cell-bound fluorescence was read at 485 nm excitation and 535 nm emission wavelengths using Tecan microtiter plate reader. Results show that in this model FGF2/VEGF stimulation caused average 5.1 ± 3.4-fold induction of endothelial migration into angioreactors when comparing PBS treated control groups, fdr adjusted t- test p = 8.1 x 10-6 (Figure 8B and 8C). We found that SB IOI-Fc intraperitoneal treatments inhibited FGF2/VEGF induced angiogenesis. In a treatment group receiving intraperitoneally SMUT- Fc at 25 mg/kg angiogenesis was inhibited close to unstimulated basal level compared to PBS treatment, p = 1.1 x 10-4, effect size 4.38 (95% CI 3.07 to 10.37). When considering SB 101- Fc doses where we performed more than one experiment, we found that 0.5 and 5 mg/kg SB IOI-Fc administration resulted in angiogenesis inhibition, but response was less robust, with apparent inhibition in 3 experiments out of 4. Due to variation in 0.5 and 5 mg/kg dosing groups, SB IOI-Fc treatments don't differentiate significantly - 25 mg/kg dosing shows medium sized effect compared to 0.5 or 5 mg/kg, but lower boundary of 95% CI-s of effect size is in both cases close to zero (Figure 8C). We used irrelevant rhlgG mAb or rhlgG-Fc as control treatments. Both these molecules were purified alike to SBIOI-Fc. The 0.5 mg/kg SB IOI-Fc treatment showed significant inhibitory effect compared to pooled 0.5 mg/kg rhlgG/rhlgG-Fc control treatments, p = 0.034, effect size 1.54 (95% CI 0.58 to 8.25). In groups receiving intraperitoneally 5 or 15 mg/ml doses of rhlgG-Fc angiogenesis response remained in the same boundaries as PBS and 0.5 mg/ml rhlgG control treatments.

It was tested whether SB lOl-Fc-mediated angiogenesis inhibition could be achieved by continous delivery via micro osmotic pumps. Alzet 100 μΐ osmotic pumps, filled in sterile conditions with SBIOI-Fc (0.3 mg/lOO μΐ) or rhlgG-Fc (0.3 mg/lOO μΐ) were implanted through a midline abdominal incision into intraperitoneal cavity of mouse, anesthetized with a combination of ketamine (Bioketan, 75 mg/kg) and medetomide (Dorbene Vet, 1 mg/kg). Wound was closed in two layers using 6-0 monofilament absorbable sutures (Monosyn;

Braun/Aesculap, Tuttlingen, Germany) and a single wound clip (7 mm Autoclips; BD, Sparks, MD). Anaesthesia was reversed with atipamezole (Antisedan 5 mg/kg,

subcutaneously), mice received also subcutaneously analgesic ketoprofen (Ketofen, 5 mg/kg). Results show that angiogenesis was inhibited in mice carrying SBlOl-Fc-filled pumps compared to rhlgG-Fc control treatment (Figure 8B).

Example 8. SBIOI-Fc blocks endothelial cell response to growth factor stimulation

Real-time endothelial growth in impedance based assay was measured to test whether SBIOI-Fc can interfere with mitogenic stimulation of endothelial cells. Growing cells cover the surface of gold electrodes of the measurement arrays resulting in increased electrical impedance (i.e. complex resistance) which can be continuously monitored. The measurement of cell layer resistance was performed using ECIS ΖΘ and 96WE1_PET plates (Applied Biophysics, Troy, NY, USA). Before use 96-well plate was pretreated with 10 mM cysteine (200 μΐ/well) for 12 min at RT, washed 2 times with PBS (175 μΐ/well) and thereafter coated with 0.1% gelatin in 150 mM NaCl (200 μΐ/well) for 1 hour at RT. HUVEC were seeded at density 5000 cells/well. After 30 hours cells were switched to starvation media— M199 supplemented with 1% FBS, 25 mM Hepes and 4 mM L-glutamine for 16 hours. After starving, cells were treated with different concentrations of hlgG-Fc or SBIOI-Fc in 155 μΐ of 5% FBS containing HUVEC growth media (5% FBS, M199, 4 mM L-glutamine, 12.5 μ^πύ heparin, 10 mM Hepes, 7.5 μg/ml ECGS) for 1 hour at 37°C. Thereafter the cells were stimulated with 25 ng/ml of VEGF. For that 20 μΐ of 8.75 x growth factor solution in 5% FBS containing HUVEC growth media was added to cells, so that final media volume was 175 μΐ/well. The growth of stimulated HUVEC was further monitored in ECIS for 72 hours at 37°C. We characterized growth of quiescent HUVEC in response to stimulation with different concentrations of VEGF (2 to 251 ng/ml). We observed three different in vitro growth pattern, 2 and 8 ng/ml VEGF induced growth but were not sufficient to sustain long-time survival, 25 ng/ml VEGF provided long-time support (72+ hours) for growth and survival. Whereas, very high VEGF concentrations (79 and 251 ng/ml) resulted in suppressed maximum density, but long-time survival (Figure 9A). To test SBIOI-Fc effect on HUVEC growth we used 25 ng/ml VEGF, as this concentration induced reproducibly robust long-time response. We found that treatment of quiescent HUVEC with 8 to 16 μΜ SB IOI-Fc doses resulted in significant inhibition of growth in response to stimulation with 25 ng/ml VEGF (Figure 9B). Example 9. SBIOI-Fc inhibits VEGFR2 level

Based on the findings that SB 101 binds to cell surface vimentin and that SB 101 is endocytosed and it inhibits cells to VEGF stimulation it is suggested that SB 101 may affect cell growth factor receptor level or activation. To test this hypothesis, we assessed the effect of SB 101-Fc-treatment on VEGF receptor 2 protein level and activation.

HUVEC cells (passage 4-6, Cell Applications, Inc.) were cultured in 6-well cell culture plates in 20% FBS containing M199 medium supplemented with 4 mM L-glutamine, 50 μg/ml heparin, 10 mM Hepes, 30 μg/ml ECGS (Millipore). Cells were seeded at density 80000 cells/well. After 24 hours cells were starved in M199 media supplemented with 1% FBS, 25 mM Hepes and 4 mM L-glutamine for overnight. After starving cells were treated with different concentrations of hlgG-Fc or SBIOI-Fc (0-150 μg/ml) in 10% FBS containing HUVEC growth media for 1 hour at 37°C and thereafter stimulated respectively with 8 ng/ml VEGF-165 (Serotec) for 10 min at 37°C. For subsequent western blot analysis, cells were lysed in 60-80 μΐ of lx Laemmli's sample buffer. The proteins (20-30 μΐ of sample) were separated on 7.5-10% SDS-PAGE gradient gel and transferred to PDVF membrane at 350 mA 1.5 hours. The membranes were blocked in 5% whey in 0.1% Tween20-TBS (TBST) at RT for 1 hour. The membranes were then incubated with primary antibodies in 2% whey- TBST (5% BSA-TBST for phospho-antibodies) overnight at 4°C followed by appopriate HRP- conjugated secondary Ab (Jackson Immunoresearch) incubation at 1/10000 dilution for 1 hour at RT. After every Ab incubation step the membranes were washed 4 x 10 min with TBST at RT. For reprobing, the membranes were incubated in stripping buffer (62.5 mM Tris pH 6.8, 2% SDS, 100 mM 2-mercaptoefhanol) for 30 min at 50°C. Results show that 1 hour treatment of endothelial cells prior VEGF stimulation with SB IOI-Fc causes dose dependently inhibition of VEGFR2 activation at phospho-Tyrosine 1175 (Figure 10).

Example 10. Testing PEG-SB 101 effect on in vivo angiogenesis

In order to test in vivo anti- angiogenic effect of PEG-SB 101 we use Directed in Vivo Angiogenesis Assay kit (DIVAA™; Trevigen, USA) according to manufacturer's instructions. In this assay, angiogenesis is measured by blood vessel invasion into tumor extracellular matrix filled angioreactors. Blood vessel in growth is stimulated by combination of two tumor-related angiogenic factors bFGF and VEGF premixed into the angioreactor's matrix. For angiogenesis assay, 20 μΐ angioreactors (semi closed silicone cylinders) are filled with growth factor reduced basement membrane extract (BME) containing premixed bFGF, VEGF and heparin for the induction of angiogenic response. For uninduced controls, BME containing equal volume of PBS was used for filling angioreactors. Final concentrations of bFGF and VEGF in angioreactor's matrix are 1.4ng^l and 0.47ng^l, respectively. Angioreactors are implanted sc into dorsolateral flank of nine week old Hsd:Athymic Nude- Foxnl/nu female or male mice (Harlan, Netherlands). Implantation is performed on both flanks and two angioreactors are inserted per flank. Mice are treated with PEG-SB 101 at doses 5 and 25 mg/kg or vehicle (PBS) during two weeks starting from the next day post implantation. Drugs are administered by intraperitoneal injections every second day. At day 14 mice are sacrificed and angioreactors are dissected. Angioreactor contents are retrieved and processed for quantization of endothelial cell invasion according to manufacturer's protocol. FITC -Lectin staining is used for fluorometric quantization of angioreactor-invaded endothelial cells. Cell-bound fluorescence is read at 485 nm excitation and 535 nm emission wavelengths using Tecan microtiter plate reader. PEG-SB 101 effect on angiogenesis by intraperitoneal administration at tested doses will be compared to vehicle treated controls and mean treatment effects are subjected to statistical analysis.

Example 11. Testing PEG-SB101 effect on endothelial cell response to VEGF stimulation

Real-time monitoring of endothelial growth in impedance based assay is used to test whether PEG-SB 101 can interfere with mitogenic stimulation of endothelial cells. Growing cells cover the surface of gold electrodes of the measurement arrays resulting in increased electrical impedance (i.e. complex resistance). HUVEC are seeded onto 8WE1 ECIS arrays (Applied BioPhysics, Troy, NY, USA). Next day, cells are changed into 0.5% FBS containing starvation media. After overnight serum starvation, cells are supplemented with 10% FBS and PEG-SB 101 or vehicle (PBS) alone. After 1 hour PEG-SB 101 and vehicle (PBS) treated cells are further stimulated by supplementing media with 5 ng/ml VEGF. Cell growth is continuously monitored over next 24-48 hours by measuring impedance at 16000 Hz using ECIS ΖΘ instrument (Applied BioPhysics). VEGF mitogenic stimulation causes increase in cells impedance. PEG-SB 101 treatment is expected to block such VEGF effect similarly to SB101 or SB IOI-Fc.

Example 12. Testing SB101, SBIOI-Fc or PEG-SB101 effect on endothelial cell response to growth factor stimulation and growth factor receptor levels

Real-time monitoring of endothelial growth in impedance based assay is used to test whether SB 101, SBIOI-Fc or PEG-SB101 can interfere with stimulation of endothelial cells with growth factors. SB 101, SBIOI-Fc or PEG-SB 101 interferes with hepatocyte growth factor (HGF/SF) stimulation of endothelial cells. HGF/SF stimulates endothelial cell growth, motility and angiogenesis by binding to the c-Met receptor and may confer resistance to anti VEGF/VEGFR pathway cancer therapies (Shunli Ding, Tatyana Merkulova-Rainon, Zhong Chao Han and Gerard Tobelem, BLOOD, 2003 101: 4816-4822; Shojaei F, Lee JH, Simmons BH, Wong A, Esparza CO, Plumlee PA, Feng J, Stewart AE, Hu-Lowe DD, Christensen JG, CANCER RES., 2010, 70: 10090-100). In impedance assay, growing cells cover the surface of gold electrodes of the measurement arrays resulting in increased electrical impedance (i.e. complex resistance). HUVEC are seeded onto 8WE1 ECIS arrays (Applied BioPhysics, Troy, NY, USA). Next day, cells are changed into 0.5% FBS containing starvation media. After overnight serum starvation, cells are supplemented with 10% FBS and SB lOl, PEG-SBlOl, SBIOI-Fc or vehicle (PBS) alone. After 1 hour SB lOl, PEG-SB lOl, SBIOI-Fc and vehicle (PBS) treated cells are further stimulated by supplementing media with 25 ng/ml HGF. Cell growth is continuously monitored over next 24-48 hours by measuring impedance at 100- 64000 Hz using ECIS ΖΘ instrument (Applied BioPhysics). HGF stimulation causes changes in cells impedance. SB lOl, PEG-SBlOl, SB IOI-Fc treatments is expected to block such HGF effects compared to control (vehicle or irrelevant protein) treatments.

Real-time monitoring of endothelial growth in impedance based assay is used to test whether SBlOl, SB IOI-Fc or PEG-SBlOl can interfere with growth factor stimulation of endothelial cells. Growth factors stimulate endothelial cell growth, motility and angiogenesis by binding to their receptors. Such growth factors involved in angiogenesis include TGF- alpha and FGF-2 and other factors which bind to their receptors In impedance assay, growing cells cover the surface of gold electrodes of the measurement arrays resulting in increased electrical impedance (i.e. complex resistance). HUVEC are seeded onto 8WE1 ECIS arrays (Applied BioPhysics, Troy, NY, USA). Next day, cells are changed into 0.5% FBS containing starvation media. After overnight serum starvation, cells are supplemented with 10% FBS and SBlOl, PEG-SB lOl, SB IOI-Fc or vehicle (PBS) alone. After 1 hour SB lOl, PEG-SB lOl, SBIOI-Fc and vehicle (PBS) treated cells are further stimulated by supplementing media with selected growth factor. Cell growth is continuously monitored over next 24-48 hours by measuring impedance at 100-64000 Hz using ECIS ΖΘ instrument (Applied BioPhysics). Growth factor stimulation causes changes in cells impedance. SBlOl, PEG-SB lOl, SB IOI-Fc treatments is expected to block such growth factor effects compared to control (vehicle or irrelevant protein) treatments.

The endothelial cells are treated 1 hour with SB IOI-Fc, SBlOl or PEG-SB 101 prior growth factor stimulation. For stimulation TGF-alpha and FGF-2 are used. Growth factor stimulation causes rapid endocytosis of growth factor receptors. Endocytosed receptor is either recycled back to the plasma membrane or targeted to lysosomes for degradation. After stimulation, respective growth factor receptor protein levels are measured using western blotting. SBlOl, PEG-SBlOl, SB IOI-Fc treatments is expected to reduce growth factor receptor levels in endothelial cells in this assay.

Example 13 CD44 deficiency associates with increased angiogenic response Directed in vivo angiogenesis assay was used to test FGF2/VEGF- induced angiogenic response in CD44 knockouts and wild type mice from C57BL/6, C3H or mixed genetic backgrounds (Figure 11). We found a significant 5.2 ± 2-fold angiogenesis response to FGF2 VEGF in CD44 deficient mice from mixed background, p = 0.0052, N = 2 experiments, effect size 13.5 (95% CI 13.5 to Inf). Using pooled data including all wild type strains showed apparent but much less robust angiogenesis induction . when CD44 was present, p = 0.41, N = 3, effect size 0.755 (95% CI 0.084 to 4.61). Importantly, blood vessel invasion into angioreactors in response to FGF2/VEGF was significantly lower in wild type mice than in CD44 knockouts (p = 0.0018), whereas unstimulated baseline values were not significantly different. However, athymic nude mice displayed angiogenesis induction comparable to CD44 knockouts (Figure 8B) and it has been shown that genetic heterogeneity contributes to sensitivity for angiogenic stimulation in mouse strains (Richard M. Rohan, Antonio Fernandez, Taturo Udagawa, Jenny Yuan, and Robert J. D'amato , FASEB J. 2000 14:871- 876 ) suggesting that also other factors than CD44 may affect angiogenesis induction in this assay. Nevertheless, our data suggest that CD44 deficiency associates with increased angiogenic response and therefore CD44 may function as endogenous inhibitor of angiogenesis.