VEGF DIMER MOLECULES AND COLUMNS COMPRISING A VEGF DIMER MOLECULE AS WELL AS USES, PRODUCTION METHODS AND METHODS INVOLVING THE SAME

Abstract

The present invention relates to a column comprising a vascular endothelial growth factor (VEGF) dimer molecule, a method for preparing such a column, a VEGF dimer molecule, an expression vector and a recombinant host cell encoding for a VEGF dimer, as well as uses and methods related thereto.

Claims

1. A column comprising a vascular endothelial growth factor (VEGF) dimer molecule comprising a first and a second VEGF molecule.

2. A VEGF dimer molecule comprising a first and a second VEGF molecule, wherein the first and/or the second VEGF molecule have a length of 122 to 250 amino acids.

3. The column according to claim 1 or the VEGF dimer molecule according to claim 2, wherein the VEGF dimer molecule is expressed by a eukaryotic cell.

4. The column according to any one of claims 1 or 3 or the VEGF dimer molecule according to claims 2 to 3, wherein the first and the second VEGF molecule are identical.

5. The column according to any one of claims 1 or 3 to 4 or the VEGF dimer molecule according to any one of claims 2 to 4, wherein the first and/or the second VEGF molecule lack the N-terminal signal peptide.

6. The column according to any one of claims 1 or 3 to 5 or the VEGF dimer molecule according to any one of claims 2 to 5, wherein the first and/or the second VEGF molecule have a length of 122 to 250 amino acids, preferably 125 to 225 amino acids, more preferably 140 to 200 amino acids, even more preferably 150 to 180 amino acids, especially more preferably 160 to 170 amino acids and most preferably 165 amino acids.

7. The column according to any one of claims 1 or 3 to 6 or the VEGF dimer molecule according to any one of claims 2 to 6, wherein the first VEGF molecule has at least 50% amino acid sequence identity in comparison to SEQ ID NO: 3, preferably at least 60% amino acid sequence identity in comparison to SEQ ID NO: 3, more preferably at least 80% amino acid sequence identity in comparison to SEQ ID NO: 3 and most preferably at least 90% amino acid sequence identity in comparison to SEQ ID NO: 3 and/or, wherein the second VEGF molecule has at least 50% amino acid sequence identity in comparison to SEQ ID NO: 4, preferably at least 60% amino acid sequence identity in comparison to SEQ ID NO: 4, more preferably at least 80% amino acid sequence identity in comparison to SEQ ID NO: 4 and most preferably at least 90% amino acid sequence identity in comparison to SEQ ID NO: 4.

8. The column according to any one of claims 1 or 3 to 7 or the VEGF dimer molecule according to any one of claims 2 to 7, wherein the first and the second VEGF molecule are linked by a linker.

9. The column according to any one of claim 8 or the VEGF dimer molecule according to claim 8, wherein the linker has a length of 10 to 30 amino acids, preferably a length of 11 to 25 amino acids, more preferably a length of 12 to 20 amino acids, especially more preferably a length of 13 to 17 amino acids and most preferably a length of 14 amino acids.

10. The column according to any one of claims 1 or 3 to 9 or the VEGF dimer molecule according to any one of claims 2 to 9, wherein the VEGF dimer molecule is immobilized by a covalent bond to a matrix.

11. A method for preparing a column according to any one of claims 1 or 3 to 10 comprising the steps of: a) preparing by eukaryotic expression a vascular endothelial growth factor (VEGF) dimer molecule comprising a first and a second VEGF molecule; and b) immobilizing the dimer of step a) on a matrix.

12. A method for preparing a column according to claim 11, wherein in step b) the VEGF dimer molecule is immobilized to the matrix by a covalent bond.

13. The VEGF dimer molecule according to any one of claims 2 to 10 for use in the treatment of preeclampsia, characterized in that the VEGF dimer molecule is bound to a column for apheresis.

14. An expression vector comprising a nucleic acid sequence encoding the VEGF dimer molecule according to any one of claims 2 to 10.

15. A recombinant host cell line comprising the VEGF dimer molecule according to any one of claims 2 to 10, comprising the expression vector according to claim 15 and/or comprising a nucleic acid sequence encoding the VEGF dimer molecule according to any one of claims 2 to 10.

16. A method for separating sFlt-1 from blood and/or releasing VEGF from complexes with sFlt-1 into blood comprising incubating the column according to any one of claims 1 or 3 to 9 or the VEGF dimer molecule according to any one of claims 2 to 9 with the blood and separating sFlt-1 from the blood and/or releasing VEGF from complexes with sFlt-1 into the blood.

17. Use of the column according to any one of claims 1 or 3 to 9 or the VEGF dimer molecule according to any one of claims 2 to 9 for separating sFlt-1 from blood and/or releasing VEGF from complexes with sFlt-1 into blood.

Description

DESCRIPTION OF THE FIGURES

[0146] FIGS. 1: Molecular modelling of scVEGF.sup.165 and negative staining electron microscopy. A) Schematic representation of the scVEGF.sup.165 expression plasmid and potential modes of assembly into multimers bridged by intermolecular disulfide-bonds. B) Band and ribbon representation of two single chain VEGF.sup.165-dimers assembling as tetrameric VEGF. The 14 amino acid linker is represented by red spheres. C) Visualization of purified recombinant moVEGF.sup.165 and scVEGF.sup.165 molecular structure by negative staining electron microscopy. moVEGF.sup.165 molecules appear as dumbbell-shaped dimeric structures with a few monomeric dots. scVEGF.sup.165 emerged as tetrameric structures and a few dumbbell-shaped single scVEGF.sup.165 molecules. In the expanded image the contour of the respective molecule was outlined by automated image processing.

[0147] FIG. 2: Binding characteristics of sFlt-1 capture molecules. A) Quantitative ELISA using VEGF-trap. Plot of absorbance 450 nm versus VEGF-trap concentration on a semi logarithmic scale with [VEGF-trap] on the logarithmic scale. The curve defines the equilibrium dissociation constant (Kd). The lowest Kd belongs to scVEGF.sup.165, indicating strongest interaction. Each value depicts calculated means of experimental triplicates (n=3). B) Calculated Kd values from three independent experimental replicates were summarized in order to compare the binding affinity between different VEGF/PIGF variants. scVEGF.sup.165 reproducibly yields lowest Kd values, i.e. highest binding affinity. Plotted are means with SD from 3 independent experiments. *p< 0.02.

[0148] FIGS. 3: Generation of sFlt-1 capturing apheresis columns and evaluation of substrate matrices. A) Identification of ideal apheresis setup. Equal amounts of recombinant moPIGF, moVEGF.sup.165, and scVEGF.sup.165 were coupled to strepTactin XT, Cyanogen bromide activated sepharose, or amino-linked agarose resin. The binding affinity to sFlt-1 at a concentration of 1600 pg/ml in human serum was measured. sFlt-1 concentration in the flow through was determined by quantitative sFlt-1 ELISA. For all three coupled proteins, the measurements depicted the most significant sFlt-1 depletion when using the amino-linked agarose system. B) Longitudinal stability of the scVEGF.sup.165 agarose column was assessed. Adsorption studies from human serum samples were performed at 1, 15, 30, 60, and 90 days. Within 90 days only a slight decrease in binding sFlt-1 reduction was noted.

[0149] FIGS. 4: Characterization of sFlt-1 capturing molecules employing human serum samples A) Concentration of sFlt-1 (pg/ml) in the sample (y-axis) after apheresis for all immobilized VEGF and PIGF variants as well as with the Flt-1 specific antibody (x-axis). B) Concentration of PIGF (pg/ml) in the sample (y-axis) after apheresis for all immobilized VEGF and PIGF variants as well as with the Flt-1 specific antibody (x-axis). C) Concentration of VEGF (pg/ml) in the sample (y-axis) after apheresis for all immobilized VEGF and PIGF variants as well as with the Flt-1 specific antibody (x-axis)..

[0150] FIGS. 5: Validation of sFlt-1 clearance and VEGF-/PIGF-release of scVEGF.sup.165 columns. Serum samples of ten individual patients (x-axis) with preeclampsia were treated with scVEGF.sup.165 apheresis. sFlt-1 (A), free PIGF (B), and VEGF (C) levels (y-axis) were measured before and after adsorption.

[0151] FIG. 6: Characteristics of different VEGF- and PIGF-dependent sFlt-1 apheresis strategies and as compared to the antibody-based setup. Apheresis systems employing Flt-1 specific antibody reduce sFlt-1 levels but do not liberate endogenous PIGF or VEGF (far left panel). Adsorption columns containing PIGF retain sFlt-1 and liberate low quantities of endogenous PIGF but no endogenous VEGF. Monomeric VEGF.sup.165 columns capture sFlt-1 and release PIGF and also VEGF. Enhanced binding characteristics of scVEGF.sup.165 for sFlt-1 result in most efficient adsorption of sFlt-1 and competitive release of PIGF and VEGF (far right panel).

[0152] FIG. 7: Purification of recombinant proteins shown by Coomassie stained sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Coomassie staining of the supernatant of the stably transfected cells expressing the indicated proteins under induced expression upon doxycycline treatment. Initial medium containing the proteins was referred to as input. While recombinant proteins were not detected in the wash as well as flow-through, a clear purification profile of the proteins could be detected in the elution fraction.

[0153] FIG. 8: Stability of immobilized VEGF/PIGF variants on agarose apheresis columns. Strep-tag specific antibody was employed to precipitate recombinant strep-tagged proteins from flow through of serum-treated scVEGF.sup.165 columns. Precipitates were subjected to SDS-PAGE and immunoblot using strep-specific antibody. No strep-tagged protein was detected in the serum flow-through, indicating that no recombinant protein disengages from the column under apheresis conditions.

[0154] FIG. 9: Evaluation of the functionality of recombinant sFlt-1, VEGF, and PIGF variants by SDS-PAGE followed by western blot. From top to bottom: specific antibody for phosphorylated ERK1/2, specific antibody for ERK1/2, specific antibody for glyceraldehyde 3-phosphate dehydrogenase (GAPDH). MAP-activation was detected by immunoblot analysis with a specific antibody for phosphorylated ERK1/ 2; equal loading and cell confluency was controlled by immunoblot for total ERK1/2 and GAPDH.

EXAMPLES

Material and Methods

Cloning, Recombinant Protein Expression and Purification

[0155] For cloning scVEGF.sup.165 and scPIGF, uncloned, double-stranded linear DNA fragments containing represented nucleotide sequence were customized and ordered in the form of GeneArt String DNA fragment (ThermoFisher scientific, Germany). The designed sequence bears two identical amino acid sequence of monomeric VEGF.sup.165 except the second monomer lacking the signal peptide sequence. Two monomers are being connected to each other using a linker with amino acid sequence of GSTSGSGKSSEGKG (showed as underlined letters below). DNA fragments were amplified using primers (see Table 1) and cloned (using restriction enzymes Nhel and BamHI) into the sleeping beauty transposon expression vector, carrying an N-terminal strep tag. Table 1: Primer list used for cloning scVEGF.sup.165 and scPIGF into expression vector.

TABLE-US-00001 Primer type Sequence SEQ ID NO: scVEGF.sup.165 forward 5′-ACAGCTAGCGCTCCTATGGCTGAAGGCGG-3′ 10 scVEGF.sup.165 reverse 5′-TGTGGATCCCCGTCTGGGCTTATCGCAGC-3′ 11 scPIGF forward 5′-ACAGCTAGCCTGCCTGCTGTTCCTCCTC-3′ 12 scPIGF reverse 5′-TGTGGATCCCTCTACGAGGCACGGCGTCG-3′ 13

[0156] The GeneArt String DNA fragment sequences of scVEGF.sup.165 and scPIGF are as follows (linker sequence underlined):

[0157] GeneArt String DNA fragment sequence of scVEGF.sup.165 (SEQ ID NO: 7):

TABLE-US-00002 GCTAGCGCTCCTATGGCTGAAGGCGGAGGACAGAATCACCACGAGGTGGTCAAGTTCATGGACGTGTAC CAGCGGAGCTACTGTCACCCCATCGAGACACTGGTGGACATCTTCCAAGAGTACCCCGACGAGATCGAG TACATCTTCAAGCCCAGCTGCGTGCCCCTGATGAGATGTGGCGGCTGCTGCAATGACGAAGGCCTGGAA TGTGTGCCCACCGAGGAATCCAACATCACCATGCAGATCATGCGGATCAAGCCCCACCAGGGCCAGCAT ATCGGCGAGATGTCTTTCCTGCAGCACAACAAGTGCGAGTGCAGACCCAAGAAGGACCGGGCCAGACAA GAGAATCCTTGCGGCCCTTGCAGCGAGCGGAGAAAGCACCTGTTTGTGCAGGACCCTCAGACCTGCAAG TGCTCCTGCAAGAACACCGACAGCCGGTGCAAAGCCAGACAGCTGGAACTGAACGAGCGGACCTGCAGA TGCGACAAGCCTAGAAGAGGCAGCACAAGCGGCAGCGGCAAAAGCTCTGAAGGCAAGGGAACGCGTGCC CCAATGGCAGAAGGTGGCGGCCAGAACCACCATGAGGTCGTGAAGTTTATGGATGTCTATCAGCGGTCC TACTGCCATCCTATCGAAACCCTGGTCGATATTTTTCAAGAGTATCCGGATGAGATTGAGTATATTTTC AAACCCTCCTGTGTGCCGCTCATGCGCTGCGGCGGATGCTGTAATGATGAGGGACTTGAGTGCGTGCCA ACCGAAGAGTCTAATATTACGATGCAGATTATGAGAATCAAACCGCATCAAGGGCAGCATATTGGGGAA ATGAGCTTCCTCCAGCATAACAAATGTGAATGCCGGCCGAAGAAGGACAGAGCCCGGCAAGAAAACCCA TGCGGCCCCTGTTCCGAGAGGCGGAAACATCTGTTCGTTCAAGATCCCCAGACCTGTAAATGTAGCTGT AAAAACACCGACTCCAGGTGCAAGGCCCGGCAACTCGAGCTGAACGAGAGAACATGTCGCTGCGATAAG CCCAGACGGGGATCCACA

[0158] GeneArt String DNA fragment sequence of scPIGF (SEQ ID NO: 8):

TABLE-US-00003 acagctagcCTGCCTGCTGTTCCTCCTCAACAATGGGCCCTGTCTGCCGGCAATGGCAGCTCTGAAGTT GAGGTGGTGCCCTTCCAAGAAGTGTGGGGCAGAAGCTACTGCAGAGCCCTGGAAAGACTGGTGGACGTG GTGTCTGAGTACCCCAGCGAGGTGGAACACATGTTCAGCCCTAGCTGCGTGTCCCTGCTGAGATGCACA GGCTGTTGCGGCGACGAGAATCTGCACTGCGTGCCAGTGGAAACCGCCAACGTGACAATGCAGCTGCTG AAAATCAGAAGCGGCGACAGACCCAGCTACGTGGAACTGACCTTCAGCCAGCACGTCCGCTGCGAGTGT AGACCCCTGCGGGAAAAGATGAAGCCCGAGAGATGCGGAGATGCCGTGCCTAGAAGAGGCAGCACATCT GGCTCTGGCAAGAGCAGCGAAGGCAAGGGACTTCCTGCTGTGCCACCACAGCAGTGGGCACTGAGTGCC GGAAATGGCTCCTCTGAGGTGGAAGTCGTGCCTTTTCAAGAAGTCTGGGGACGCTCCTACTGTCGCGCT CTTGAGAGACTGGTCGATGTCGTCAGCGAGTACCCCTCCGAAGTCGAGCACATGTTTTCCCCATCCTGT GTGTCTCTGCTGCGGTGTACCGGATGCTGCGGGGATGAGAACCTGCATTGTGTGCCTGTCGAGACAGCC AATGTCACCATGCAGCTCCTCAAGATCAGATCCGGCGATCGGCCCTCCTACGTCGAGCTGACATTTTCT CAGCACGTTCGATGCGAATGCCGGCCTCTGCGCGAGAAAATGAAGCCTGAACGCTGTGGCGACGCCGTG CCTCGTAGAggatccaca

[0159] Protein sequence scPIGF (linker underlined) (SEQ ID NO: 9):

TABLE-US-00004 TASLPAVPPQQWALSAGNGSSEVEWPFQEVWGRSYCRALERLVDWSEYPSEVEHMFSPSCVSLLRCT GCCGDENLHCVPVETANVTMQLLKIRSGDRPSYVELTFSQHVRCECRPLREKMKPERCGDAVPRRGSTS GSGKSSEGKGLPAVPPQQWALSAGNGSSEVEVVPFQEVWGRSYCRALERLVDVVSEYPSEVEHMFSPSC VSLLRCTGCCGDENLHCVPVETANVTMQLLKIRSGDRPSYVELTFSQHVRCECRPLREKMKPERCGDAV PRRGST

[0160] For the production of recombinant protein stable HEK293T Epstein-Barr virus nuclear antigen (EBNA) cell lines were generated employing the sleeping beauty transposon system and the protein was purified as previously described (Agarwal et al., 2012). Briefly, the vectors were transfected into the HEK293T EBNA cells using FuGENE® HD transfection reagent (Promega GmbH, Madison, USA) for the duration of 3 days with fetal bovine serum (FBS)- free medium. Supernatant of the cells was harvested on the following days, filtered and the proteins carrying the Strep tag were purified via Strep-Tactin®XT (IBA Lifesci-ence, Göttingen, Germany) resin at room temperature (RT). The proteins were then eluted by Biotin containing Tris-buffered saline (TBS)- buffer (IBA Lifescience, Göttingen, Germany), and the aliquots were stored at -80° C.

Negative Stain Electron Microscopy

[0161] The structure of monomeric and dimeric VEGF was visualized by negative staining electron microscopy as described previously (Bober et al., 2010). Briefly, samples (usual concentrations 10 - 20 nM) were incubated on carbon-coated grids for 1 min, washed and then stained with 0.75% uranyl formate for 1 min. The grids were rendered hydrophilic by glow discharge at low pressure in air. Specimens were examined in a Philips/FEI CM 100 TWIN transmission electron microscope operated at 60 kV accelerating voltage. Images were recorded with a side-mounted Olympus Veleta camera with a resolution of 2048 x 2048 pixels (2 k x 2 K) and the ITEM acquisitions software.

Human Umbilical Vein Endothelial Cell (HUVEC) Cell Culture and Mitogen Treatment

[0162] HUVEC passage 3 were cultured on collagen II coated 6 well plate with Endopan 3 kit medium containing 5% FBS, hydrocortisone, VEGF, hFGF-B, R3-IGF-1, ascorbic acid, hEGF, GA-1000 (Pan biotech, Bavaria, Germany) until they reached 80% confluency. The cells were starved with 3% FBS and no additive growth factor supplement for 24 hours. On the next day, the cells were stimulated with the recombinant proteins as indicated. After incubation for 10 minutes at 37° C., the cells were directly lysed on the plate using 30 .Math.l 2x Laemmli buffer, and loaded on a sodium dodecyl sulfate (SDS) polyacrylamide gel, followed by western blot.

Solid Phase Enzyme-Linked Immunosorbent Assay (ELISA) Style Assay

[0163] ELISA style binding assays were performed as described previously (Agarwal et al., 2012). In brief, 10 .Math.g/ml of recombinant VEGF, PIGF and their mutant variants were individually immobilized together with bovine serum albumin (BSA) on 96 well plate (Transparente Immuno Standardmodule, Thermo scientific, Denmark), while matrix loaded only with BSA served as a control. A binding saturation curve was generated by applying increasing concentrations of VEGFR1/R2 trap ranging from 0,1 to 750 nM, followed by detection of the ligand-receptor complex using a horseradish peroxidase (HRP)-conjugated antibody against the mouse Fc domain of VEGFR1/R2 trap. Luminescence was quantified at optical density (OD) 450 nm using a by spectrophotometer (Thermo ScientificMultiskan GO, Finland).

ELISA Measurements

[0164] Human sFlt-1, free VEGF, and PIGF levels were quantified using specific commercial ELISA kits (R&D Systems, Minneapolis, USA). The assays were performed as specified by the manufacturer. Fluorescence was quantified using a spectrophotometer (Thermo ScientificMul-tiskan GO, Finland). Data was analyzed with GraphPad prism 7 software (GraphPad Soft-ware Inc., La Jolla, CA, USA).

Generation of Specific VEGF/PIGF Columns

[0165] The immobilization of the recombinant protein on Strep-Tactin®XT matrix (IBA Lifescience, Göt-tingen, Germany) was performed on Polyprep® chromatography columns (Bio-Rad laboratories, USA) according to the manufacturers’ instructions.

[0166] The recombinant proteins containing a Strep tag, were incubated with the recommended amount of resin in TRIS 50 mM, NaCl 150 mM, pH=8 overnight. For immobilization of recombinant protein on Cyanogen bromide activated resin (Merck, Germany) on Polyprep® chromatography columns (Bio-Rad laboratories, USA), the proteins where dialyzed in coupling buffer containing 100 mM NaHCO.sub.3, 500 mM NaCl, pH=8.3 for 2 days. The beads were activated with 30 ml cold 1 mM HCl for 15 min. The recommended amount of resin was incubated with the dialyzed proteins overnight at 4° C. The unbound surface of the beads was blocked by Tris-HCl 0.1 M for 2 hours at RT.

[0167] For the immobilization of recombinant protein on an agarose matrix (AminoLink™ Plus Immobilization Kit, Thermo Fisher scientific, USA), the manufacturers’ instruction was followed. Briefly, the recombinant proteins were incubated in coupling buffer pH=10, added to the resin and incubated for 4 hours with 50 mM NaCNBH.sub.3 in coupling buffer pH=7.2. Blocking of the unbound surface of the resin was performed by incubation with 50 mM NaCNBH.sub.3 in quenching buffer. In all cases, the protein concentrations were measured before and after the coupling to determine the coupling efficiency. An equal concentration of around 500 mg/ml from each recombinant protein was used for immobilization to the corresponding resin.

[0168] For the assessment of the longitudinal stability of the scVEGF165 agarose column, 1.8 ml of VEGF coupled agarose resin was generated, divided into 6 aliquots of 300 .Math.l and stored at 4° C. Adsorption studies were performed after 1, 15, 30, 60, and 90 days.

Precipitation of Strep-Tagged Proteins From Flow Through

[0169] Strep-tagged proteins were precipitated from the flow through after apheresis treatment of serum samples by adding Strep-Tactin®XT (IBA Lifescience) beads. For the positive control, recombinant strep-tagged VEGF was added to the serum flow through. After washing the beads twice with TRIS 50 mM, NaCl 150 mM, pH=8, 30 .Math.l 2 x Laemmli buffer was directly added to the beads followed by SDS polyacrylamide gel electrophoresis, and immunoblot blot using Strep-tag specific antibody.

Ethics Approval and Informed Consent

[0170] Written informed consent was obtained from all participants according to protocols 10-238 and 09-258 as reviewed and approved by the ethics committee of the University Hospital Cologne. Preeclampsia was defined as new onset of hypertension >140/90 mmHg, proteinuria >0.3 g/g Creatinine, and sFlt-1/PIGF > 85 at EGA ≤32 weeks.

Example 1 - Molecular Modelling and Structural Analysis of Single Chain VEGF.SUP.165 Dimers

[0171] Intermolecular disulfide bonds stabilize the structure of VEGF and PIGF homodimers and unfold the receptor binding sites. These unique properties could be utilized to engineer higher order multimers of VEGF and PIGF with enhanced binding affinity to sFlt-1.

[0172] Expression constructs containing a VEGF.sup.165 lacking the N-terminal sequence followed by a short inert 14 amino acid linker and a second VEGF.sup.165 lacking the N-terminal signal peptide (single chain VEGF dimers = scVEGF.sup.165) equipped with a cleavable Strep-tag® for purification (FIG. 1A) were designed (SEQ ID NO: 7). It was speculated that the short 14-amino acid linker hampered assembly of scVEGF.sup.165 in a 40 kDa dimer. In contrast, an open structure of two or more scVEGF.sup.165 molecules linked by intermolecular disulfide bonds would allow formation of tetramers, or multimeric chains of VEGF. Molecular modelling approaches of the tetrameric quaternary assembly of scVEGF.sup.165 were based on structural restraints of the 14-amino acid linker and negative stain electron micrographs (FIGS. 1B+C).

[0173] Monomeric wildtype VEGF.sup.165 (moVEGF.sup.165, full-length VEGF.sup.165 lacking the N-terminal sequence), and scVEGF.sup.165 constructs were cloned as detailed above (see also Table 1), expressed in human embryonic kidney cells (HEK 293T), and the protein was purified under physiologic conditions using Strep-TactinⓇ technology (FIG. 7). The constructs were biologically functional as assessed in experiments investigating mitogen-activated protein kinase (MAP)-kinase activation in HUVEC cells in culture (FIG. 9). Activation of map kinase pathway can be detected after treatment with recombinant VEGF and PLGF variants. Co-treatment of VEGF and sFlt-1 abrogates MAP-activation (FIG. 9).

[0174] To visualize the quaternary molecular structure of scVEGF.sup.165 in comparison to moVEGF.sup.165, negative staining electron microscopy was employed (FIG. 1C). As expected for monomeric expressed VEGF, moVEGF.sup.165 assembled in dumbbell-shaped dimeric complexes with a few monomeric molecules represented as single dots (left panel FIG. 1C). In striking contrast, scVEGF.sup.165 appeared as uniformly ordered cloverleaf-shaped structures representing complexes of scVEGF.sup.165 in a 2:2 configuration, i.e. tetramers of VEGF.sup.165.

[0175] In addition to moVEGF.sup.165 and scVEGF.sup.165, monomeric PIGF (moPIGF) and single chain PIGF dimers (scPIGF) constructs were generated and expressed and tested as previously described (FIG. 7 top right and bottom right).

[0176] Subsequently, binding characteristics of moVEGF.sup.165, scVEGF.sup.165, moPIGF and scPIGF were explored. To screen for strong interactors, the binding affinity and static binding capacity of VEGF and PIGF constructs to VEGF-trap were quantified as the equilibrium dissociation constant (Kd) by ELISA based serial dilution experiments and protein loading in high excess, respectively. The VEGF-Trap as used in the present application is a chimeric protein containing the second binding domain of the VEGFR-1 receptor and the third domain of the VEGFR-2 receptor fused to the Fc segment of a human IgG backbone resulting in a very high VEGF binding affinity (Kd≈1 pM).

[0177] As expected from previous reports, moVEGF.sup.165 displayed higher affinity (2.6 fold) as compared to moPIGF (Christinger et al., 2004). Strikingly, scVEGF.sup.165 however showed 11% higher affinity as compared to moVEGF.sup.165. Interestingly, affinity of scPIGF was not significantly different compared to moPIGF (FIGS. 2A+B).

[0178] Results: It is a very surprising finding that the binding affinity of scVEGF.sup.165 to VEGF-trap is 11% higher in comparison to that moVEGF.sup.165, especially in light of the fact that the binding affinity of scPIGF in comparison to moPIGF is approximately equal. Further, the surprisingly large variation in binding capacity found in the different constructs is remarkable.

Example 2- Evaluation of Substrate Matrices and Stability of sFlt-1 Capturing Apheresis Columns

[0179] To analyze coupling efficiency of recombinant proteins to different resins, the recombinant protein concentration before and after coupling to the resin was determined. The results are shown in Table 2.

[0180] Table 2: Representative data showing recombinant protein concentration before and after coupling to resin, to determine coupling efficiency of recombinant proteins to different resin.

TABLE-US-00005 Protein name [Protein](.Math.g/ ml) before coupling [Protein](.Math.g/ml) after coupling Percentage moVEGF.sup.165 coupled agarose 519.1 0.192 99.96 moVEGF.sup.165 coupled CnBr 518.9 0.124 99.97 moVEGF.sup.165 coupled StrepTactin 518.5 0.187 99.96 moPIGF coupled agarose 487.9 0.161 99.967 moPIGF coupled CnBr 490 0.117 99.97 moPIGF coupled StrepTactin 489.2 0.158 99.96 moVEGF.sup.165 coupled agarose 513 0.113 99.977 scVEGF.sup.165 coupled agarose 538 0.189 99.96 VEGF DR coupled agarose 523 0.154 99.97 moPLGF coupled agarose 512 0.139 99.97 scPIGF coupled agarose 524 0.143 99.97 Flt-1 antibody coupled agarose 543 0.182 99.96

[0181] To optimize the dynamic binding characteristics of the apheresis columns, sFlt-1 adsorption from human serum by immobilized moPIGF and moVEGF.sup.165 was tested using Streptactin XT®, Cyanogen bromide activated sepharose or aldehyde-activated agarose for immobilization of the ligand. Significant adsorption of sFlt-1 was noted for all resins on both moPIGF- as well as moVEGF.sup.165-based columns. However, the largest reduction of 86.16 % sFlt-1 in a single run was achieved with columns where scVEGF.sup.165 was immobilized on agarose matrix (FIG. 3A). Stability of the scVEGF.sup.165-based agarose column was assessed at 1, 15, 30, 60, and 90 days after generation and no significant loss of affinity noted (FIG. 3B).

[0182] Results: Binding of a ligand to an agarose matrix led for all ligands (moPIGF, moVEGF.sup.165 and scVEGF.sup.165) to the largest adsorption of sFlt-1 from human serum in comparison to binding to a Streptactin XT® matrix or a sepharose matrix. Moreover, sFlt-1 adsorption from human serum was significantly higher for scVEGF.sup.165 in comparison to moPIGF or moVEGF.sup.165 (with moVEGF.sup.165 having a slightly better sFlt-1 adsorption from human serum in comparison to moPIGF). The largest reduction of 86.16 % sFlt-1 in a single run was achieved with columns where scVEGF.sup.165 was immobilized on agarose matrix. Moreover, the examination of longitudinal stability of scVEGF.sup.165 agarose column revealed only a slight decrease in binding sFlt-1.

Example 3 - Characterization of sFlt-1 Capturing Ligands

[0183] In a next step, the efficacy of sFlt-1 reduction from patient serum was assessed for the different VEGF- and PIGF-variants immobilized on agarose apheresis columns.

[0184] Mini columns carrying aldehyde-activated agarose were equally loaded with VEGF- or PIGF-variants (Table 2). In addition, one antibody-based apheresis column was generated to provide a reference control. sFlt-1, VEGF-, and PIGF-concentrations were determined in a serum sample of a patient with preeclampsia using commercial ELISA kits before and after single runs over a control column or columns equipped with the respective VEGF- or PIGF-variant or Flt-1 specific antibody. Whereas no difference between sFlt-1-, PIGF-, and VEGF-concentrations was noted after running the sample over the control column, sFlt-1 levels were reduced by varying amounts depending on the specific ligand (FIG. 4A). With moVEGF.sup.165 immobilized on the column 77.15% reduction of sFlt-1 was achieved in a single passage of patient serum, which equaled reduction achieved by the antibody column. In contrast and as expected (Christinger et al., 2004) due to the lower binding affinity and capacity of moPIGF as compared to moVEGF.sup.165, adsorption with the moPIGF column only yielded 46.18% reduction of sFlt-1. For moVEGF columns liberation of PIGF and VEGF was noted, whereas for moPIGF columns only release of PIGF but no release of VEGF was detected and the antibody column released neither VEGF nor PIGF (FIGS. 4B+C).

[0185] Strikingly, in the adsorption experiments employing patient serum, the modified single chain VEGF-dimers (scVEGF.sup.165) boasted sFlt-1 reduction to 89.87% in a single run (FIG. 4A), while at the same time, releasing VEGF and PIGF in large amounts (FIGS. 4B+C). Interestingly, for PIGF the optimized multimerization strategy did not affect sFlt-1 binding and PIGF release to the same extent (FIG. 4A). Only minor enhancement of sFlt-1 binding and no significant difference in PIGF release was documented for the scPIGF column as compared to the moPIGF column (FIGS. 4A+B). Expectedly, as for moPIGF the scPIGF columns did not release endogenous VEGF (FIG. 4C).

[0186] The notion that increased VEGF or PIGF levels in the flow through might represent leakage of recombinant protein from the column was invalidated by purification of Strep-tagged proteins from the flow through of columns carrying Strep-tagged VEGF or PIGF after passage of serum samples. For all constructs tested, no leaked protein was detected in western blot analysis employing Strep-specific primary antibody (FIG. 9).

[0187] Results: All columns with immobilized VEGF and PIGF variants as well as with the Flt-1 specific antibody reduced sFlt-1 in the sample. The reduction of sFlt-1 was greatest for scVEGF.sup.165 columns (p<0.0001), all other VEGF and PIGF variants as well as the Flt-1 specific antibody reduced sFlt-1 less efficiently (FIG. 4A). Apheresis with VEGF and PIGF columns resulted in release of PIGF most pronounced for scVEGF.sup.165 (p<0.0001), while apheresis with moVEGF.sup.165 only resulted in a release of PIGF half that much of scVEGF.sup.165. Apheresis with sFlt-1 antibody did not result in a significant PIGF release (FIG. 4B). VEGF release was only present in apheresis columns containing VEGF variants, again most pronounced for scVEGF.sup.165 (p<0.0001). Apheresis of serum samples using columns with PIGF variants and Flt-1 specific antibody did not result in VEGF liberation (FIG. 4C).

Example 4 - Validation of scVEGF.SUP.165.-Based Apheresis in Independent Patient Samples

[0188] The yield of sFlt-1 reduction and VEGF-/PIGF-release by scVEGF.sup.165 columns was substantiated in serum samples of 10 independent patients at different gestational ages with clinically suspected preeclampsia and high sFlt-⅟PIGF ratios.

[0189] Results: In all patient samples, sFlt-1 reduction of 88% (mean) (median 92.2%; SD 5.27) (FIG. 5A), PIGF release of 20-fold (mean) compared to the initial levels (median 14.57 fold; SD 5.40) (FIG. 5B) as well as VEGF release of 9.07 fold (mean) compared to the initial levels (median 8.74.; SD 4.15) (FIG. 5C) was achieved in a single run over the scVEGF.sup.165 column.

[0190] Apheresis systems employing Flt-1 specific antibody reduce sFlt-1 levels but do not liberate endogenous PIGF or VEGF (FIG. 6A). Adsorption columns containing PIGF retain sFlt-1 and liberate low quantities of endogenous PIGF but not endogenous VEGF (FIG. 6B). Monomeric VEGF.sup.165 columns capture sFlt-1 and release larger quantities of PIGF and also VEGF (FIG. 6C). Enhanced binding characteristics of scVEGF.sup.165 for sFlt-1 result in most efficient adsorption of sFlt-1 and competitive release of PIGF and VEGF leading to restitution of the angiogenic balance in preeclampsia (FIG. 6D).

[0191] In general, the unique characteristic of scVEGF.sup.165-based apheresis compared to antibody-mediated apheresis and moVEGF- or PIGF-based approaches is the enhanced affinity for sFlt-1 and thus the capability to efficiently release VEGF and PIGF (FIG. 6).

References

[0192] Sibai BM. Prevention of preeclampsia: a big disappointment. Am J Obstet Gynecol. 1998;179:1275-1278.

[0193] Redman CW. Latest Advances in Understanding Preeclampsia. Science. 2005;308:1592-1594.

[0194] Walker JJ. Pre-eclampsia. Lancet. 2000;356:1260-1265.

[0195] Maynard SE, Min J-Y, Merchan J, Lim K-H, Li J, Mondal S, Libermann TA, Mor-gan JP, Sellke FW, Stillman IE, Epstein FH, Sukhatme VP, Karumanchi SA. Ex-cess placental soluble fms-like tyrosine kinase 1 (sFlt1) may contribute to endo-thelial dysfunction, hypertension, and proteinuria in preeclampsia. J Clin Invest. 2003;111:649-658.

[0196] Levine RJ, Maynard SE, Qian C, Lim K-H, England LJ, Yu KF, Schisterman EF, Thadhani R, Sachs BP, Epstein FH, Sibai BM, Sukhatme VP, Karumanchi SA. Circulating angiogenic factors and the risk of preeclampsia. N Engl J Med. 2004;350:672-683.

[0197] ACOG Practice Bulletin No. 202: Gestational Hypertension and Preeclampsia. Obstetrics & Gynecology. 2019;133:e1-e25.

[0198] Makris A, Yeung KR, Lim SM, Sunderland N, Heffernan S, Thompson JF, Ili-opoulos J, Killingsworth MC, Yong J, Xu B, Ogle RF, Thadhani R, Karumanchi SA, Hennessy A. Placental Growth Factor Reduces Blood Pressure in a Utero-placental Ischemia Model of Preeclampsia in Nonhuman Primates. Hyperten-sion. 2016;67:1263-1272.

[0199] Li Z, Zhang Y, Ying Ma J, Kapoun AM, Shao Q, Kerr I, Lam A, O′Young G, Sannajust F, Stathis P, Schreiner G, Karumanchi SA, Protter AA, Pollitt NS. Re-combinant vascular endothelial growth factor 121 attenuates hypertension and improves kidney damage in a rat model of preeclampsia. Hypertension. 2007;50:686-692.

[0200] Eddy AC, Bidwell III GL, George EM. Pro-angiogenic therapeutics for preeclampsia. Biol Sex Differ. 2018; 9: 36.

[0201] Turanov AA, Lo A, Hassler MR, Makris A, Ashar-Patel A, Alterman JF, Coles AH, Haraszti RA, Roux L, Godinho BMDC, Echeverria D, Pears S, Iliopoulos J, Shanmugalingam R, Ogle R, Zsengeller ZK, Hennessy A, Karumanchi SA, Moore MJ, Khvorova A. RNAi modulation of placental sFLT1 for the treatment of preeclampsia. Nature Biotechnology. 2018;36:1164-1173.

[0202] Fan X, Rai A, Kambham N, Sung JF, Singh N, Petitt M, Dhal S, Agrawal R, Sut-ton RE, Druzin ML, Gambhir SS, Ambati BK, Cross JC, Nayak NR. Endometrial VEGF induces placental sFLT1 and leads to pregnancy complications. Journal of Clinical Investigation. 2014;124:4941-4952.

[0203] Thadhani R, Kisner T, Hagmann H, Bossung V, Noack S, Schaarschmidt W, Jank A, Kribs A, Cornely OA, Kreyssig C, Hemphill L, Rigby AC, Khedkar S, Lindner TH, Mallmann P, Stepan H, Karumanchi SA, Benzing T. Pilot Study of Extracorporeal Removal of Soluble Fms-Like Tyrosine Kinase 1 in Preeclampsia. Circulation. 2011;124:940-950.

[0204] Thadhani R, Hagmann H, Schaarschmidt W, Roth B, Cingoez T, Karumanchi SA, Wenger J, Lucchesi KJ, Tamez H, Lindner T, Fridman A, Thome U, Kribs A, Danner M, Hamacher S, Mallmann P, Stepan H, Benzing T. Removal of Soluble Fms-Like Tyrosine Kinase-1 by Dextran Sulfate Apheresis in Preeclampsia. J Am Soc Nephrol. 2016;27:903-913.

[0205] Trapiella-Alfonso L, Alexandre L, Fraichard C, Pons K, Dumas S, Huart L, Gaucher JF, Hebert-Schuster M, Guibourdenche J, Fournier T, Vidal M, Broutin I, Lecomte-Raclet L, Malaquin L, Descroix S, Tsatsaris V, Gagey-Eilstein N, Lecarpentier E. VEGF (Vascular Endothelial Growth Factor) Functionalized Magnetic Beads in a Microfluidic Device to Improve the Angiogenic Balance in Preeclampsia. Hypertension. 2019 Jul;74(1):145-153. doi: 10.1161/HYPERTENSIONAHA.118.12380. Epub 2019 May 13.

[0206] Jain A, Cheng K. The principles and applications of avidin-based nanoparticles in drug delivery and diagnosis. Journal of Controlled Release. 2017;245:27-40.

[0207] Claffey KP, Senger DR, Spiegelman BM. Structural requirements for dimeriza-tion, glycosylation, secretion, and biological function of VPF/VEGF. Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology. 1995;1246:1-9.

[0208] Peretz D, Gitay-Goren H, Safran M, Kimmel N, Gospodarowicz D, Neufeld G. Glycosylation of vascular endothelial growth factor is not required for its mito-genic activity. Biochemical and Biophysical Research Communications. 1992;182:1340-1347.

[0209] Kuriakose A, Chirmule N, Nair P. Immunogenicity of Biotherapeutics: Causes and Association with Posttranslational Modifications. Journal of Immunology Research. 2016;2016:1-18.

[0210] Tatusova et al. FEMS Microbiol. Lett. 1999; 174: 247-250.

[0211] Mizushima and Nagata, pEF-BOS, a powerful mammalian expression vector. Nucleic Acids Res. (1990) 18(17), 5322.

[0212] Mulligan RC, Howard BH, Berg P. Synthesis of rabbit beta-globin in cultured monkey kidney cells following infection with a SV40 beta-globin recombinant genome. Nature. 1979; 277(5692):108-14.

[0213] Agarwal P, Zwolanek D, Keene DR, Schulz J-N, Blumbach K, Heinegard D, Zaucke F, Paulsson M, Krieg T, Koch M, Eckes B. Collagen XII and XIV, New Partners of Cartilage Oligomeric Matrix Protein in the Skin Extracellular Matrix Suprastructure. Journal of Biological Chemistry. 2012;287:22549-22559.

[0214] Bober M, Enochsson C, Collin M, Mörgelin M. Collagen VI Is a Subepithelial Adhesive Target for Human Respiratory Tract Pathogens. Journal of Innate Immunity. 2010;2:160-166.

[0215] Christinger HW, Fuh G, de Vos AM, Wiesmann C. The Crystal Structure of Pla-cental Growth Factor in Complex with Domain 2 of Vascular Endothelial Growth Factor Receptor-1. Journal of Biological Chemistry. 2004;279:10382-10388.