Bispecific anti-VEGF and anti-TrkB binding molecules for the treatment of ocular diseases

Abstract

The present invention relates to binding molecules that specifically bind to vascular endothelial growth factor (VEGF) and tropomyosin receptor kinase B (TrkB), and to their use in medicine, pharmaceutical compositions containing them, and methods of using them as medicaments for the treatment and / or prevention of eye diseases.

Description

[Technical field]

[0001] FIELD OF THEINVENTION The present invention relates to binding molecules which bind to vascular endothelial growth factor (VEGF) and tropomyosin receptor kinase B (TrkB) and to their use in medicine, pharmaceutical compositions containing them and methods of their use as medicaments for the treatment and / or prevention of eye diseases.

[0002] 2. Background of the Invention Macular degeneration, particularly age-related macular degeneration (AMD), is the leading cause of blindness in developed countries. Its advanced stages are divided into wet neovascular AMD and dry geographic atrophy (GA). wAMD is characterized by increased ocular neovascularization and vascular leakage as well as progressive neurodegeneration and photoreceptor loss. Anti-VEGF treatment is the current standard of care for this disease. Treatment typically consists of drugs that target VEGF (Avastin, Lucentis, Eylea, and more recently, Beovu), which are administered by injection into the back of the eye.

[0003] VEGF sequestration reduces vascular leakage, which initially improves vision, but the retina remains compromised, leading to further cell loss. Although intravitreal anti-VEGF therapy has been a breakthrough for wAMD over the past decade, no treatment options exist for geographic atrophy, a slowly progressive disease that ultimately leads to blindness.

[0004] Within two years of wAMD diagnosis, nearly one-third of patients develop progressive geographic atrophy despite successful anti-VEGF treatment. It is the later stages that are associated with progressive visual impairment. Patients lose their vision and independence. Their daily life is severely affected by the inability to drive, read, watch television, and recognize faces. wAMD patients are often not treated in a real-world setting and must return to the clinic for frequent intravitreal injections.

[0005] Currently, there is no treatment available for any type of geographic atrophy, including that which occurs during wAMD. Furthermore, there is some evidence that more frequent anti-VEGF treatments may be associated with a higher rate of geographic atrophy development.

[0006] Therefore, there is a high unmet need to provide novel therapeutic biological molecules that can be frequently used in anti-VEGF treatment, while at the same time reducing the risk of concurrent geographic atrophy development.

[0007] Summary of the Invention The present invention is based on the concept of combining within a single binding molecule an antigen-binding site that specifically binds vascular endothelial growth factor (VEGF) and an antigen-binding site that specifically binds tropomyosin receptor kinase B (TrkB).

[0008] In a first aspect, the present invention relates to a binding molecule comprising at least one antigen-binding site that specifically binds to vascular endothelial growth factor (VEGF) and at least one antigen-binding site that specifically binds to tropomyosin receptor kinase B (TrkB). In one embodiment of the first aspect, the binding molecule is bispecific and tetravalent. In a further embodiment of the binding molecule of the first aspect or any of its embodiments, the at least one antigen-binding site that specifically binds to VEGF is an immunoglobulin (Ig) molecule. In a further embodiment of the binding molecule of the first aspect or any of its embodiments, the at least one antigen-binding site that specifically binds to TrkB is fused to the C-terminus of the heavy chain of the Ig molecule. In a further embodiment of the binding molecule of the first aspect or any of its embodiments, the at least one antigen-binding site that specifically binds to TrkB comprises one or more scFvs. In a further embodiment of the binding molecule of the first aspect or any of its embodiments, the at least one antigen-binding site that specifically binds VEGF is an immunoglobulin (Ig) molecule and the at least one antigen-binding site that specifically binds TrkB comprises one or more scFvs. In a related embodiment, the one or more scFvs have a VL-VH orientation from N-terminus to C-terminus. In yet another related embodiment, the one or more scFvs are fused to the C-terminus of the heavy chain of the Ig molecule. In a further embodiment of the binding molecule of the first aspect or any of its embodiments, the Ig molecule is a monoclonal antibody, a human monoclonal antibody, a humanized monoclonal antibody, a chimeric antibody, an (scFv)2 or a fragment of an antibody, such as an F(ab')2 fragment. In a further embodiment of the binding molecule of the first aspect or any of its embodiments, the Ig molecule is an IgG or an F(ab')2. In another further embodiment of the binding molecule of the first aspect or any of its embodiments, one or more scFvs are fused to the Ig molecule by a peptide linker, preferably a peptide linker having a length of about 4-20 amino acids.

[0009] In a further embodiment of the binding molecule of the first aspect or any of its embodiments, the antigen binding site that specifically binds to TrkB comprises: a) an antigen-binding site comprising a light chain CDR comprising the amino acid sequences of SEQ ID NO: 201 (CDR1), SEQ ID NO: 202 (CDR2) and SEQ ID NO: 203 (CDR3); and An antigen-binding site comprising heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 204 (CDR1), SEQ ID NO: 205 (CDR2) and SEQ ID NO: 206 (CDR3); or An antigen-binding site comprising heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 207 (CDR1), SEQ ID NO: 208 (CDR2) and SEQ ID NO: 209 (CDR3); or An antigen-binding site comprising heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 210 (CDR1), SEQ ID NO: 211 (CDR2) and SEQ ID NO: 212 (CDR3). is selected from.

[0010] In a further embodiment of the binding molecule of the first aspect or any of its embodiments, the antigen binding site that specifically binds to TrkB comprises: a) an antigen-binding site comprising a light chain variable domain comprising an amino acid sequence that is at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 213 or 215, and a heavy chain variable domain comprising an amino acid sequence that is at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 214 or 216. is selected from.

[0011] In a further embodiment of the binding molecule of the first aspect or any of its embodiments, the binding molecule comprises the amino acid sequence of SEQ ID NO:222 or the amino acid sequence of SEQ ID NO:223 or the amino acid sequence of SEQ ID NO:224 or the amino acid sequence of SEQ ID NO:225.

[0012] In a further embodiment of the binding molecule of the first aspect or any of its embodiments, the antigen-binding site that specifically binds to VEGF comprises antigen-binding sites a) to c): a) an antigen-binding site comprising a light chain CDR comprising the amino acid sequences of SEQ ID NO: 145 (CDR1), SEQ ID NO: 146 (CDR2) and SEQ ID NO: 147 (CDR3); and An antigen-binding site comprising heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 148 (CDR1), SEQ ID NO: 149 (CDR2) and SEQ ID NO: 150 (CDR3); or An antigen-binding site comprising heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 151 (CDR1), SEQ ID NO: 152 (CDR2) and SEQ ID NO: 153 (CDR3); or an antigen-binding site comprising heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 154 (CDR1), SEQ ID NO: 155 (CDR2), and SEQ ID NO: 156 (CDR3); b) an antigen-binding site comprising a light chain CDR comprising the amino acid sequences of SEQ ID NO: 157 (CDR1), SEQ ID NO: 158 (CDR2) and SEQ ID NO: 159 (CDR3); and An antigen-binding site comprising heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 160 (CDR1), SEQ ID NO: 161 (CDR2) and SEQ ID NO: 162 (CDR3); or An antigen-binding site comprising heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 163 (CDR1), SEQ ID NO: 164 (CDR2) and SEQ ID NO: 165 (CDR3); or an antigen-binding site comprising heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 166 (CDR1), SEQ ID NO: 167 (CDR2), and SEQ ID NO: 168 (CDR3); c) an antigen-binding site comprising a light chain CDR comprising the amino acid sequences of SEQ ID NO: 169 (CDR1), SEQ ID NO: 170 (CDR2) and SEQ ID NO: 171 (CDR3); and An antigen-binding site comprising heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 172 (CDR1), SEQ ID NO: 173 (CDR2) and SEQ ID NO: 174 (CDR3); or An antigen-binding site comprising heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 175 (CDR1), SEQ ID NO: 176 (CDR2) and SEQ ID NO: 177 (CDR3); or An antigen-binding site comprising heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 178 (CDR1), SEQ ID NO: 179 (CDR2) and SEQ ID NO: 180 (CDR3). is selected from the group consisting of:

[0013] In a further embodiment of the binding molecule of the first aspect or any of its embodiments, the antigen-binding site that specifically binds to VEGF comprises antigen-binding sites a) to d): a) an antigen-binding site comprising a light chain variable domain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 181, and a heavy chain variable domain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 182; b) an antigen-binding site comprising a light chain variable domain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 183, and a heavy chain variable domain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 184; c) an antigen-binding site comprising a light chain variable domain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 189, and a heavy chain variable domain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 190, 191, 192 or 194; d) an antigen-binding site comprising a light chain variable domain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 193, and a heavy chain variable domain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 190, 191, 192 or 194. is selected from the group consisting of:

[0014] In a further embodiment of the binding molecule of the first aspect or any of its embodiments, the antigen-binding site that specifically binds to VEGF comprises antigen-binding sites a) to d): a) an antigen-binding site comprising a light chain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 185, and a heavy chain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 186; b) an antigen-binding site comprising a light chain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 187, and a heavy chain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 188; c) an antigen-binding site comprising a light chain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 195, and a heavy chain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 196, 197, 198 or 200; d) an antigen-binding site comprising a light chain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 199, and a heavy chain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 196, 197, 198 or 200. is selected from the group consisting of:

[0015] In a second aspect, the present invention provides a method for the preparation of a ribozyme comprising the steps of: , 111, 113, 115, 117, 119, 121, 123, 125, 127, 129, 131, 133, 135, 137, 139, 141 or 143; and SEQ ID NO: 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 44, 46, 48, 50, 52, 54, 56, 58, 60, 62, 64, 66, 68, 70, 72, 74, 76, 78, 80, 82, 84, 86, 88, 90, 92, 94, 96, 98, 100, 102, 104, 106, 108, 110, 112, 114, 116 , 118, 120, 122, 124, 126, 128, 130, 132, 134, 136, 138, 140, 142 or 144, and a heavy chain comprising an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO:1, 118, 120, 122, 124, 126, 128, 130, 132, 134, 136, 138, 140, 142 or 144.

[0016] The present invention relates to a binding molecule comprising:

[0017] In a fourth aspect, the present invention relates to a TrkB binding molecule comprising or consisting of two scFvs, wherein each scFv specifically binds to TrkB.

[0018] In a further embodiment of the fourth aspect, the TrkB binding molecule further comprises an Ig molecule.

[0019] In a further embodiment of the fourth aspect or any of its embodiments, the Ig molecule is a monoclonal antibody, a human monoclonal antibody, a humanized monoclonal antibody, a chimeric antibody, a fragment of an antibody, such as an Fv, Fab, Fab' or F(ab')2 fragment, a single chain antibody, such as a single chain variable fragment (scFv), a small modular immunopharmaceutical (SMIP), a domain antibody, a nanobody, or a diabody.

[0020] In a further embodiment of the fourth aspect or any of its embodiments, each scFv is fused to the C-terminus of a heavy chain of an Ig molecule. In a further embodiment of the fourth aspect or any of its embodiments, the Ig molecule is an IgG, F(ab) or F(ab')2. In a further embodiment of the fourth aspect or any of its embodiments, the Ig molecule comprises or consists of an Fc region. In a further embodiment of the fourth aspect or any of its embodiments, the TrkB binding molecule is bispecific and tetravalent.

[0021] In a fifth aspect, the present invention relates to an scFv that specifically binds to TrkB, comprising (i) a light chain variable domain comprising the amino acid sequence of SEQ ID NO: 213 and a heavy chain variable domain comprising the amino acid sequence of SEQ ID NO: 214, or (ii) a light chain variable domain comprising the amino acid sequence of SEQ ID NO: 215 and a heavy chain variable domain comprising the amino acid sequence of SEQ ID NO: 216.

[0022] In a further embodiment of the fifth aspect, the scFv comprises the amino acid sequence of SEQ ID NO:222 or the amino acid sequence of SEQ ID NO:223 or the amino acid sequence of SEQ ID NO:224 or the amino acid sequence of SEQ ID NO:225.

[0023] In a sixth aspect, the present invention provides a method for producing a composition comprising the steps of: (i) a light chain variable domain comprising the amino acid sequence of SEQ ID NO: 189; and A heavy chain variable domain comprising the amino acid sequence of SEQ ID NO: 191; or A heavy chain variable domain comprising the amino acid sequence of SEQ ID NO: 192; or Heavy chain variable domain comprising the amino acid sequence of SEQ ID NO: 194 (ii) a heavy chain variable domain comprising the amino acid sequence of SEQ ID NO: 190; and A light chain variable domain comprising the amino acid sequence of SEQ ID NO: 192; or A light chain variable domain comprising the amino acid sequence of SEQ ID NO: 193 The present invention relates to an antibody molecule that specifically binds to VEGF, comprising:

[0024] In a further embodiment of the sixth aspect, the antibody molecule comprises a light chain comprising the amino acid sequence of SEQ ID NO: 195 or 199 and a heavy chain comprising the amino acid sequence of SEQ ID NO: 196, 197, 198 or 200.

[0025] In a seventh aspect, the invention relates to an isolated nucleic acid molecule encoding (i) the heavy chain or the heavy chain variable domain and / or (ii) the light chain or the light chain variable domain of a binding molecule of any one of the preceding aspects and embodiments.

[0026] In an eighth aspect, the present invention relates to an isolated nucleic acid molecule encoding (i) the heavy chain or the heavy chain variable domain and / or (ii) the light chain or the light chain variable domain of a TrkB binding molecule of any one of the preceding aspects and embodiments.

[0027] In a ninth aspect, the present invention relates to an isolated nucleic acid molecule encoding (i) the heavy chain variable domain and / or (ii) the light chain variable domain of an scFv of any one of the preceding aspects and embodiments.

[0028] In a tenth aspect, the invention relates to an isolated nucleic acid molecule encoding (i) the heavy chain or a heavy chain variable domain and / or (ii) the light chain or a light chain variable domain of an antibody molecule of any one of the preceding aspects and embodiments.

[0029] In an eleventh aspect, the present invention relates to a viral vector comprising the isolated nucleic acid molecule of the seventh, eighth, ninth or tenth aspect.

[0030] In a twelfth aspect, the present invention relates to an expression vector comprising the nucleic acid molecule of the eleventh aspect.

[0031] In a thirteenth aspect, the present invention relates to a host cell transfected with an expression vector of the twelfth aspect.

[0032] In a fourteenth aspect, the present invention provides a method for producing a composition comprising the steps of: (a) culturing a host cell of the thirteenth aspect under conditions allowing expression of a molecule: (b) recovering the molecule, and optionally, (c) further purifying and / or modifying and / or formulating the molecule. The present invention relates to a method of producing the binding molecule, TrkB binding molecule, scFv or antibody molecule of any of the preceding aspects and embodiments.

[0033] In a fifteenth aspect, the present invention relates to a binding molecule, TrkB binding molecule, scFv or antibody molecule of any of the preceding aspects and embodiments for use in medicine.

[0034] In an embodiment of the fifteenth aspect, the present invention relates to a binding molecule, a TrkB binding molecule, an scFv or an antibody molecule for use of the fifteenth aspect, wherein said use is the treatment of an ocular or retinal disease or a neurodegenerative disease.

[0035] In a further embodiment of the fifteenth aspect, the invention relates to a binding molecule, a TrkB-binding molecule, an scFv or an antibody molecule for use of the fifteenth aspect, wherein the use is for the treatment of a neurological / neuro-ocular or retinal disease.

[0036] In another embodiment of the fifteenth aspect, the invention relates to a method for treating macular degeneration, age-related macular degeneration, diabetic retinopathy, diabetic macular edema, retinitis pigmentosa, hereditary retinal dystrophies, hereditary macular dystrophies, myopic degeneration, geographic atrophy, retinal artery occlusion, endophthalmitis, uveitis, cystoid macular edema, choroidal neovascular membrane secondary to any retinal disease, optic neuropathy, glaucoma, retinal detachment, toxic retinopathy, radiation retinopathy and traumatic retinopathy, precursor and mild to moderate Alzheimer's disease. The fifteenth aspect relates to the binding molecule, TrkB binding molecule, scFv or antibody molecule for use in treating Alzheimer's disease, delaying disease progression in patients with Alzheimer's disease, Huntington's disease, Parkinson's disease, major depressive disorder, schizophrenia, cognitive impairment associated with schizophrenia, preventing first-episode psychosis in individuals with attenuated psychotic syndromes, preventing relapse in patients with schizophrenia, treatment-resistant depression, bulimia, obesity or metabolic syndrome.

[0037] In an embodiment relating to the fifteenth aspect, the present invention relates to a binding molecule, a TrkB-binding molecule, an scFv or an antibody molecule for use according to the fifteenth aspect, wherein the use is for the treatment of macular degeneration, in particular wet age-related macular degeneration (wAMD).

[0038] In an embodiment of the fifteenth aspect, the present invention relates to a binding molecule, a TrkB binding molecule, an scFv or an antibody molecule for use of the fifteenth aspect, wherein the use is for the treatment of retinal vein occlusion (RVO).

[0039] In an embodiment of the fifteenth aspect, the invention relates to a binding molecule, a TrkB-binding molecule, an scFv or an antibody molecule for use of the fifteenth aspect, wherein the use is for the treatment and / or prevention of geographic atrophy.

[0040] In a sixteenth aspect, the present invention relates to a pharmaceutical composition comprising a pharma- ceutically acceptable carrier and the binding molecule, TrkB binding molecule, scFv or antibody molecule of any of the preceding aspects and embodiments.

[0041] In a seventeenth aspect, the present invention relates to a method for treating or preventing an ocular or retinal disease or a neurodegenerative disease, comprising administering to a patient in need thereof a therapeutically effective amount of a binding molecule, TrkB binding molecule, scFv or antibody molecule according to any of the preceding aspects and embodiments.

[0042] In an eighteenth aspect, the invention relates to the use of a binding molecule, TrkB binding molecule, scFv or antibody molecule according to any of the preceding aspects and embodiments for the manufacture of a pharmaceutical composition for treating or preventing an eye or retinal disease or a neurodegenerative disease. [Brief description of the drawings]

[0043] [Figure 1] Figure 1: Diagram of the design of single VEGF-TrkB binding molecules, series 1 to series 4. [Figure 2A-B] Figures 2A-B: (A) TrkB phosphorylation (Y706 / 707) or (B) ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of the indicated molecules. Data represent the mean + / - SEM. [Figure 3A-D] Figures 3A-D: TrkB phosphorylation (Y706 / 707) was measured in CHO cells stably expressing (A) cynomolgus monkey TrkB, (B) rabbit TrkB, (C) rat TrkB or (D) mouse TrkB after incubation with increasing concentrations of the indicated molecules. Data represent the mean + / - SEM. [Figure 4A-C] Figures 4A-C: Trk phosphorylation (Y706 / 707) was measured in CHO cells stably expressing (A) human TrkA, (B) human TrkB, or (C) human TrkC after incubation with increasing concentrations of the indicated molecules. The natural ligands NGF for TrkA, BDNF for TrkB, and NT-3 for TrkC were used as controls. Data represent the mean + / - SEM. [Figure 5A-B] Figure 5A-B: (A) CHO cells stably expressing human TrkB were incubated with the indicated concentrations of the native TrkB ligand BDNF (duplicates), 1 nM BDNF with the indicated concentrations of the agonistic TrkB antibody C2 (triplicates), or 1 nM BDNF with the indicated concentrations of Doppelmab TPP-11736 (triplicates). (B) TrkB internalization was assessed by immunofluorescence staining of surface TrkB receptors followed by confocal microscopy analysis. Dark and bright fields in the heatmap represent high and low percentages of cells above the fluorescence threshold, respectively. [Figure 6] Figure 6: ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of the indicated molecules, with or without preincubation with 200 ng / mL human VEGF-A. Incubation with VEGF-A alone served as control. Data represent the mean. Error bars are omitted for clarity. [Figure 7] Figure 7: Neuroprotective function of TrkB activation in a rat model of diabetes-induced retinal neurodegeneration. Animals were treated with STZ to induce hyperglycemia. Retinal function was then assessed by electroretinogram (ERG) and rod-driven B-wave implicit time delay immediately before and 2 weeks after intravitreal application of the agonistic TrkB antibody C2 or Doppelmab TPP-11736. Mean + / - SEM; nsp>0.05, not significant; **p<0.01; ***p<0.001; one-way ANOVA with Tukey's multiple comparison test. Anti-TNP served as isotype control antibody. [Figure 8A-B] Figures 8A-B: VEGF-A uptake was assessed by measuring VEGF receptor 2 phosphorylation (Y1175-VEGFR2). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated molecules. (A) Comparison of Doppelmab TPP-11735, -736, -737 and -738. (B) Comparison of TPP-11736 and TPP-11738 with Eylea. Non-stimulated cells (basal) and 50ng / mL human VEGF without antibody treatment served as controls. Data represent the mean + / - SEM. [Figure 9A-B] Figures 9A-B: VEGF-A uptake was assessed by ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated molecules. (A) Comparison of Doppelmab TPP-11735, -736, -737 and -738. (B) Comparison of TPP-11736, -738 with Eylea. Non-stimulated cells (basal) and 50ng / mL human VEGF without antibody treatment served as controls. Data represent the mean + / - SEM. [Figure 10] Figure 10: VEGF-A uptake was assessed by measuring p38 MAPK phosphorylation (T180 / Y182). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated molecules. Unstimulated cells (basal) and 50 ng / mL human VEGF without antibody treatment served as controls. Data represent the mean + / - SEM. [Figure 11A-E]Figures 11A-11E: Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 10 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated molecules: (A) TPP-11735, (B) TPP-11736, (C) Eylea, (D) TPP-11737, (E) TPP-11738. Molecular concentrations are shown in mol / L. VEGF-A uptake was assessed by image-based quantification of HRMEC cell number. Relative cell numbers are shown. Cell number at t=0 was set to 1. Unstimulated cells served as control (basal). Data represent mean + / - SEM. [Figure 12] Figure 12: TrkB phosphorylation (Y706 / 707) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of the indicated molecules. Data represent the mean + / - SEM. [Figure 13] Figure 13: ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of the indicated molecules. Data represent the mean + / - SEM. [Figure 14A-B] Figures 14A-B: VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 phosphorylation (Y1175) or (B) ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated molecules. Non-stimulated cells (basal) and 50 ng / mL human VEGF without antibody treatment served as controls. Data represent the mean + / - SEM. [Figure 15]Figure 15: VEGF-A uptake was assessed by image-based quantification of HRMEC cell number. For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 10 ng / mL human VEGF with or without preincubation with 0.5 nM Doppelmab or 1 nM Eylea. Images were recorded every 4 hours for a total of 96 hours. Relative cell numbers are shown. Cell number at t=0 was set to 1. Unstimulated cells served as control (basal). Data represent mean + / - SEM. [Figure 16] Figure 16: Endothelial sprouting was assessed by confocal microscopy and spheroid circumferences obtained from maximum projections of Z-stacks are shown. For this purpose, spheroids of human retinal microvascular endothelial cells (HRMEC) were embedded in a collagen matrix. Endothelial sprouting was induced for 24 h by incubation with 50 ng / mL human VEGF with or without a 24 h preincubation with 2.5 nM Doppelmab or 5 nM Eylea as indicated. Unstimulated cells served as control (basal). Data represent the mean + / - SEM. n [Figure 17A-C] Figures 17A-C: VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 phosphorylation (Y1175), (B) ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) or (C) p38 MAPK phosphorylation (T180 / Y182). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of Doppelmab TPP-11736 or TPP-13788 (B20 IgG). 50 ng / mL human VEGF without antibody treatment served as control. Data represent mean + / - SEM. [Figure 18]Figure 18: Endothelial sprouting was assessed by confocal microscopy and spheroid circumferences obtained from maximum projections of Z-stacks are shown. For this purpose, spheroids of human retinal microvascular endothelial cells (HRMEC) were embedded in a collagen matrix. Endothelial sprouting was induced for 24 h by incubation with 50 ng / mL human VEGF with or without preincubation with 2.5 nM Doppelmab TPP-11736 or 2.5 nM TPP-13788 (B20 IgG). Unstimulated cells served as control (basal). Data represent the mean + / - SEM. [Figure 19A-C] Figures 19A-C: (A) TrkB phosphorylation (Y706 / 707) or (B and C) ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of the indicated molecules. Data represent the mean + / - SEM. [Figure 20A-B] Figures 20A-B: TrkB phosphorylation (Y706 / 707) was measured in CHO cells stably expressing (A) cynomolgus monkey TrkB or (B) rat TrkB after incubation with increasing concentrations of the indicated molecules. Data represent the mean + / - SEM. [Figure 21A-C] Figures 21A-C: VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 phosphorylation (Y1175), (B) ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) or (C) Src phosphorylation (Y419). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated molecules. 50 ng / ml human VEGF without antibody treatment served as control. Data represent mean + / - SEM. [Figure 22]Figure 22: Endothelial sprouting was assessed by confocal microscopy and spheroid circumferences obtained from maximum projections of Z-stacks were shown. For this purpose, spheroids of human retinal microvascular endothelial cells (HRMEC) were embedded in a collagen matrix. Endothelial sprouting was induced for 24 h by incubation with 50 ng / mL human VEGF with or without preincubation with 2.5 nM of the indicated molecules. Unstimulated cells served as control (basal). Data represent the mean + / - SEM. ns p>0.05 not significant, ****p<0.0001. [Fig. 23A-D] Figures 23A-D: VEGF-A capture was assessed by image-based quantification of HRMEC cell number. For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 10 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated molecules. Unstimulated cells (basal) and 50 ng / mL human VEGF without antibody treatment served as controls. Data represent the mean + / - SEM. (A) TPP-11736, (B) TPP-14936, (C) TPP-14937 or (D) Eylea. Bound molecule / Eylea concentration is shown in mol / L. Images were recorded every 4 hours for a total of 84 hours. Relative cell numbers are shown. Cell number at t=0 was set to 1. Unstimulated cells served as control (basal). Data represent the mean + / - SEM. [Fig. 24A-B] Figures 24A-B: (A) TrkB phosphorylation (Y706 / 707) or (B) ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of the second series of the indicated molecules. Data represent the mean + / - SEM. [Diagram 25] FIG. 25: Cartoon showing the proposed underlying effect of VEGF-induced clustering using a single binding molecule increased the efficacy and potency of TrkB activation. [Figure 26A-B]Figures 26A-B: TrkB phosphorylation (Y706 / 707) was measured in CHO cells stably expressing (A) cynomolgus TrkB or (B) rat TrkB after incubation with increasing concentrations of a second series of the indicated molecules. Data represent the mean + / - SEM. [Figure 27A-C] Figures 27A-C: TrkB phosphorylation (Y706 / 707) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of Doppelmab (A) TPP-14940, (B) TPP-14941, or (C) C2 antibody, with or without preincubation with 200 ng / mL human VEGF-A. Data represent mean + / - SEM. [Figure 28A-C] Figures 28A-C: ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of Doppelmab (A) TPP-14940, (B) TPP-14941 or (C) C2 antibody with or without preincubation with 200 ng / mL human VEGF-A. Data represent mean + / - SEM. [Figure 29A-B] Figures 29A-B: (A) TrkB phosphorylation (Y706 / 707) or (B) ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of either human VEGF-A or increasing concentrations of BDNF alone or with a fixed concentration of 200 ng / mL hVEGF and increasing concentrations of BDNF. Data represent the mean + / - SEM. [Fig. 30A-C] Figures 30A-C: TrkB phosphorylation (Y706 / 707) was measured in CHO cells stably expressing cynomolgus monkey TrkB after incubation with increasing concentrations of Doppelmab (A) TPP-14940, (B) TPP-14941, or (C) C2 antibody, with or without preincubation with 200 ng / mL human VEGF-A. Data represent mean + / - SEM. [Fig. 31A-C] Figures 31A-C: ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) was measured in CHO cells stably expressing cynomolgus TrkB after incubation with increasing concentrations of Doppelmab (A) TPP-14940, (B) TPP-14941 or (C) C2 antibody with or without preincubation with 200 ng / mL human VEGF-A. Data represent mean + / - SEM. [Fig. 32A-C] Figures 32A-C: VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 phosphorylation (Y1175), (B) ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) or Src phosphorylation (Y419). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated molecules. 50 ng / ml human VEGF without antibody treatment served as control. Data represent mean + / - SEM. [Figure 33A-C] Figures 33A-C: Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 10 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated molecules: (A) TPP-11736, (B) TPP-14938, (C) TPP-14939. Molecular concentrations are shown in mol / L. VEGF-A uptake was assessed by image-based quantification of HRMEC cell number. Relative cell numbers are shown. Cell number at t=0 was set to 1. Unstimulated cells served as control (basal). Data represent mean + / - SEM. [Fig. 34A-C]Figures 34A-C: VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 phosphorylation (Y1175), (B) ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) or Src phosphorylation (Y419). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated molecules. 50 ng / ml human VEGF without antibody treatment served as control. Data represent mean + / - SEM. [Fig. 35A-D] Figures 35A-D: VEGF-A uptake was assessed by image-based quantification of HRMEC cell number. For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 10 ng / mL human VEGF with or without preincubation with increasing concentrations of (A) TPP-14936, (B) TPP-14937, (C) TPP-14940 or (D) TPP-14941. Molecular concentrations are shown in mol / L. Images were recorded every 4 hours for a total of 84 hours. Relative cell numbers are shown. Cell number at t=0 was set to 1. Non-stimulated cells served as control (basal). Data represent mean + / - SEM. [Fig. 36A-C] Figures 36A-C: VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 phosphorylation (Y1175), (B) ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) or Src phosphorylation (Y419). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated molecules. 50 ng / ml human VEGF without antibody treatment served as control. Data represent mean + / - SEM. [Figure 37A-D]Figures 37A-D: VEGF-A uptake was assessed by image-based quantification of HRMEC cell number. For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 10 ng / mL human VEGF with or without preincubation with increasing concentrations of (A) TPP-14938, (B) TPP-14939, (C) TPP-14940 or (D) TPP-14941. Molecular concentrations are shown in mol / L. Images were recorded every 4 hours for a total of 84 hours. Relative cell numbers are shown. Cell number at t=0 was set to 1. Non-stimulated cells served as control (basal). Data represent mean + / - SEM. [Figure 38] Figure 38: Endothelial sprouting was evaluated by confocal microscopy and spheroid circumferences obtained from maximum projections of Z-stacks were shown. For this purpose, spheroids of human retinal microvascular endothelial cells (HRMEC) were embedded in a collagen matrix. Endothelial sprouting was induced by incubation with 50 ng / mL human VEGF for 24 h with or without 24 h preincubation with 2.5 nM of the indicated molecules. Unstimulated cells served as control (basal). Data represent mean + / - SEM. ns p>0.05 not significant; §p<0.0001 compared to 50 ng / mL hVEGF; #p>0.05 not significant compared to 50 ng / mL hVEGF. ***p<0.001; one-way ANOVA with Tukey's multiple comparison test. [Fig. 39A-B]Figure 39A-B: (A) Time protocol showing experimental procedure. 15 min after intravitreal (ivt) administration of anti-VEGF compounds (13 or 26 pmol / eye Eylea or TPP-14940) or control (26 pmol TPP-11737), 13 pmol / eye human VEGF-A was administered by ivt injection. PBS injection served as control. 24 h later, 1 mL / kg Evans Blue (EB) solution (45 mg / mL in 0.9% saline) was administered by intravenous (iv) injection for 30 min, after which the eyes were isolated and fixed. At the same time points, plasma samples were collected to ensure equal systemic EB exposure. (B) Quantification of VEGF-A-induced hyperpermeability in Brown Norway rat retinas was performed by measuring EB extravasation in retinal flat mounts by confocal microscopy. The eyes were cut along the ora serrata, the lens and vitreous were removed, and the eyecups were fixed in paraformaldehyde (4%) for 1 h at 4°C, then transferred to PBS overnight at 4°C. The retina was separated from the outer segment (sclera and choroid), transferred to a glass slide, and cut four times to achieve a flat cloverleaf-like structure. The tissue was covered with mounting medium (Vectashield H-1200 containing the DNA stain DAPI) and a coverslip was placed on top to obtain a retinal flat mount. The samples were excited at a wavelength of 639 nm, and the emission of Evans blue at 669 nm was recorded with a LSM 700 confocal laser scanning microscope (Carl Zeiss, Jena; gain 800, laser intensity 2%, 5 stacks 60 μm) to obtain images of the retinal flat mount with maximum intensity projection. Analysis of the total fluorescence intensity was performed after opening the images in the program ImageJ with a threshold of 30. ***p<0.001; *p<0.05; ns p>0.05; #p>0.05 Not significantly different from TPP-11737+PBS. One-way ANOVA with Tukey's multiple comparison test. n=9-17. [Fig. 40A-D]Figures 40A-D: (A and B) TrkB phosphorylation (Y706 / 707) or (C or D) ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of BDNF or increasing concentrations of the indicated molecules. [Fig. 41A-B] Figure 41A-B: (A) CHO cells stably expressing human TrkB were incubated with the indicated concentrations of the natural TrkB ligand BDNF (duplicates) or 1 nM BDNF with the indicated concentrations of Doppelmab (each in triplicates). (B) TrkB internalization was assessed by immunofluorescence staining of surface TrkB receptors followed by confocal microscopy analysis. Dark and bright fields in the heatmap represent high and low percentages of cells above the fluorescence threshold, respectively. [Diagram 42] Figure 42: VEGF-A uptake was assessed by measuring VEGF receptor 2 phosphorylation (Y1175). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated molecules. 50 ng / ml human VEGF without molecule treatment served as control. Data represent mean + / - SEM. [Diagram 43] Figure 43: Endothelial sprouting was evaluated by confocal microscopy and spheroid circumferences obtained from maximum projections of Z-stacks were shown. For this purpose, spheroids of human retinal microvascular endothelial cells (HRMEC) were embedded in a collagen matrix. Endothelial sprouting was induced for 24 h by incubation with 50 ng / mL human VEGF with or without preincubation with 2.5 nM of the indicated molecules. Unstimulated cells served as control (basal). Data represent mean + / - SEM. ns p>0.05 not significant; §p<0.0001 compared to 50 ng / mL hVEGF; #p>0.05 not significant compared to 50 ng / mL hVEGF. ***p<0.001; one-way ANOVA with Tukey's multiple comparison test. [Fig. 44A-D]Figure 44A-D: TrkB phosphorylation (Y706 / 707) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of Doppelmab. (A) TPP-22180 vs. TPP-22204; (B) TPP-22192 vs. TPP-22216; (C) TPP-22190 vs. TPP-22214; (D) TPP-22191 vs. TPP-22215. Data represent mean + / - SEM. [Fig. 45A-B] Figures 45A-B: TrkB phosphorylation (Y706 / 707) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of Doppelmab. (A) TPP-22180 vs. TPP-22204; (B) TPP-22192 vs. TPP-22216. Data represent the mean + / - SEM. [Fig. 46A-C] Figure 46A-C: TrkB phosphorylation (Y706 / 707) was measured in CHO cells stably expressing (A) human TrkA, (B) human TrkB or (C) human TrkC after incubation with increasing concentrations of C2 antibody or Doppelmab TPP-22204 or TPP-22214. The natural ligands NGF for TrkA, BDNF for TrkB and NT-3 for TrkC were used as controls. Data represent the mean + / - SEM. [Fig. 47A-B] Figures 47A-B: TrkB phosphorylation (Y706 / 707) was measured in CHO cells stably expressing the human TrkB receptor after incubation with (A) increasing concentrations of C2 antibody with or without a constant concentration of 0.3 nM, 1 nM or 3 nM BDNF or (B) increasing concentrations of Doppelmab TPP-22204 with or without a constant concentration of 0.3 nM, 1 nM or 3 nM BDNF. Data represent the mean + / - SEM. [Fig. 48A-D]Figures 48A-D: TrkB phosphorylation (Y706 / 707) was measured in CHO cells stably expressing the human TrkB receptor after incubation with increasing concentrations of Doppelmab TPP-22214 with or without preincubation with (A) 200 ng / mL human VEGF-A (hVEGF), (B) 50 ng / mL hVEGF, (C) 10 ng / mL hVEGF or (D) 2 ng / mL hVEGF. Data represent the mean + / - SEM. [Fig. 49A-D] Figures 49A-D: ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of Doppelmab TPP-22214 with or without preincubation with (A) 200 ng / mL human VEGF-A (hVEGF), (B) 50 ng / mL hVEGF, (C) 10 ng / mL hVEGF or (D) 2 ng / mL hVEGF. Data represent the mean + / - SEM. [Figure 50] Figure 50: The size of Doppelmab complexes was assessed using size-exclusion chromatography coupled with a multi-angle light scattering detector. Differential refractive index (black, dark grey, light grey and dashed lines) and light scattering were monitored over the time it took for the protein to elute from the size-exclusion column. The light scattering data is used to determine the molar mass at each point. The molar mass at the midpoint of each peak is indicated by an asterisk and measured using the right axis. TPP-22214 alone (dashed line) or in complex with VEGF at various ratios was studied (black 20:1, dark grey 4:1, light grey 1:1). A-D represent schematics of possible complexes based on the measured molar masses. [Fig. 51A-C]Figure 51A-C: TrkB internalization was assessed by immunofluorescence staining of surface TrkB receptors without cell permeabilization followed by confocal microscopy analysis. CHO cells stably expressing human TrkB were incubated with (A) increasing concentrations of the natural TrkB ligand BDNF, (B) increasing concentrations of TPP-22214 or (C) 1 nM BDNF with increasing concentrations of TPP-22214. Data represent the percentage of cells with surface TrkB staining intensity above threshold. Mean + / - SEM. [Figure 52] Figure 52: Neuroprotective function of TrkB activation in a rat model of diabetes-induced retinal neurodegeneration. Animals were treated with STZ to induce hyperglycemia. Retinal function was then assessed by electroretinogram (ERG) and rod-driven B-wave implicit time delay immediately before and 2 weeks after intravitreal application of agonistic TrkB antibody C2 or Doppelmab TPP-11736. Mean + / - SEM; nsp>0.05, not significant (ns); ****p<0.0001; one-way ANOVA with Tukey's multiple comparison test. Anti-TNP served as isotype control antibody. [Fig. 53A-B] Figure 53A-B: VEGF-A uptake was assessed by measuring VEGF receptor 2 phosphorylation (Y1175-VEGFR2). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated molecules. (A) Comparison of Doppelmab TPP-14941, TPP-22216, TPP-22192, TPP-22204 and TPP-22180. (B) Comparison of Doppelmab TPP-14940, TPP-14941, TPP-22190, TPP-22214, TPP-22191 and TPP-22215. Non-stimulated cells (basal) and 50 ng / ml human VEGF without antibody treatment served as controls. Data represent the mean + / - SEM. [Fig. 54A-B]Figure 54A-B: VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 (VEGFR2) phosphorylation (Y1175) or (B) ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of TPP-22214 or Eylea. 50 ng / ml human VEGF without molecular treatment served as control. Data represent mean + / - SEM. [Fig. 55A-E] Figure 55A-E: Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 10 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated molecules: (A) TPPP-22204, (B) TPP-22214c, (C) TPP-22216 or (D) Eylea. Molecule concentrations are shown in mol / L. VEGF-A uptake was assessed by image-based quantification of HRMEC cell number. Images were recorded every 4 hours for a total of 84 hours. Relative cell numbers are shown. Cell number at t=0 was set to 1. Unstimulated cells served as control (basal). Data represent mean + / - SEM. (E) Plots of the difference between the area under the growth curve and the area under the basal curve against the concentration of Eylea or TPP-22214 are shown. [Fig. 56A-B] Figure 56A-B: VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 (VEGFR2) phosphorylation (Y1214) or (B) p38-MAPK phosphorylation (T180 / Y182). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50ng / mL human VEGF with or without preincubation with increasing concentrations of TPP-22214 or Eylea. 50ng / ml human VEGF without molecular treatment served as control. Data represent mean + / - SEM. [Fig. 57A-B]Figure 57A-B: Human retinal microvascular endothelial cell (HRMEC) spheroids were embedded in collagen matrix. Endothelial sprouting was induced by incubation with 50ng / mL human VEGF for 24 hours with or without preincubation with 2.5nM TPP-22204, TPP-22214, TPP-22215, TPP-22216 or 5nM Eylea. Endothelial sprouting was assessed by confocal microscopy and spheroid circumferences obtained from maximum projections of Z-stacks are shown. Unstimulated cells served as control (base). Data represent mean + / - SEM. ns p>0.05 not significant; ****p<0.0001. (B) Representative maximum projection images from spheroids after 24 h of sprouting under basal conditions or after stimulation with 50 ng / mL human VEGF with or without preincubation with 2.5 nM TPP-22214 or 5 nM Eylea are shown. Bar = 100 μm. [Fig. 58A-B] Figure 58A-B: VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 phosphorylation (Y951) or (B) Src phosphorylation (Y419). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50ng / mL human VEGF with or without preincubation with increasing concentrations of TPP-22214 or Eylea. 50ng / ml human VEGF without antibody treatment served as control. Data represent mean + / - SEM. [Fig. 59A-B]Figure 59A-B: TPP-22214 prevented human VEGF-A-induced hyperpermeability in rat retina. (A) Time protocol showing experimental procedure. 15 min after intravitreal (ivt) administration of anti-VEGF compounds (13 or 26 pmol / eye Eylea or TPP-14940) or control (26 pmol TPP-11737), 13 pmol / eye human VEGF-A was administered by ivt injection. PBS injection served as control. 24 h later, 1 mL / kg Evans Blue (EB) solution (45 mg / mL in 0.9% saline) was administered by intravenous (iv) injection for 30 min, after which the eyes were isolated and fixed. At the same time points, plasma samples were collected to ensure equal systemic EB exposure. (B) Quantification of VEGF-A-induced hyperpermeability in Brown Norway rat retinas was performed by measuring EB extravasation in retinal flat mounts by confocal microscopy. The eyes were cut along the ora serrata, the lens and vitreous were removed, and the eyecups were fixed in paraformaldehyde (4%) for 1 h at 4°C, then transferred to PBS overnight at 4°C. The retina was separated from the outer segment (sclera and choroid), transferred to a glass slide, and cut four times to achieve a flat cloverleaf-like structure. The tissue was covered with mounting medium (Vectashield H-1200 containing the DNA stain DAPI) and a coverslip was placed on top to obtain a retinal flat mount. The samples were excited at a wavelength of 639 nm, and the emission of Evans blue at 669 nm was recorded with a LSM 700 confocal laser scanning microscope (Carl Zeiss, Jena; gain 800, laser intensity 2%, 5 stacks 60 μm) to obtain images of the retinal flat mount with maximum intensity projection. Analysis of the total fluorescence intensity was performed after opening the images in the program ImageJ with a threshold of 30. ***p<0.001; *p<0.05; ns p>0.05. One-way ANOVA with Tukey's multiple comparison test. n=9–17. Incubation with Eylea at a molar ratio of 67:1. VEGF is shown for comparison. [Fig. 60A-B]Figure 60A-60B: (A) Functional characterization of TrkB extracellular domain (TrkB-ECD). CHO cells stably expressing human TrkB were incubated with increasing concentrations of the natural ligand BDNF or 10 nM BDNF with increasing concentrations of TrkB-ECD. TrkB activation was assessed by measuring TrkB phosphorylation (Y706 / 707). (B) Effect of TPP-22214 binding to TrkB-ECD on inhibition of VEGF-induced VEGFR2 phosphorylation. Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF, 50 ng / mL human VEGF with increasing concentrations of TPP-22214, or 50 ng / mL human VEGF with increasing concentrations of TPP-22214 and 100 nM TrkB-ECD. HRMEC incubation with increasing concentrations of TrkB-ECD in the presence or absence of 50 ng / mL VEGF and unstimulated cells (basal) served as controls. VEGF-A uptake was assessed by measuring VEGF receptor 2 phosphorylation (Y1175). Data represent the mean + / - SEM. [Fig. 61A-B] Figure 61A-B: (A) TrkB phosphorylation (Y706 / 707) or (B) ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of the indicated molecules. Data represent the mean + / - SEM. [Fig. 62A-C] Figure 62A-C: TrkB phosphorylation (Y706 / 707) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of Doppelmab (A) TPP-23457, (B) TPP-23459 or (C) TPP-6830 antibody with or without preincubation with 200 ng / mL human VEGF-A. Data represent mean + / - SEM. [Fig. 63A-C]Figure 63A-C: ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) was measured in CHO cells stably expressing human TrkB after incubation with increasing concentrations of Doppelmab (A) TPP-23457, (B) TPP-23459 or (C) TPP-6830 antibody with or without preincubation with 200 ng / mL human VEGF-A. Data represent mean + / - SEM. [Fig. 64A-B] Figure 64A-B: TrkB phosphorylation (Y706 / 707) was measured in CHO cells stably expressing the human TrkB receptor after incubation with (A) increasing concentrations of C2 antibody with or without a constant concentration of 0.3 nM, 1 nM or 3 nM BDNF or (B) increasing concentrations of Doppelmab TPP-23457 with or without a constant concentration of 0.3 nM, 1 nM or 3 nM BDNF. Data represent the mean + / - SEM. [Fig. 65A-B] Figure 65A-B: ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2) was measured in CHO cells stably expressing human TrkB after incubation with (A) increasing concentrations of C2 antibody with or without a constant concentration of 0.3 nM, 1 nM or 3 nM BDNF or (B) increasing concentrations of Doppelmab TPP-23457 with or without a constant concentration of 0.3 nM, 1 nM or 3 nM BDNF. Data represent the mean + / - SEM. [Fig. 66A-B] Figure 66A-B: VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 phosphorylation (Y1175) or (B) ERK1 / 2 phosphorylation (T202 / Y204-ERK1) (T185 / Y187-ERK2). For this purpose, human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50ng / mL human VEGF with or without preincubation with increasing concentrations of TPP-22215, TPP-23457, TPP-23459 or Eylea. 50ng / ml human VEGF without molecular treatment served as control. Data represent mean + / - SEM.

[0044] Detailed Description of the Invention The present invention is based on the concept of combining, in a single binding molecule, an antigen-binding site that specifically binds vascular endothelial growth factor (VEGF) with an antigen-binding site that specifically binds tropomyosin receptor kinase B (TrkB).

[0045] It is important to point out that until the present invention, the preparation of a binding molecule targeting these two antigens has not been disclosed or even indirectly contemplated. Thus, the present inventors have prepared a binding molecule comprising at least one antigen-binding site that specifically binds to VEGF and at least one antigen-binding site that specifically binds to TrkB.

[0046] Our initial goal was to design a single binding molecule with comparable efficacy and / or potency when compared to each individual binder, i.e., individual VEGF or TrkB binders that bind only to their respective targets. This was already considered difficult in itself, since it was expected, but not fully understood, that formatting the individual binders within a single binding molecule would adversely affect the original efficacy and / or potency of the individual binders within the single binding molecule.

[0047] Surprisingly, it is found that combining TrkB activation with VEGF capture in a single binding molecule leads to an unexpected increase in both the effectiveness and potency of TrkB activation.The single binding molecule of the present invention shows full TrkB agonist activity, in contrast to the partial agonist activity of individual TrkB binders.Furthermore, the potency of TrkB activation is further increased after VEGF is bound to the single binding molecule.

[0048] Without wishing to be bound by theory, it is believed that VEGF-induced clustering by a single binding molecule of the present invention may be responsible for the observed increase in the potency of TrkB activation. In addition, regardless of the proposed VEGF-induced clustering mechanism, it is clear that the design of the TrkB binding site in the binding molecule of the present invention supports the optimal conformation of TrkB binding that results in the observed full TrkB agonist activity. Advantageously, by combining VEGF scavenging (resulting in inhibition of vascular dysfunction and leakage) with activation of neuroprotective TrkB receptors (resulting in reduced neuronal cell death), patients immediately benefit from a single injection. This is important because retina specialists do not like to inject the same eye with more than one treatment on the same day. Thus, two separate treatments addressing wAMD and GA require two separate treatment visits per eye. In summary, the binding molecules of the present invention are useful for treating and / or preventing loss of visual function, thereby improving quality of life.

[0049] In one aspect, the invention provides binding molecules, in particular molecules having at least one antigen binding site that specifically binds to vascular endothelial growth factor VEGF, preferably VEGF-A, and at least one antigen binding site that specifically binds to tropomyosin receptor kinase B (TrkB), having one or more of the properties described herein below.

[0050] In another aspect, a binding molecule of the invention binds to human TrkB with high affinity. In an embodiment related to this aspect, a binding molecule of the invention binds to human TrkB with high affinity. D In another embodiment, a binding molecule of the invention binds to human TrkB with a K D In another embodiment, a binding molecule of the invention binds to human TrkB with a K D In another embodiment, a binding molecule of the invention binds to human TrkB with a K D In another embodiment, a binding molecule of the invention binds to human TrkB with a K DIn an embodiment related to this aspect, the binding molecule of the invention binds to human TrkB with a K D In another embodiment, a binding molecule of the invention binds to human TrkB with a K D In another embodiment, a binding molecule of the invention binds to human TrkB with a K D In another embodiment, the binding molecules of the invention bind to human TrkB with a K D Binds at <50 nM.

[0051] In another aspect, a binding molecule of the invention activates TrkB with high potency under conditions comprising preincubation with 200 ng / mL human VEGF-A. In an embodiment related to this aspect, a binding molecule of the invention activates human TrkB via EC 50 In a further embodiment, the binding molecules of the invention activate human TrkB at EC 50 In a further embodiment, the binding molecules of the invention activate human TrkB at EC 50 In a further embodiment, the binding molecules of the invention activate human TrkB at EC 50 In a further embodiment, the binding molecules of the invention activate human TrkB at EC 50 In a further embodiment, the binding molecules of the invention activate human TrkB at EC 50 In a further embodiment, the binding molecules of the invention activate human TrkB at EC 50 In a further embodiment, the binding molecules of the invention activate human TrkB at EC 50 In a further embodiment, the binding molecules of the invention activate human TrkB at EC 50 In a further embodiment, the binding molecules of the invention activate human TrkB at EC 50 Active at <10 nM.

[0052] In another aspect, a binding molecule of the invention binds with high affinity to human VEGF, preferably VEGF-A. In an embodiment related to this aspect, a binding molecule of the invention binds with high affinity to human VEGF-A. D In another embodiment, a binding molecule of the invention binds to human VEGF-A with a K D In another embodiment, a binding molecule of the invention binds to human VEGF-A with a K D In another embodiment, a binding molecule of the invention binds to human VEGF-A with a K D In another embodiment, a binding molecule of the invention binds to human VEGF-A with a K D In another embodiment, a binding molecule of the invention binds to human VEGF-A with a K D In another embodiment, a binding molecule of the invention binds to human VEGF-A with a K D Binds at <400 pM.

[0053] In another aspect, a binding molecule of the invention inhibits VEGF-A phosphorylation (Tyr1175) with high potency. In an embodiment related to this aspect, a binding molecule of the invention inhibits human VEGF-A phosphorylation (Tyr1175) with IC 50 In an embodiment of this aspect, the binding molecules of the invention inhibit human VEGF-A phosphorylation (Tyr1175) at IC 50 In an embodiment related to this aspect, the binding molecules of the invention inhibit human VEGF-A phosphorylation (Tyr1175) with an IC 50 In an embodiment related to this aspect, the binding molecules of the invention inhibit human VEGF-A phosphorylation (Tyr1175) with an IC 50 In an embodiment related to this aspect, the binding molecules of the invention inhibit human VEGF-A phosphorylation (Tyr1175) at IC 50 In an embodiment related to this aspect, the binding molecules of the invention inhibit human VEGF-A phosphorylation (Tyr1175) at IC 50In an embodiment related to this aspect, the binding molecules of the invention inhibit human VEGF-A phosphorylation (Tyr1175) at IC 50 In an embodiment related to this aspect, the binding molecules of the invention inhibit human VEGF-A phosphorylation (Tyr1175) at IC 50 Inhibited at <300pM.

[0054] In another embodiment, the binding molecules of the invention are more potent at inducing activation of TrkB downstream signaling pathways than the natural TrkB ligand, BDNF, hi a further embodiment, the binding molecules of the invention regulate gene expression through TrkB-mediated signaling pathways in a pattern comparable to that of BDNF.

[0055] In yet another embodiment, a binding molecule of the invention does not reduce BDNF-induced ERK phosphorylation.In a further embodiment, a binding molecule of the invention is specific for TrkB phosphorylation and / or activation and does not non-specifically phosphorylate / activate TrkA or TrkC.

[0056] In one aspect, the binding molecules of the invention are useful for preventing neurodegeneration and loss of retinal function in disease-relevant animal models. In a further aspect, the binding molecules of the invention stimulate the TrkB-dependent survival signaling pathway, thereby providing neuroprotection, thereby protecting neurons, glial cells and / or neurovascular units in the retina of patients with, for example, macular degeneration, age-related macular degeneration, geographic atrophy or diabetic retinopathy.

[0057] In another aspect, the binding molecules of the invention regenerate axons / dendrites and / or synapses in the retina after onset, e.g., in macular degeneration, age-related macular degeneration, geographic atrophy, or diabetic retinopathy, thereby resulting in neuronal regeneration. In one aspect, the binding molecules of the invention can be formulated at high concentrations for intravitreal injection into the eye.

[0058] In another aspect, the binding molecules of the invention show superior capture of VEGF-A compared to the current anti-VEGF standard of care compound, aflibercept (Eylea), with respect to inhibition of VEGF-induced signaling, endothelial cell proliferation / sprouting and / or VEGF-induced hyperpermeability in rat retina.

[0059] In another embodiment, the binding molecules of the invention exhibit substantially longer vitreous half-lives in rabbit PK studies, e.g., compared to Eylea or other IgG antibodies, and are therefore suitable for quarterly or less frequent injections per year.

[0060] In another aspect, the binding molecules of the present invention are useful for targeting the high unmet need of patients with wAMD and at risk of developing geographic atrophy.In a related embodiment, the binding molecules of the present invention are uniquely useful for treating patients with wAMD and preventing the onset of geographic atrophy.In a further related embodiment, the onset of geographic atrophy is delayed or the severity of the disease is reduced, i.e., the onset and / or progression rate of geographic atrophy is reduced.In this regard, treatment of patients with the binding molecules of the present invention can be started before a definitive diagnosis of geographic atrophy, i.e., for patients at risk of developing geographic atrophy.

[0061] In another embodiment, the binding molecules of the invention are useful for targeting the high unmet need of patients with retinal vein occlusion (CRVO).

[0062] In another embodiment, a binding molecule of the invention improves ERG deficit implicit time when compared to the current standard of care (Eylea).

[0063] In one embodiment, binding of a binding molecule of the invention to TrkB does not trigger receptor internalization, hi a related embodiment, binding of a binding molecule of the invention to TrkB will result in less receptor internalization compared to the natural ligand, BDNF.

[0064] In one embodiment, binding of VEGF-A to a binding molecule of the invention increases the potency of TrkB activation.

[0065] The increase in potency can be measured, for example, by a single binding molecule, e.g., by measuring EC20 values ​​for TrkB or Erk1 / 2 phosphorylation in a suitable cell model after treatment with a single binding molecule, e.g., in CHO cells overexpressing (human) TrkB (see Examples). 50 The EC50 / EC50 binding activity can be measured in vitro by determining the EC50 / EC50 activity of the single binding molecule. Single binding molecules should be tested with or without preincubation of cells with an appropriate concentration, e.g., 200 ng / mL, 50 ng / mL, 10 ng / mL or 2 ng / mL human VEGF-A (hVEGF). After preincubation of cells with human VEGF-A, single binding molecules of the invention have no effect on TrkB or Erk1 / 2 phosphorylation. 50 When measured by , the cells exhibit an increase in potency, at least about a 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 7-fold, 8-fold, 9-fold, or at least about a 10-fold increase in potency, compared to cells that have not been pre-incubated with VEGF-A.

[0066] In another embodiment, the binding molecules of the invention exhibit full TrkB agonist activity, and thus are as effective at activating TrkB as the natural ligand, BDNF.

[0067] In another embodiment, a binding molecule of the invention is capable of simultaneously activating the TrkB receptor and capturing VEGF, preferably VEGF-A.

[0068] definition Terms not specifically defined herein should be given the meaning that would be given to them by one of ordinary skill in the art in light of this disclosure and the context, except that as used herein, the following terms have the meanings indicated and follow the following conventions, unless otherwise indicated.

[0069] As used herein, the term "antigen-binding site" includes a heavy chain variable domain (VH) and a light chain variable domain (VL) derived from an antibody. In such a case, each variable domain comprises three CDRs. In one aspect, the antigen-binding site or specific portion of the protein of the invention is generally derived from an antibody. The general structure of an antibody or immunoglobulin molecule is well known to those skilled in the art.

[0070] "Antibodies" or "immunoglobulin molecules" (also known as immunoglobulins, abbreviated Ig) are gamma globulin proteins that can be found in the blood or other body fluids of vertebrates and are used by the immune system to recognize and neutralize foreign substances, such as bacteria and viruses. They are typically formed from basic structural units, each with two large heavy chains and two small light chains, forming, for example, monomers with one unit, dimers with two units, or pentamers with five units. Antibodies can bind to other molecules or structures, known as antigens, by non-covalent interactions. This binding is specific in the sense that the antibody will bind only to a particular structure with high affinity. The unique part of an antigen that is recognized by an antibody is called the epitope or antigenic determinant. The part of an antibody that binds to an epitope is sometimes called the paratope and is present in the so-called variable domain or variable region (Fv) of the antibody. The variable domain contains three so-called complementarity determining regions (CDRs) spaced by framework regions (FR).

[0071] In the context of the present invention, reference to CDRs is based on various definitions, e.g. CCG, also called IMGT (Lefranc MP, Pommie C, Ruiz M, Giudicelli V, Foulquier E, Truong L, Thouvenin-Contet V, Lefranc G. "IMGT unique numbering for immunoglobulin and T cell receptor variable domains and Ig superfamily V-like domains." Dev Comp Immunol. 2003 Jan;27(1):55-77; Giudicelli V, Brochet X, Lefranc MP. "IMGT / V-QUEST: IMGT standardized analysis of the immunoglobulin (IG) and T cell receptor (TR) nucleotide sequences". Cold Spring Harb Protoc. 2011;2011(6):695-715). An alternative definition of CDR is based on Chothia (Chothia and Lesk, J. Mol. Biol. 1987, 196: 901-917) along with Kabat (EA Kabat, TT Wu, H. Bilofsky, M. Reid-Miller and H. Perry, Sequence of Proteins of Immunological Interest, National Institutes of Health, Bethesda (1983)).

[0072] As used herein, the expressions "variable domain" or "variable region" or Fv refer to each of a pair of light and heavy chains that are directly involved in binding the antibody to the antigen. The variable domain of the light chain is abbreviated as "VL" and the variable domain of the heavy chain is abbreviated as "VH". The variable light and heavy chain domains have the same general structure, each domain containing four framework (FR) regions whose sequences are widely conserved and connected by three HVRs (or CDRs). The framework regions adopt a β-sheet conformation and the CDRs may form loops connecting the β-sheet structure. The CDRs of each chain are held in a three-dimensional structure by the framework regions and, together with the CDRs of the other chain, form the antigen-binding site. The heavy and light chain CDR3 regions of an antibody play a particularly important role in the binding specificity / affinity of the antibodies of the invention and therefore provide a further object of the invention. As used in this application, the term "constant domain" or "constant region" refers to the sum of the domains of an antibody other than the variable regions. Such constant domains and regions are well known in the current state of the art and are described, for example, in Kabat et al. ("Sequence of proteins of immunological interest", US Public Health Services, NIH Bethesda, MD, Publication No. 91).

[0073] The "Fc portion" of an antibody is not directly involved in the binding of the antibody to an antigen, but exhibits various effector functions. "Fc portion of an antibody" is a term well known to those skilled in the art and is defined based on papain cleavage of an antibody. Depending on the amino acid sequence of the constant region of their heavy chain, antibodies or immunoglobulins are divided into classes: IgA, IgD, IgE, IgG and IgM. According to the heavy chain constant region, the different classes of immunoglobulins are called α, δ, ε, γ and μ, respectively. Some of these can be further divided into subclasses (isotypes), e.g., IgG1, IgG2, IgG3 and IgG4, IgAl and IgA2. The Fc portion of an antibody is directly involved in ADCC (antibody-dependent cell-mediated cytotoxicity) and CDC (complement-dependent cytotoxicity) based on complement activation, Clq binding and Fc receptor binding. Complement activation (CDC) is initiated by the binding of the complement factor Clq to the Fc portion of most IgG antibody subclasses. The effect of antibodies on the complement system depends on the particular conditions, but binding to Clq occurs through a defined binding site in the Fc portion. Such binding sites are known in the state of the art and are described, for example, in Boakle et al., Nature 282 (1975) 742-743; Lukas et al., J. Immunol. 127 (1981) 2555-2560; Brunhouse and Cebra, Mol. Immunol. 16 (1979) 907-917; Burton et al., Nature 288 (1980) 338-344; Thommesen et al., Mol. Immunol. 37 (2000) 995-1004; Idusogie et al., J. Immunol.164 (2000) 4178-4184; Hezareh et al., J. Virology 75 (2001) 12161- 12168; Morgan et al., Immunology 86 (1995) 319-324; EP No. 0307434.Such binding sites are, for example, L234, L235, D270, N297, E318, K320, K322, P331 and P329 (numbering according to the EU index of Kabat, see below). The most important of these residues mediating C1q and Fc gamma receptor binding in IgG1 are L234 and L235 (Hezareh et al., J. Virology 75 (2001) 12161- 12168). Antibodies of subclasses IgG1 and IgG3 usually exhibit complement activation and Clq and C3 binding. On the other hand, IgG2 and IgG4 do not activate the complement system and do not bind Clq and C3.

[0074] In the art, antibodies have been further developed and have become versatile tools in medicine and technology. Thus, in the context of the present invention, the term "antibody molecule" or "antibody" or "Ig molecule" (used synonymously herein) not only includes antibodies as they can be found in nature, but also includes antibodies that contain, for example, two light chains and two heavy chains or only two heavy chains as in camelids, and further encompasses all molecules that contain at least one paratope with binding specificity for an antigen and structural similarity to the variable domain of an immunoglobulin.

[0075] Thus, an antibody or immunoglobulin or Ig molecule can include a monoclonal antibody, a human antibody, a humanized antibody, a chimeric antibody, a fragment of an antibody, in particular an Fv, Fab, Fab' or F(ab')2 fragment, a single chain antibody, in particular a single chain variable fragment (scFv), a small modular immunopharmaceutical (SMIP), a domain antibody, a nanobody, a diabody. An antibody may have effector functions, such as ADCC or CDC, which are usually mediated by the Fc portion (antibody constant region) of the antibody, or may have no effector function, for example, by lacking an Fc portion or having a blocked or masked Fc portion, essentially an Fc portion that is not recognized or is not fully recognized by immune cells or immune system components, such as the complement system. A monoclonal antibody (mAb) is a monospecific antibody with identical amino acid sequence. They can be produced by hybridoma technology from hybrid cell lines (called hybridomas) that represent clones of the fusion of a specific antibody-producing B cell with a myeloma (B cell cancer) cell (Kohler G, Milstein C. Continuous culture of fused cells secreting antibody of predefined specificity. Nature 1975; 256:495-7). Alternatively, monoclonal antibodies can be produced by recombinant expression in host cells (Norderhaug L, Olafsen T, Michaelsen TE, Sandlie I. (May 1997). "Versatile vectors for transient and stable expression of recombinant antibody molecules in mammalian cells." J Immunol Methods 204 (1): 77-87; see also below. A "recombinant antibody" or "recombinant binding molecule" is an antibody or binding molecule produced by recombinantly engineered host cells, which may be isolated or purified.

[0076] For applications in humans, it is often desirable to reduce the immunogenicity of antibodies originally derived from other species, e.g. mice. This can be done by constructing chimeric antibodies or by a process called "humanization". In this context, a "chimeric antibody" is understood to be an antibody that comprises a sequence portion (e.g. variable domain) derived from one species (e.g. mouse) fused to a sequence portion (e.g. constant domain) derived from a different species (e.g. human). A "humanized antibody" is an antibody that comprises a variable domain originally derived from a non-human species, in which certain amino acids have been mutated so that the overall sequence of the variable domain more closely resembles the sequence of the human variable domain. Methods for chimerizing and humanizing antibodies are well known in the art (Billetta R, Lobuglio AF. "Chimeric antibodies". Int Rev Immunol. 1993;10(2-3):165-76; Riechmann L, Clark M, Waldmann H, Winter G (1988). "Reshaping human antibodies for therapy". Nature: 332:323).

[0077] A "humanized" antibody refers to an antibody that comprises amino acid residues from a non-human hypervariable region (HVR) and amino acid residues from a human FR. In certain embodiments, a humanized antibody will comprise substantially all of at least one, and typically two, variable domains, in which all or substantially all of the HVRs (e.g., complementarity determining regions (CDRs)) correspond to those of a non-human antibody and all or substantially all of the framework regions (FRs) correspond to those of a human antibody. A humanized antibody can optionally comprise at least a portion of an antibody constant region derived from a human antibody. A "humanized form" of an antibody, e.g., a non-human antibody, refers to an antibody that has been subjected to humanization.

[0078] Furthermore, techniques have been developed for generating antibodies based on sequences derived from the human genome, for example by the use of phage display or transgenic animals (WO 90 / 05144; D. Marks, HR Hoogenboom, TP Bonnert, J. McCafferty, AD Griffiths and G. Winter (1991) "By-passing immunisation. Human antibodies from V-gene libraries displayed on phage." J.Mol.Biol., 222, 581-597; Knappik et al., J. Mol. Biol. 296: 57-86, 2000; S. Carmen and L. Jermutus, "Concepts in antibody phage display". Briefings in Functional Genomics and Proteomics 2002 1(2):189-203; Lonberg N, Huszar D. "Human antibodies from transgenic mice". Int Rev Immunol. 1995;13(1):65-93.; M, Taussig MJ. "Production of human antibody repertoires in transgenic mice". Curr Opin Biotechnol. 1997 Aug;8(4):455-8.) Such antibodies are "human antibodies" in the context of the present invention.

[0079] Antibodies may also include fragments of immunoglobulins that retain antigen-binding properties, such as Fab, Fab' or F(ab')2 fragments. Such fragments can be obtained by fragmentation of immunoglobulins, for example by proteolytic digestion or by recombinant expression of such fragments. For example, immunoglobulin digestion can be achieved by common techniques, such as by using papain or pepsin (WO 94 / 29348). Papain digestion of antibodies typically produces two identical antigen-binding fragments, the so-called Fab fragments, each of which has a single antigen-binding site and a residual Fc fragment. Pepsin treatment results in F(ab')2. In Fab molecules, the variable domains are each fused to an immunoglobulin constant domain, preferably of human origin. Thus, the heavy chain variable domain may be fused to a CH1 domain (the so-called Fd fragment) and the light chain variable domain may be fused to a CL domain. Fab molecules can be produced by recombinant expression of the respective nucleic acid in a host cell. See below.

[0080] Many techniques have been developed to place immunoglobulin variable domains or molecules derived from such variable domains in different molecular contexts, which should also be considered as "antibodies" of the present invention. Generally, these antibody molecules are smaller in size compared to immunoglobulins and may contain a single amino acid chain or several amino acid chains. For example, single chain variable fragments (scFv) are fusions of the variable regions of immunoglobulin heavy and light chains linked together with a short linker, usually serine (S) or glycine (G) (WO 88 / 01649; WO 91 / 17271; Huston et al; International Reviews of Immunology, Volume 10, 1993, 195 - 217). "Single domain antibodies" or "nanobodies" have an antigen-binding site in a single Ig-like domain (WO 94 / 04678; WO 03 / 050531; Ward et al., Nature. 1989 Oct 12;341(6242):544-6; Revets et al., Expert Opin Biol Ther. 5(1):111-24, 2005). One or more single domain antibodies with binding specificities for the same or different antigens can be linked together. Diabodies are bivalent antibody molecules consisting of two amino acid chains containing two variable domains (WO 94 / 13804; Holliger et al., Proc Natl Acad Sci US A. 1993 Jul 15;90(14):6444-8). Other examples of antibody-like molecules are the immunoglobulin superfamily antibodies (IgSF; Srinivasan and Roeske, Current Protein Pept. Sci. 2005, 6(2): 185-96). A different concept leads to the so-called small modular immunopharmaceuticals (SMIPs), which contain an Fv domain linked to a single-chain hinge and effector domain lacking the constant domain CH1 (WO 02 / 056910).

[0081] As used herein, the term "bind" or "specifically bind" refers to the binding of an antibody or an antigen-binding site (e.g., in a binding molecule described herein) to an epitope of an antigen in an in vitro assay. Affinity is the interaction between a single antigen-binding site on an antibody molecule and a single epitope. It is represented by the equilibrium association constant Ka=k / k or the equilibrium dissociation constant K=k / k.

[0082] The binding affinity of an antibody molecule can be increased by a process known as affinity maturation (Marks et al., 1992, Biotechnology 10:779-783; Barbas, et al., 1994, Proc. Nat. Acad. Sci, USA 91:3809-3813; Shier et al., 1995, Gene 169:147-155). Thus, affinity matured antibodies are also encompassed by the present invention.

[0083] An epitope is a region of an antigen that is bound by an antibody or antigen-binding moiety. The term "epitope" includes any polypeptide determinant capable of specific binding to an antibody or antigen-binding moiety. In certain embodiments, epitopic determinants include chemically active surface groups of molecules, such as amino acids, glycan side chains, phosphoryl or sulfonyl, and in certain embodiments may have specific three-dimensional structural characteristics and / or specific charge characteristics. Conformational and non-conformational epitopes are distinguished in that the binding to the former but not the latter is lost in the presence of denaturing solvents. As used herein, the terms "bind" and "specifically bind" refer to the binding of an antibody or antigen-binding moiety to an epitope of an antigen in an in vitro assay using purified wild-type antigen, preferably a plasmon resonance assay (BIAcore®, GE-Healthcare Uppsala, Sweden).

[0084] A "single chain Fv fragment" (scFv) is a polypeptide comprising an antibody heavy chain variable domain (VH), a linker and an antibody light chain variable domain (VL), wherein the antibody domains and the linker have one of the following orders from N-terminus to C-terminus: a) VH-linker-VL, b) VL-linker-VH, wherein the linker is a polypeptide 15 to 25 amino acids in length, preferably 20 amino acids.

[0085] As used herein, the term "identical" or "% identity" in the context of two or more nucleic acid or polypeptide sequences refers to two or more sequences or subsequences that have the same or a certain percentage of nucleotides or amino acid residues that are the same when compared and aligned for maximum correspondence. To determine % identity, the sequences are aligned for optimal comparison purposes (e.g., gaps can be introduced into the sequence of the first amino acid or nucleic acid sequence for optimal alignment with the second amino acid or nucleic acid sequence). The amino acid residues or nucleotides at corresponding amino acid positions or nucleotide positions are then compared. If a position in the first sequence is occupied by the same amino acid residue or nucleotide as the corresponding position in the second sequence, the molecules are identical at that position. The % identity between two sequences is a function of the number of identical positions shared by the sequences (i.e., % identity = number of identical positions / total number of positions (e.g., overlapping positions) x 100). In some embodiments, the two sequences being compared are the same length (e.g., excluding additional sequences that extend beyond the sequence being compared), after gaps are introduced into the sequences, if necessary. For example, when variable region sequences are compared, leader and / or constant domain sequences are not considered. For sequence comparison between two sequences, a "corresponding" CDR refers to the CDR that is in the same position in both sequences (e.g., CDR-H1 of each sequence).

[0086] The determination of the percent identity or percent similarity between two sequences can be achieved using a mathematical algorithm. A preferred, non-limiting example of a mathematical algorithm utilized for comparing two sequences is the algorithm of Karlin and Altschul, 1990, Proc. Natl. Acad. Sci. USA 87:2264-2268, modified in Karlin and Altschul, 1993, Proc. Natl. Acad. Sci. USA 90:5873-5877. Such an algorithm is incorporated into the NBLAST and XBLAST programs of Altschul et al., 1990, J. Mol. Biol. 215:403-410. BLAST nucleotide searches can be performed using the NBLAST program, score=100, wordlength=12, to obtain nucleotide sequences homologous to the nucleic acid encoding the protein of interest. BLAST protein searches can be performed using the XBLAST program, score=50, wordlength=3, to obtain amino acid sequences homologous to the protein of interest. To obtain gapped alignments for comparison purposes, Gapped BLAST can be utilized as described in Altschul et al., 1997, Nucleic Acids Res. 25:3389-3402. Alternatively, PSI-Blast can be used to perform an iterative search that detects distant relationships (Id.) between molecules. When utilizing BLAST, Gapped BLAST, and PSI-Blast programs, the default parameters of each program (e.g., XBLAST and NBLAST) can be used. Another preferred, non-limiting example of a mathematical algorithm utilized for sequence comparison is the algorithm of Myers and Miller, CABIOS (1989). Such an algorithm is incorporated into the ALIGN program (version 2.0), which is part of the GCG sequence alignment software package. When utilizing the ALIGN program for comparing amino acid sequences, a PAM120 weight residue table, a gap length penalty of 12, and a gap penalty of 4 can be used.Further algorithms for sequence analysis are known in the art and include ADVANCE and ADAM as described in Torellis and Robotti, 1994, Comput. Appl. Biosci. 10:3-5, and FASTA as described in Pearson and Lipman, 1988, Proc. Natl. Acad. Sci. USA 85:2444-8. Within FASTA, ktup is a control option that sets the sensitivity and speed of the search. If ktup=2, similar regions in the two sequences being compared are found by looking at pairs of aligned residues. If ktup=1, single aligned amino acids are examined. ktup can be set to 2 or 1 for protein sequences or 1-6 for DNA sequences. If ktup is not specified, the default is 2 for proteins and 6 for DNA. Alternatively, protein sequence alignments can be performed using the CLUSTAL W algorithm, as described in Higgins et al., 1996, Methods Enzymol. 266:383-402.

[0087] A "therapeutically effective amount" of an administered molecule is, for example, the minimum amount necessary to prevent, ameliorate, or treat the clinical symptoms of an ocular or retinal disease or a neurodegenerative disease, particularly the minimum amount effective in these disorders.

[0088] As used herein, the terms "monovalent," "bivalent," and "tetravalent" refer to the number of antigen-binding elements in a binding molecule (one, two, or four, respectively).

[0089] As used herein, the terms "monospecific" and "bispecific" refer to the number (1, 2) of different antigens or epitopes to which a binding molecule specifically binds.

[0090] By way of example, a typical monoclonal antibody (MAb) is bivalent and monospecific, having two antigen-binding arms that both recognize the same epitope. A bispecific antibody has two antigen-binding sites capable of recognizing and binding to two different antigens or epitopes.

[0091] TrkB binding Tropomyosin receptor kinase B (TrkB), also known as tyrosine receptor kinase B or BDNF / NT-3 growth factor receptor or neurotrophic tyrosine kinase receptor type 2, is a protein that in humans is encoded by the NTRK2 gene (Genbank ID: 4915). TrkB is the receptor for brain-derived neurotrophic factor (BDNF).

[0092] Neurotrophic tyrosine kinase receptor B (TrkB; gene symbol: NTRK2) is expressed by retinal neurons and glial cells. In normal retina, TrkB signaling counters cellular stress and promotes cell survival. In diseased eyes, such as diabetic retinopathy or geographic atrophy, there is loss and dysfunction of retinal neurons and glial cells that leads to visual impairment and loss of vision. Activation of TrkB signaling above basal levels (which are reduced in diabetic retinopathy) can counteract the loss and dysfunction of neurons and glial cells, thus improving visual function. Furthermore, TrkB activation has the potential to regenerate lost synaptic connections in diseased eyes, thereby promoting the recovery of visual function. Upon ligand binding, TrkB undergoes homodimerization and subsequent autophosphorylation. Depending on the phosphorylation site (Y516, Y702, Y706, Y707 or Y817), various signaling pathways are activated, including the activity of different subforms of PLCγ1 or AKT and ERK that regulate separate and overlapping signaling cascades that induce axon / neurite outgrowth, improve synaptic plasticity or increase cell survival.

[0093] The TrkB binding component of the binding molecules of the invention, e.g., scFv or scFv2, specifically binds to native or recombinant human TrkB. By binding to TrkB, the binding molecules of the invention activate TrkB signaling and thus, the binding molecules of the invention act as TrkB agonists.

[0094] The binding molecules of the invention can recognize specific "TrkB antigenic epitopes" and "TrkB epitopes." In particular, the binding molecules of the invention bind to an epitope in the extracellular domain of human TrkB. Preferably, the TrkB-binding component of the binding molecule is a TrkB agonist, more preferably a full agonist.

[0095] The extracellular domain of human TrkB essentially comprises the following sequence (SEQ ID NO:226):

[0096] CPTSCKCSASRIWCSDPSPGIVAFPRLEPNSVDPENITEIFIANQKRLEIINEDDVEAYVGLRNLTIVDSGLKFVAHKAFLKNSNLQHINFTRNKLTSLSRKHFRHLDLSELILVGNPFTCSCDIMWIKTLQEAKSSPDTQDLYCLNESSKNIPLANLQIPNCGLPSANLAAPNLTVEEGKSITLSCSVAGDPVPNMYW DVGNLVSKHMNETSHTQGSLRITNISSDDSGKQISCVAENLVGEDQDSVNLTVHFAPTITFLESPTSDHHWCIPFTVKGNPKPALQWFYNGAILNESKYICTKIHVTNHTEYHGCLQLDNPTHMNNGDYTLIAKNEYGKDEKQISAHFMGWPGIDDGANPNYPDVIYEDYGTAANDIGDTTNRSNEIPSTDVTDKTGREH

[0097] As used herein, the terms "TrkB antigenic epitope" and "TrkB epitope" refer to a molecule (e.g., a peptide) or a fragment of a molecule capable of binding to the antigen binding site of a binding molecule or binding molecule fragment of the invention that specifically binds to tropomyosin receptor kinase B. These terms further include, for example, a TrkB antigenic determinant recognized by any of the binding molecules or binding molecule fragments of the invention. A binding molecule or binding molecule fragment of the invention may be: A light chain CDR comprising the amino acid sequences of SEQ ID NO: 201 (CDR1), SEQ ID NO: 202 (CDR2) and SEQ ID NO: 203 (CDR3); Heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 204 (CDR1), SEQ ID NO: 205 (CDR2) and SEQ ID NO: 206 (CDR3); or Heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 207 (CDR1), SEQ ID NO: 208 (CDR2) and SEQ ID NO: 209 (CDR3); or Heavy chain CDRs comprising the amino acid sequences of SEQ ID NO: 210 (CDR1), SEQ ID NO: 211 (CDR2) and SEQ ID NO: 212 (CDR3) The combination of light chain CDRs and heavy chain CDRs is selected from:

[0098] A TrkB antigen epitope may be comprised in a protein, a protein fragment, a peptide, etc. Epitopes are most commonly proteins, short oligopeptides, oligopeptide mimetics (i.e., organic compounds that mimic the antibody binding properties of the TrkB antigen), or combinations thereof.

[0099] VEGF binding Vascular endothelial growth factor (VEGF), also called VEGF-A or vascular permeability factor (VPF), is one of the most important angiogenic factors. VEGF belongs to a gene family that includes placental growth factor (PlGF), VEGF-B, VEGF-C, VEGF-D, VEGF-E and VEGF-F. Alternative splicing of the mRNA of a single human VEGF gene results in at least six isoforms (VEGF121, VEGF145, VEGF165, VEGF183, VEGF189 and VEGF206), with VEGF165 being the most abundant isoform.

[0100] Two VEGF tyrosine kinase receptors (VEGFRs) that interact with VEGF have been identified: VEGFR-1 (also known as Flt-1) and VEGFR-2 (also known as KDR or FIK-1). VEGFR-1 has the highest affinity for VEGF, while VEGFR-2 has a somewhat lower affinity for VEGF. A detailed description of VEGF has been provided by Ferrara (Endocrine Rev. 2004, 25: 581-611), and its interactions with its receptors and its functions in normal and pathological processes can be found in Hoeben et al. Pharmacol. Rev. 2004, 56: 549-580.

[0101] VEGF has been reported to be an important regulator of both normal and abnormal angiogenesis (Ferrara and Davis-Smyth, Endocrine Rev. 1997, 18: 4-25; Ferrara J. MoL Med. 1999, 77: 527-543). Compared to other growth factors that contribute to the process of angiogenesis, VEGF is unique in its high specificity for endothelial cells within the vasculature.

[0102] VEGF is particularly involved in eye diseases.VEGF concentration in ocular fluid is highly correlated with the presence of active proliferation of blood vessels in patients with diabetic retinopathy and other ischemia-related retinopathies.In addition, studies have demonstrated the localization of VEGF in the choroidal neovascular membrane of patients with age-related macular degeneration (AMD).

[0103] The binding molecules of the invention, e.g., the VEGF binding moieties of immunoglobulin (Ig) molecules, have specificity for VEGF in that they specifically bind to one or more epitopes within the VEGF molecule. By binding to VEGF, the binding molecules of the invention act as VEGF antagonists. Preferably, the binding molecules of the invention specifically bind to VEGF-A.

[0104] Specific binding of a VEGF binding component to its antigen, VEGF, can be determined by any suitable method known per se, including, for example, the assays described herein, scatter assays and / or competitive binding assays, such as radioimmunoassays (RIA), enzyme immunoassays (EIA and ELISA) and sandwich competition assays and their various modifications known per se in the art. The same is true when a TrkB binding component binds to its antigen.

[0105] With respect to the antigen VEGF, the VEGF binding moieties, e.g., immunoglobulin (Ig) molecules of the present invention are not limited with respect to species. Thus, the immunoglobulin (Ig) molecules preferably bind human VEGF when intended for therapeutic purposes in humans. However, immunoglobulin (Ig) molecules that bind VEGF from another mammalian species are also within the scope of the present invention. An immunoglobulin (Ig) molecule that binds to a form of VEGF may cross-react with VEGF from one or more other species that has a sequence different from that of humans. For example, an immunoglobulin (Ig) molecule that binds to human VEGF may exhibit cross-reactivity with VEGF from one or more other species of primates and / or VEGF from one or more species of animals used in animal models for diseases (e.g., the species and animal models mentioned herein), particularly diseases and disorders related to VEGF-mediated effects on angiogenesis, e.g., monkey, mouse, rat, rabbit, pig, dog. Such cross-reactive immunoglobulin (Ig) molecules are advantageous in research and / or drug development. This is because it is possible to test the immunoglobulin single variable domains of the invention in recognised disease models, such as monkeys, in particular cynomolgus or rhesus monkeys, or mice and rats.

[0106] Preferably, the VEGF binding component recognizes an epitope in a region of the VEGF of interest that has a high degree of identity with human VEGF, taking into account cross-reactivity with one or more VEGF molecules from species other than human that are intended for use as animal models during the development of therapeutic VEGF antagonists.

[0107] Thus, a binding molecule of the invention, e.g., a VEGF binding component of an immunoglobulin (Ig) molecule, recognizes an epitope located in whole or in part in a region of VEGF that is involved in binding to its receptor. According to a preferred embodiment, the VEGF binding component of a binding molecule of the invention blocks VEGF receptor activation, preferably substantially, and most preferably completely.

[0108] As described above, the ability of a VEGF binding component to block the interaction between VEGF and its receptor can be determined by an amplified luminescence proximity homogeneous assay (AlphaScreen®), a competitive ELISA or a surface plasmon resonance (SPR)-based assay (Biacore®), as described in the Examples.

[0109] TrkB and VEGF binding molecules The present invention relates to binding molecules with binding specificities for at least two different targets. In the context of the present invention, the binding molecules are derived from antibodies. Techniques for preparing binding molecules include, but are not limited to, recombinant co-expression of two immunoglobulin heavy-light chain pairs with different specificities (see Milstein and Cuello, Nature 305: 537 (1983), WO 93 / 08829 and Traunecker et al., EMBO J. 10: 3655 (1991)) and "knobs-in-holes" engineering (see, for example, US Pat. No. 5,731,168). The binding molecules of the invention can also be synthesized by manipulating electrostatic steering effects to prepare antibody Fc-heterodimeric molecules (WO 2009 / 089004); cross-linking two or more antibodies or fragments (see, e.g., US Pat. No. 4,676,980 and Brennan et al., Science, 229: 81 (1985)); using leucine zippers to produce bispecific antibodies (see, e.g., Kostelny et al., Immunol., 148(5): 1547-1553 (1992)); using "diabody" technology to prepare bispecific antibody fragments (see, e.g., Hollinger et al., Proc. Natl. Acad. Sci. USA, 90:6444- 6448 (1993)) and using single-chain Fv (sFv) dimers (see, e.g., Gruber et al., / . Immunol., 152:5368 (1994)) and also by preparing trispecific antibodies as described, for example, in Tutt et al. / . Immunol. 147: 60 (1991).

[0110] In one embodiment, the binding molecule comprises or consists of at least one antigen binding site that specifically binds vascular endothelial growth factor (VEGF), preferably VEGF-A, and at least one antigen binding site that specifically binds tropomyosin receptor kinase B (TrkB). In a related embodiment, the binding molecule is bispecific and tetravalent.

[0111] In another embodiment, the binding molecule comprises or consists of at least one antigen-binding site that specifically binds vascular endothelial growth factor (VEGF) and at least one antigen-binding site that specifically binds tropomyosin receptor kinase B (TrkB), where the at least one antigen-binding site that specifically binds VEGF is an immunoglobulin (Ig) molecule, preferably an IgG molecule. In a related embodiment, the binding molecule is bispecific and tetravalent.

[0112] In another embodiment, the binding molecule comprises or consists of at least one antigen binding site that specifically binds vascular endothelial growth factor (VEGF) and at least one antigen binding site that specifically binds tropomyosin receptor kinase B (TrkB), wherein the at least one antigen binding site that specifically binds TrkB comprises one or more scFvs, preferably in a VL-VH orientation from N-terminus to C-terminus. In a related embodiment, the binding molecule is bispecific and tetravalent.

[0113] In another embodiment, the binding molecule comprises or consists of at least one antigen binding site that specifically binds vascular endothelial growth factor (VEGF) and at least one antigen binding site that specifically binds tropomyosin receptor kinase B (TrkB), where the at least one antigen binding site that specifically binds VEGF is an immunoglobulin (Ig) molecule, preferably an IgG molecule, and where the at least one antigen binding site that specifically binds TrkB is fused to the C-terminus of the heavy chain of the Ig molecule. In a related embodiment, the binding molecule is bispecific and tetravalent.

[0114] In another embodiment, the binding molecule comprises or consists of at least one antigen binding site that specifically binds vascular endothelial growth factor (VEGF) and at least one antigen binding site that specifically binds tropomyosin receptor kinase B (TrkB), wherein the at least one antigen binding site that specifically binds VEGF is an immunoglobulin (Ig) molecule, preferably an IgG molecule, and wherein the at least one antigen binding site that specifically binds TrkB comprises one or more scFvs, preferably in a VL-VH orientation from N-terminus to C-terminus. In a related embodiment, the binding molecule is bispecific and tetravalent.

[0115] In another embodiment, the binding molecule comprises or consists of at least one antigen binding site that specifically binds vascular endothelial growth factor (VEGF) and at least one antigen binding site that specifically binds tropomyosin receptor kinase B (TrkB), where the at least one antigen binding site that specifically binds VEGF is an immunoglobulin (Ig) molecule, preferably an IgG molecule, and where the at least one antigen binding site that specifically binds TrkB comprises one or more scFvs, preferably in a VL-VH orientation from N-terminus to C-terminus, fused to the C-terminus of the heavy chain of the Ig molecule. In a related embodiment, the binding molecule is bispecific and tetravalent.

[0116] In a preferred embodiment, a binding molecule of the invention comprises or consists of (i) two heavy chains, each comprising or consisting of a heavy chain variable region specific for VEGF, a constant IgG domain and an scFv specific for TrkB, and (ii) two light chains, each comprising or consisting of a light chain variable region specific for VEGF.

[0117] In addition, these single chain Fv fragments can be further stabilized by incorporation of disulfide bonds between the VH and VL domains, within the VH domain or within the VL domain through the incorporation of cysteine ​​residues. The term N-terminus refers to the first amino acid of the polypeptide chain, while the term C-terminus refers to the last amino acid at the C-terminus of the polypeptide chain. Thus, an embodiment of the invention is one in which one or more scFvs comprise an additional cysteine ​​residue for disulfide bond formation.

[0118] In one embodiment, the invention provides a binding molecule that is a polyspecific binding protein comprising (i) an Ig molecule having two heavy chains and two light chains and that specifically binds VEGF, and (ii) two scFv molecules (scFvs), each of which specifically binds TrkB. Preferably, each heavy chain of the Ig molecule has one scFv fused to its C-terminus, thereby forming a bispecific, tetravalent binding protein.

[0119] In one embodiment, the present invention provides a method for producing a composition comprising: (i) From the N-terminus to the C-terminus, - a heavy chain variable domain specific for VEGF (e.g., a murine, humanized or human VH domain) - constant domain of IgG (e.g. human IgG1 or IgG4) - a peptide linker (e.g., a GS minilinker) and - an scFv specific for TrkB (e.g., an scFv comprising from N-terminus to C-terminus a VH domain (e.g., a murine, humanized or human VH domain), a linker and a VL domain (e.g., a murine, humanized or human VL domain) or vice versa, a VL domain, a linker and a VH domain); and two heavy chains comprising (ii) each from the N-terminus to the C-terminus a light chain variable domain specific for VEGF (e.g., a murine, humanized or human VL domain); Light chain constant domain (e.g., human kappa chain) and two light chains containing The present invention provides a binding molecule (also referred to herein as a multispecific binding protein or modified Ig molecule) having the following structure:

[0120] The present invention provides binding molecules having at least one antigen binding site that specifically binds vascular endothelial growth factor (VEGF) and at least one antigen binding site that specifically binds tropomyosin receptor kinase B (TrkB).

[0121] Methods for preparing binding sites that bind specific target antigens are well known in the art, and one of skill in the art can readily use these methods to devise antigen-binding sites with the required specificity for VEGF or TrkB target antigens.

[0122] Methods for producing antibodies and antibody fragments are well known in the art. For example, antibodies can be produced by any one of several methods utilizing inducing in vivo production of antibody molecules, screening immunoglobulin libraries (Orlandi et al, 1989. Proc. Natl. Acad. Sci. USA 86:3833-3837; Winter et al 1991, Nature 349:293-299), or the production of monoclonal antibody molecules by cell lines in culture. These include, but are not limited to, the hybridoma technique, the human B cell hybridoma technique, and the Epstein-Barr Virus (EBV)-hybridoma technique (Kohler et al 1975. Nature 256:4950497; Kozbor et al 1985. J. Immunol. Methods 81 :31-42; Cote et al 1983. Proc. Natl. Acad. Sci. USA 80:2026-2030; Cole et al 1984. Mol. Cell. Biol. 62:109-120).

[0123] It will be routine for one of skill in the art to prepare antibodies having binding sites with the requisite specificity for VEGF and / or TrkB target antigens, as well as the binding molecules described herein, using methods known in the art and described herein. Isolation of binding domains from such antibodies is routine practice, and indeed further information regarding methods that can be used to generate the antibodies and binding molecules described herein is provided in the accompanying Examples.

[0124] The inventors have devised binding molecules against VEGF / TrkB according to the invention, as shown in Table 1, and have prepared a selection of these molecules, which are discussed in the accompanying Examples.

[0125] [Table 1] TIFF0007633397000002.tif250169 TIFF0007633397000003.tif247169 TIFF0007633397000004.tif249169 TIFF0007633397000005.tif246169 TIFF0007633397000006.tif248169 TIFF0007633397000007.tif249169 TIFF0007633397000008.tif247169 TIFF0007633397000009.tif249169 TIFF0007633397000010.tif248169 TIFF0007633397000011.tif249169 TIFF0007633397000012.tif248169 TIFF0007633397000013.tif249169 TIFF0007633397000014.tif248169 TIFF0007633397000015.tif249169 TIFF0007633397000016.tif249169 TIFF0007633397000017.tif244169 TIFF0007633397000018.tif243169 TIFF0007633397000019.tif248169 TIFF0007633397000020.tif249169 TIFF0007633397000021.tif249169 TIFF0007633397000022.tif247169 TIFF0007633397000023.tif250169 TIFF0007633397000024.tif248169 TIFF0007633397000025.tif249169 TIFF0007633397000026.tif246169 TIFF0007633397000027.tif248169 TIFF0007633397000028.tif249169 TIFF0007633397000029.tif248169 TIFF0007633397000030.tif248169 TIFF0007633397000031.tif249169 TIFF0007633397000032.tif249169 TIFF0007633397000033.tif243169 TIFF0007633397000034.tif246169 TIFF0007633397000035.tif210169

[0126] In further embodiments, the binding molecule or TrkB binding molecule of the invention may be based on any of the TrkB binders disclosed below, which are also disclosed in WO 2018 / 224630.

[0127] [Table 2] TIFF0007633397000037.tif248169 TIFF0007633397000038.tif245169 TIFF0007633397000039.tif247169 TIFF0007633397000040.tif84169

[0128] In a preferred embodiment, the binding molecule comprises at least one antigen-binding site that specifically binds to vascular endothelial growth factor (VEGF) and at least one antigen-binding site that specifically binds to tropomyosin receptor kinase B (TrkB), wherein the antigen-binding site that specifically binds to TrkB comprises a light chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO:201 (CDR1), SEQ ID NO:202 (CDR2) and SEQ ID NO:203 (CDR3) and a heavy chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO:204 (CDR1), SEQ ID NO:205 (CDR2) and SEQ ID NO:206 (CDR3) or a heavy chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO:207 (CDR1), SEQ ID NO:208 (CDR2) and SEQ ID NO:209 (CDR3) or a heavy chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO:210 (CDR1), SEQ ID NO:211 (CDR2) and SEQ ID NO:212 (CDR3). and a light chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO: 1 (CDR1), SEQ ID NO: 146 (CDR2) and SEQ ID NO: 147 (CDR3), and a heavy chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO: 148 (CDR1), SEQ ID NO: 149 (CDR2) and SEQ ID NO: 150 (CDR3), or a heavy chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO: 151 (CDR1), SEQ ID NO: 152 (CDR2) and SEQ ID NO: 153 (CDR3), or a heavy chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO: 154 (CDR1), SEQ ID NO: 155 (CDR2) and SEQ ID NO: 156 (CDR3).

[0129] In a further preferred embodiment, the binding molecule comprises at least one antigen-binding site that specifically binds vascular endothelial growth factor (VEGF) and at least one antigen-binding site that specifically binds tropomyosin receptor kinase B (TrkB), wherein the antigen-binding site that specifically binds TrkB has at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98%, 99%, 100%, 101%, 102%, 103%, 104%, 105%, 106%, 107%, 108%, 109%, 110%, 111%, 112%, 113%, 114%, 115%, 116%, 117%, 118%, 119%, 120%, 121%, 122%, 123%, 124%, 125%, 126%, 127%, 128%, 129%, 130%, 131%, 132%, 133%, 134%, 135%, 136%, 137%, 138%, 139%, 140%, 141%, 142%, 143%, 144%, 145%, 146%, 147%, 148%, 149%, 150%, 151%, 152%, 153%, 154%, 155%, 156%, 157%, 158%, 159%, 160%, 161%, 162%, 163%, 164%, 165%, 166%, 167%, 168%, 169%, 170%, 171 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 214 or 216, and and a heavy chain variable domain comprising or consisting of an amino acid sequence that is at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 189 or 193. and a heavy chain variable domain comprising or consisting of an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 190, 191, 192 or 194.

[0130] In a further preferred embodiment, the binding molecule comprises at least one antigen-binding site that specifically binds to vascular endothelial growth factor (VEGF) and at least one antigen-binding site that specifically binds to tropomyosin receptor kinase B (TrkB), wherein the antigen-binding site that specifically binds to TrkB comprises a light chain variable domain comprising or consisting of the amino acid sequence of SEQ ID NO: 213 or 215 and a heavy chain variable domain comprising or consisting of the amino acid sequence of SEQ ID NO: 214 or 216, and wherein the antigen-binding site that specifically binds to VEGF comprises a light chain variable domain comprising or consisting of the amino acid sequence of SEQ ID NO: 189 or 193 and a heavy chain variable domain comprising or consisting of the amino acid sequence of SEQ ID NO: 190, 191, 192 or 194.

[0131] In a further preferred embodiment, the binding molecule comprises at least one antigen-binding site that specifically binds to vascular endothelial growth factor (VEGF) and at least one antigen-binding site that specifically binds to tropomyosin receptor kinase B (TrkB), wherein the antigen-binding site that specifically binds to TrkB comprises a light chain variable domain comprising or consisting of the amino acid sequence of SEQ ID NO:213 and a heavy chain variable domain comprising or consisting of the amino acid sequence of SEQ ID NO:214, and wherein the antigen-binding site that specifically binds to VEGF comprises a light chain variable domain comprising or consisting of the amino acid sequence of SEQ ID NO:193 and a heavy chain variable domain comprising or consisting of the amino acid sequence of SEQ ID NO:191.

[0132] In a further preferred embodiment, the binding molecule comprises at least one antigen binding site that specifically binds vascular endothelial growth factor (VEGF) and at least one antigen binding site that specifically binds tropomyosin receptor kinase B (TrkB), wherein the antigen binding site that specifically binds TrkB is an antigen that is at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 213 or 215. and a heavy chain variable domain comprising or consisting of an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 214 or 216, wherein the light chain CDRs in the light chain variable domain are selected from the group consisting of SEQ ID NO: 201 (CDR1), SEQ ID NO: 202 (CDR2) and SEQ ID NO: 203 (CDR3). and the heavy chain CDR in the heavy chain variable domain consists of the amino acid sequences of SEQ ID NO: 204 (CDR1), SEQ ID NO: 205 (CDR2) and SEQ ID NO: 206 (CDR3); or the heavy chain CDR in the heavy chain variable domain consists of the amino acid sequences of SEQ ID NO: 207 (CDR1), SEQ ID NO: 208 (CDR2) and SEQ ID NO: 209 (CDR3); or the heavy chain CDR in the heavy chain variable domain consists of the amino acid sequences of SEQ ID NO: 210 (CDR1), SEQ ID NO: 211 (CDR2) and SEQ ID NO: 212 (CDR3), wherein The antigen-binding site that heterologously binds to the ribozyme comprises a light chain variable domain that comprises or consists of an amino acid sequence that is at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 189 or 193, and a light chain variable domain that comprises or consists of an amino acid sequence that is at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 190, 191, 192 or 194.and a heavy chain variable domain comprising or consisting of an amino acid sequence that is 94%, 95%, 96%, 97%, 98% or at least 99% identical thereto, wherein the light chain CDRs in the light chain variable domain consist of the amino acid sequences of SEQ ID NO: 145 (CDR1), SEQ ID NO: 146 (CDR2) and SEQ ID NO: 147 (CDR3), and the heavy chain CDRs in the heavy chain variable domain consist of the amino acid sequences of SEQ ID NO: 148 (CDR1), SEQ ID NO: 149 (CDR2) and SEQ ID NO: 150 (CDR3), or the heavy chain CDRs in the heavy chain variable domain consist of the amino acid sequences of SEQ ID NO: 151 (CDR1), SEQ ID NO: 152 (CDR2) and SEQ ID NO: 153 (CDR3), or the heavy chain CDRs in the heavy chain variable domain consist of the amino acid sequences of SEQ ID NO: 154 (CDR1), SEQ ID NO: 155 (CDR2) and SEQ ID NO: 156 (CDR3).

[0133] In a further preferred embodiment, the binding molecule comprises at least one antigen-binding site that specifically binds to vascular endothelial growth factor (VEGF) and at least one antigen-binding site that specifically binds to tropomyosin receptor kinase B (TrkB), wherein the antigen-binding site that specifically binds to TrkB comprises a light chain variable domain comprising or consisting of the amino acid sequence of SEQ ID NO: 213 and a heavy chain variable domain comprising or consisting of the amino acid sequence of SEQ ID NO: 214, wherein the light chain CD4 domain in the light chain variable domain R consists of the amino acid sequences of SEQ ID NO: 201 (CDR1), SEQ ID NO: 202 (CDR2) and SEQ ID NO: 203 (CDR3), and the heavy chain CDRs in the heavy chain variable domain consist of the amino acid sequences of SEQ ID NO: 204 (CDR1), SEQ ID NO: 205 (CDR2) and SEQ ID NO: 206 (CDR3), or the heavy chain CDRs in the heavy chain variable domain consist of the amino acid sequences of SEQ ID NO: 207 (CDR1), SEQ ID NO: 208 (CDR2) and SEQ ID NO: 209 (CDR3), or the heavy chain CDRs in the heavy chain variable domain consist of the amino acid sequences of SEQ ID NO: and an antigen-binding site that specifically binds to VEGF comprises a light chain variable domain comprising or consisting of the amino acid sequence of SEQ ID NO: 193, and a heavy chain variable domain comprising or consisting of the amino acid sequence of SEQ ID NO: 191, wherein the light chain CDRs in the light chain variable domain comprise the amino acid sequences of SEQ ID NO: 145 (CDR1), SEQ ID NO: 146 (CDR2), and SEQ ID NO: 147 (CDR3). and the heavy chain CDRs in the heavy chain variable domain consist of the amino acid sequences of SEQ ID NO: 148 (CDR1), SEQ ID NO: 149 (CDR2) and SEQ ID NO: 150 (CDR3), or the heavy chain CDRs in the heavy chain variable domain consist of the amino acid sequences of SEQ ID NO: 151 (CDR1), SEQ ID NO: 152 (CDR2) and SEQ ID NO: 153 (CDR3), or the heavy chain CDRs in the heavy chain variable domain consist of the amino acid sequences of SEQ ID NO: 154 (CDR1), SEQ ID NO: 155 (CDR2) and SEQ ID NO: 156 (CDR3).

[0134] In one embodiment, the binding molecule comprises at least one antigen-binding site that specifically binds to vascular endothelial growth factor (VEGF) and at least one antigen-binding site that specifically binds to tropomyosin receptor kinase B (TrkB), wherein the antigen-binding site that specifically binds to TrkB comprises a light chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO: 228 (CDR1), SEQ ID NO: 229 (CDR2) and SEQ ID NO: 230 (CDR3) and a heavy chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO: 231 (CDR1), SEQ ID NO: 232 (CDR2) and SEQ ID NO: 233 (CDR3) or a heavy chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO: 234 (CDR1), SEQ ID NO: 235 (CDR2) and SEQ ID NO: 236 (CDR3) or a heavy chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO: 237 (CDR1), SEQ ID NO: 238 (CDR3). and a light chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO: 145 (CDR1), SEQ ID NO: 146 (CDR2) and SEQ ID NO: 147 (CDR3), and a heavy chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO: 148 (CDR1), SEQ ID NO: 149 (CDR2) and SEQ ID NO: 150 (CDR3), or a heavy chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO: 151 (CDR1), SEQ ID NO: 152 (CDR2) and SEQ ID NO: 153 (CDR3), or a heavy chain CDR comprising or consisting of the amino acid sequence of SEQ ID NO: 154 (CDR1), SEQ ID NO: 155 (CDR2) and SEQ ID NO: 156 (CDR3).

[0135] In another embodiment, the binding molecule comprises at least one antigen binding site that specifically binds vascular endothelial growth factor (VEGF) and at least one antigen binding site that specifically binds tropomyosin receptor kinase B (TrkB), wherein the antigen binding site that specifically binds TrkB comprises a light chain variable domain and a heavy chain variable domain, each of which is at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NOs: 240 and 241, 242 and 243, 244 and 245, 246 and 247, 248 and 249, 250 and 251, 252 and 253, 254 and 255, or 256 and 257. and wherein the antigen binding site that specifically binds to VEGF comprises or consists of an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 189 or 193. and a heavy chain variable domain comprising or consisting of an amino acid sequence at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or at least 99% identical to the amino acid sequence of SEQ ID NO: 190, 191, 192 or 194.

[0136] In a further preferred embodiment of any of the preceding embodiments, the binding molecules are bispecific and tetravalent. For the avoidance of doubt, by binding the binding molecules of the invention to their respective targets VEGF and TrkB, the binding molecules of the invention act as VEGF antagonists or TrkB agonists, respectively.

[0137] In a further preferred embodiment of any of the preceding embodiments, the antigen binding site that specifically binds VEGF is an immunoglobulin (Ig) molecule and at least one antigen binding site that specifically binds TrkB comprises one or more scFvs.

[0138] In a further preferred embodiment of any of the preceding embodiments, one or more scFvs are fused to the C-terminus of the heavy chain of the Ig molecule.

[0139] In a further preferred embodiment of any of the preceding embodiments, the at least one antigen binding site that specifically binds VEGF is an immunoglobulin (Ig) molecule, more preferably an IgG, and the at least one antigen binding site that specifically binds TrkB comprises one or more scFvs, more preferably two scFvs.

[0140] The antibody molecules or binding molecules described herein may be fused (as fusion proteins) or otherwise linked (covalently or non-covalently) to other molecular entities that have a desired effect on the properties of the antibody molecule. For example, it may be desirable to improve the pharmacokinetic properties of the antibodies or binding molecules described herein, particularly in the case of single chain or domain antibodies, e.g., their stability in body fluids, e.g., blood. In this regard, a number of techniques have been developed, in particular to extend the half-life of such antibody molecules in the circulation, such as pegylation (WO 98 / 25971; WO 98 / 48837; WO 2004081026), fusion or other covalent linkage of the antibody molecule to another antibody molecule having affinity for a serum protein, such as albumin (WO 2004041865; WO 2004003019) or expression of the antibody molecule as a fusion protein with all or part of a serum protein, such as albumin or transferrin (WO 01 / 79258).

[0141] Since the Fc region of an antibody interacts with many Fc receptors resulting in many important functional capabilities (called "effector functions"), in certain embodiments the antibody is a full-length antibody or an antibody containing a portion of the Fc region, as long as the antibody exhibits specific binding to both the relevant portion of the antigen and the Fc receptor and complement. The choice of the type and length of the constant region depends on whether effector functions, such as complement fixation or antibody-dependent cell-mediated cytotoxicity, are desirable features and on the desired pharmacological properties of the antibody protein.

[0142] In an embodiment of the invention, the stability of the scFv portion can be improved by incorporating two cysteine ​​residues in close 3-dimensional proximity to form disulfide bonds within the scFv. To achieve stabilization via engineered disulfide bonds, the residues at these positions are preferably substituted with cysteine ​​residues.

[0143] As demonstrated in the accompanying examples, the inventors have shown that TrkB scFv with an N- to C-terminal VL-VH orientation can function in the binding molecules of the invention to activate TrkB signaling. TrkB scFv with an N- to C-terminal VH-VL orientation can also function, but this orientation may result in reduced activity. Thus, a preferred embodiment of the invention is where the order is N- to C-terminal VL-VH.

[0144] In a further preferred embodiment of the present invention, one or more scFvs that specifically bind to TrkB are fused to an Ig molecule that specifically binds to VEGF (e.g., human IgG1, IgG1(KO), IgG1FcRnmut, IgG4Pro) via a peptide linker, preferably a peptide linker having a length of about 4 to 20 amino acids (e.g., any one of 6, 9, 10, 12, or 15). Preferably, the scFv is fused to the C-terminus of the heavy chain of the Ig molecule. Preferably, the Ig molecule is an IgG.

[0145] Methods for linking scFv molecules to the C-terminus of the heavy chain of an IgG molecule or linking variable domains to scFv molecules are well known in the art. Typically, a small linker sequence of glycine and serine amino acids (called a GS mini-linker) is used. The number of amino acids in the linker can vary from 4 (GGGS) (SEQ ID NO: 227), 6 (GGSGGS) (SEQ ID NO: 217), 10 (GGGGSGGGGS) (SEQ ID NO: 218), 15 (GGGGSGGGGSGGGGS) (SEQ ID NO: 219), 20 (GGGGSGGGGSGGGGSGGGGS) (SEQ ID NO: 220) or more. In practice, the linker is usually formed by combining a nucleic acid molecule encoding the IgG of interest (which in this case would include nucleic acids encoding the heavy chain variable domain and IgG-type constant domains of the VEGF binding site) with a nucleic acid encoding the desired scFv (which in this case would include nucleic acids encoding the heavy and light chain variable domains in either a VL-VH or VH-VL orientation for the TrkB binding site) spaced by a nucleic acid molecule encoding a linker sequence (e.g. a GS mini-linker of any one of 5, 10, 15 or 20 amino acids, preferably a linker of SEQ ID NO: 218). This complete HC-scFv encoding nucleic acid molecule is then placed into an expression vector and introduced into a suitable host cell to form the complete IgG heavy chain-scFv single polypeptide, as further explained below.

[0146] Preferably, the GS minilinker between the scFV molecule and the C-terminus of the heavy chain of the IgG molecule is 10L1 (SEQ ID NO: 218).

[0147] In an embodiment of the invention, binding to the complement product C1q or Fc gamma receptor by a binding molecule of the invention is eliminated by use of an IgG4 constant region or an IgG1 constant region with directed mutagenesis of L to A at positions 234 and 235.

[0148] In an embodiment of the invention, the binding molecules of the invention may have an Fc region or relevant section thereof that is engineered to avoid unintended cross-linking by soluble Fc gamma receptors or complement C1q. In one embodiment, such binding molecules or antibody variants have a much lower affinity for Fc gamma receptors and complement C1q than the parent antibody (hereinafter, unless otherwise stated, the term "parent" in the context of an antibody molecule or in the context of an IgG or Fc region refers to the non-engineered antibody molecule, Fc region or IgG, respectively, from which the mutant (engineered) molecule is derived. Thus, an embodiment of the invention is one in which an Ig molecule comprises an Fc variant that has a reduced affinity for Fc gamma receptors or complement receptors or both, compared to the wild-type Fc region. Such Ig molecules are referred to herein as IgG1(KO).

[0149] Binding molecules comprising Fc regions or relevant sections thereof engineered to modify serum levels (half-life) by optimizing their interaction with the neonatal Fc receptor (FcRn), for example by a point mutation in the CH2 domain at position H310A or a point mutation at position H435A, are also contemplated. Such Ig molecules are referred to herein as IgG1FcRnmut.

[0150] Further contemplated are binding molecules that include Ig molecules that include hinge region mutants of IgG4 that eliminate swapping of heavy chains with other IgG4 molecules. Such Ig molecules are referred to herein as IgG4Pro.

[0151] A further aspect of the invention provides an isolated nucleic acid molecule encoding a binding molecule of the invention or an antibody molecule of the invention, or an expression vector comprising such a nucleic acid molecule.

[0152] In some embodiments, a binding molecule of the invention or an antibody molecule of the invention comprises an antibody heavy and / or light chain polypeptide. As can be appreciated by those skilled in the art, nucleic acid molecules encoding heavy chain polypeptides, light chain polypeptides, or heavy and light chain polypeptides can be readily prepared.

[0153] Nucleic acid molecules encoding the light and heavy chains can be chemically and enzymatically synthesized by polymerase chain reaction (PCR) using standard methods. First, suitable oligonucleotides can be synthesized using methods known in the art (e.g., Gait, 1984) that can be used to generate synthetic genes. Methods for generating synthetic genes from oligonucleotides are known in the art (e.g., Stemmer et al., 1995; Ye et al., 1992; Hayden et Mandecki, 1988; Frank et al., 1987).

[0154] The nucleic acid molecules of the present invention include, but are not limited to, DNA molecules encoding the polypeptide sequences shown in the sequence listing. The present invention also relates to nucleic acid molecules that hybridize to DNA molecules encoding the polypeptide sequences shown in the sequence listing under high stringency binding and washing conditions, as defined in WO 2007 / 042309. Preferred molecules (in terms of mRNA) are those that have at least 75% or 80% (preferably at least 85%, more preferably at least 90%, most preferably at least 95%) homology or sequence identity with one of the DNA molecules described herein. By way of example, with the expression of antibodies in eukaryotic cells in mind, the DNA sequences shown in the sequence listing have been designed to match the codon usage in eukaryotic cells. If it is desired to express the antibodies in E. coli, these sequences can be modified to match the E. coli codon usage. Variants of the DNA molecules of the present invention can be constructed in several different ways, for example as described in WO 2007 / 042309.

[0155] A further aspect of the present invention is (a) culturing a host cell of the invention under conditions that allow expression of a molecule comprising: (b) recovering the molecule, and optionally, The present invention provides a method for producing a binding molecule or antibody molecule described herein, comprising:

[0156] An embodiment of this aspect of the invention is that the process further comprises a step (c) of further purifying and / or modifying and / or formulating the binding molecule of the invention.

[0157] To produce a binding molecule or antibody of the invention, a DNA molecule encoding the full-length light and / or heavy chain or fragments thereof is inserted into an expression vector such that the sequence is operably linked to transcriptional and translational control sequences.

[0158] To produce a binding molecule or antibody of the invention, the skilled artisan can choose from the wide variety of expression systems known in the art, e.g., those reviewed in Kipriyanov and Le Gall, Curr Opin Drug Discov Devel. 2004 Mar;7(2):233-42.

[0159] Expression vectors include plasmids, retroviruses, cosmids, EBV-derived episomes, etc. Expression vectors and expression control sequences are selected to be compatible with the host cell. The antibody light chain gene and the antibody heavy chain gene or the gene for the heavy chain of the binding molecule described herein (e.g., a gene comprising an immunoglobulin heavy chain sequence attached at its C-terminus to an scFv sequence) can be inserted into separate vectors. In certain embodiments, DNA sequences for both the light and heavy chain sequences are inserted into the same expression vector. Convenient vectors are those that encode a functionally complete human CH or CL immunoglobulin sequence, with appropriate restriction sites engineered as described above to allow easy insertion and expression of any VH or VL sequence. The constant chains are usually kappa or lambda for antibody light chains, and can be any IgG isotype (IgG1, IgG2, IgG3, IgG4) or other immunoglobulins, including allelic variants, for antibody heavy chains, but are not limited thereto.

[0160] The recombinant expression vector can also encode a signal peptide that facilitates secretion of the antibody chain (e.g., the heavy and light chains of a binding molecule or antibody described herein) from a host cell. The DNA encoding the antibody chain can be cloned into the vector such that the signal peptide is linked in-frame to the amino terminus of the mature antibody chain DNA. The signal peptide can be an immunoglobulin signal peptide or a heterologous peptide derived from a non-immunoglobulin protein. Alternatively, the DNA sequence encoding the antibody chain (e.g., the heavy and light chains of a binding molecule or antibody described herein) can already contain a signal peptide sequence.

[0161] In addition to the DNA sequence encoding the antibody chains (e.g., the heavy and light chains of the binding molecules or antibodies described herein), the recombinant expression vector carries regulatory sequences including promoters, enhancers, termination and polyadenylation signals and other expression control elements that control the expression of the antibody chains in a host cell. Examples of promoter sequences (exemplified for expression in mammalian cells) are promoters and / or enhancers from (CMV) (e.g., CMV Simian Virus 40 (SV40) (e.g., SV40 promoter / enhancer), adenovirus (e.g., adenovirus major late promoter (AdMLP)), polyoma, and strong mammalian promoters such as the native immunoglobulin promoter and the actin promoter. Examples of polyadenylation signals are BGH polyA, SV40 late or early polyA, alternatively, the 3'UTR of an immunoglobulin gene, etc. can be used.

[0162] Recombinant expression vectors may also carry sequences regulating replication of the vector in host cells (e.g., origins of replication) and selectable marker genes. Nucleic acid molecules encoding the heavy chain or antigen-binding portion thereof and / or the light chain or antigen-binding portion thereof of the binding molecules or antibodies described herein, and vectors containing these DNA molecules, can be introduced into host cells, e.g., bacterial cells or higher eukaryotic cells, e.g., mammalian cells, according to transfection techniques well known in the art, including liposome-mediated transfection, polycation-mediated transfection, protoplast fusion, microinjection, calcium phosphate precipitation, electroporation, or transfer with viral vectors.

[0163] Preferably, the nucleic acid molecules encoding the heavy and light chains of a binding molecule or antibody described herein are present on two vectors that are co-transfected into a host cell, preferably a mammalian cell.

[0164] Thus, a further aspect provides a host cell comprising an expression vector comprising a nucleic acid molecule encoding the heavy chain and an expression vector comprising a nucleic acid molecule encoding the light chain of a binding molecule or antibody described herein.

[0165] Mammalian cell lines available as hosts for expression are well known in the art and include, inter alia, Chinese hamster ovary (CHO, CHO-DG44) cells, NSO, SP2 / 0 cells, HeLa cells, baby hamster kidney (BHK) cells, monkey kidney cells (COS), human cancer cells (e.g., Hep G2), A549 cells, 3T3 cells or derivatives / progeny of any such cell lines. Other mammalian cells can be used, including (but not limited to) human, mouse, rat, monkey and rodent cell lines, or other eukaryotic or prokaryotic cells, including (but not limited to) yeast, insect and plant cells, e.g., bacteria. The binding molecules of the invention are produced by culturing the host cells for a period of time sufficient to allow expression of the binding molecule in the host cells.

[0166] The binding molecules and antibody molecules described herein are preferably recovered from the culture medium as secreted polypeptides or may be recovered from host cell lysates, for example, when expressed without a secretion signal. To obtain substantially homogeneous preparations of the binding molecules or antibodies described herein, the binding molecules or antibody molecules described herein should be purified using standard protein purification methods used for recombinant proteins and host cell proteins. By way of example, state of the art purification methods useful for obtaining the binding molecules and antibodies of the invention include, as a first step, the removal of cells and / or particulate cell debris from the culture medium or lysate. The binding molecules or antibodies are then purified from contaminating soluble proteins, polypeptides and nucleic acids, for example, by fractionation on immunoaffinity or ion exchange columns, ethanol precipitation, reverse phase HPLC, Sephadex chromatography, chromatography on silica or cation exchange resins. As a final step in the process to obtain the VEGF and TrkB single binding molecules described herein, the purified binding molecules can be dried, for example lyophilized, as described below for therapeutic applications.

[0167] A further aspect of the invention provides a binding molecule of the invention for use in medicine, and this use in medicine and for the treatment of the diseases described below is also to be understood as including the TrkB binding molecules, scFvs and antibodies of the invention.

[0168] The binding molecules of the invention are, for example, indicated for use in the treatment / management of ocular or retinal diseases or neurodegenerative diseases, preferably for the treatment of neuro / neuro-ocular or retinal diseases. In a further aspect, the invention relates to a method for the treatment and / or prevention of ocular or retinal diseases or neurodegenerative diseases, comprising administering to a human (e.g., an individual suffering from wAMD or at risk of developing geographic atrophy) an effective amount of a binding molecule of the invention, thereby ameliorating one or more symptoms of the ocular or retinal disease or neurodegenerative disease.

[0169] Terms such as "treatment" and "therapy" as used herein are meant to include therapeutic and prophylactic or suppressive measures of a disease or disorder that result in any clinically desirable or beneficial effect, including (but not limited to) alleviation or relief of one or more symptoms, regression of a disease or disorder, slowing or halting its progression. Thus, for example, the term treatment includes administering a binding molecule before or after the onset of symptoms of a disease or disorder, thereby preventing or eliminating one or more signs of a disease or disorder. As another example, the term includes administering a binding molecule after clinical manifestation of a disease to combat symptoms of the disease. Additionally, administration of a binding molecule after the onset of clinical symptoms and after the onset of clinical symptoms, where administration affects a clinical parameter of a disease or disorder, such as the extent of tissue damage or the amount or extent of metastasis, includes "treatment" or "therapy" as used herein, regardless of whether the treatment results in an improvement of the disease. Furthermore, so long as the compositions of the invention, alone or in combination with another therapeutic agent, reduce or ameliorate at least one symptom of the disorder being treated compared to the symptom in the absence of the binding molecule, the result should be considered an effective treatment of the underlying disorder, regardless of whether all symptoms of that disorder are reduced.

[0170] The binding molecules of the invention can be administered to subjects having or at risk of having an eye or retinal disease. The invention further provides the use of the binding molecules in the manufacture of a medicament for the prevention and / or treatment of an eye or retinal disease. As used herein, the term "subject" refers to any mammalian patient to which the binding molecules can be administered, including, for example, humans and non-human mammals, such as primates, rodents, and dogs. Subjects specifically intended for treatment using the methods described herein include humans. The binding molecules can be administered alone or in combination with other compositions.

[0171] In one embodiment, the binding molecules of the invention are useful in treating macular degeneration, age-related macular degeneration, diabetic retinopathy, diabetic macular edema, retinitis pigmentosa, hereditary retinal dystrophies, hereditary macular dystrophies, myopic degeneration, geographic atrophy, geographic atrophy secondary to age-related macular degeneration, retinal artery occlusion, endophthalmitis, uveitis, cystoid macular edema, choroidal neovascular membrane secondary to any retinal disease, optic neuropathy, glaucoma, retinal detachment, toxic retinopathy, radiation retinopathy. and for use in treating traumatic retinopathy, prodromal and mild to moderate Alzheimer's disease, slowing disease progression in patients with Alzheimer's disease, Huntington's disease, Parkinson's disease, major depressive disorder, schizophrenia, cognitive impairment associated with schizophrenia, preventing first-episode psychosis in individuals with attenuated psychotic syndromes, preventing relapse in patients with schizophrenia, treatment-resistant depression, bulimia, obesity, or metabolic syndrome.

[0172] In a preferred embodiment, the binding molecules of the invention have utility in the treatment of wAMD. More preferably, the binding molecules of the invention have utility in the treatment of wAMD and in the treatment of geographic atrophy. In a further preferred embodiment, the binding molecules of the invention have utility in the treatment of geographic atrophy secondary to age-related macular degeneration. In a further preferred embodiment, the binding molecules of the invention have utility in the treatment of geographic atrophy or in the treatment of patients at risk of developing geographic atrophy. In a further preferred embodiment, the binding molecules of the invention have utility in the prevention of geographic atrophy. In a most preferred embodiment, the binding molecules of the invention have utility in the treatment of wAMD in patients at risk of developing geographic atrophy.

[0173] In another embodiment, the binding molecules of the invention may be useful in the treatment of hearing loss, in particular the treatment of cisplatin-induced hearing loss as well as noise-induced and age-related hearing loss.

[0174] The binding molecules of the invention are administered by any suitable means, including intravitreal, oral, parenteral, subcutaneous, intraperitoneal, intrapulmonary and intranasal. Parenteral infusions include intramuscular, intravenous, intraarterial, intraperitoneal or subcutaneous administration. In addition, the binding molecules of the invention are suitably administered by pulse infusion, particularly in decreasing doses. In one aspect, administration is provided by injection, most preferably intravenous or subcutaneous injection, depending in part on whether administration is brief or chronic. Preferably, the binding molecules of the invention are provided by intravitreal injection into the eye.

[0175] In a further aspect, the binding molecules of the invention are used in combination with a device useful for administration of the binding molecule, such as a syringe, injection pen or other device, hi a further aspect, the binding molecules of the invention are included in a kit of parts, which also includes, for example, a package insert with instructions for use of the binding molecule.

[0176] The effectiveness of the binding molecules of the invention and compositions comprising same can be tested using any suitable in vitro assay, cell-based assay, in vivo assay and / or animal model known per se, or any combination thereof, depending on the specific disease involved. Suitable assays and animal models will be apparent to those skilled in the art and include, for example, the assays and animal models used in the examples below.

[0177] The actual pharmacologic or therapeutic dose will, of course, depend on factors known to those of skill in the art, such as the age and weight of the patient, the route of administration, and the severity of the disease. In any event, the binding molecules of the invention will be administered in a dosage and manner that allows for the delivery of a pharmacologic effective amount based on the patient's unique condition.

[0178] The binding molecules of the invention can be used on their own or in combination with other pharmacologically active ingredients, such as state of the art or standard therapeutic compounds.

[0179] Thus, a further aspect of the invention provides a pharmaceutical composition comprising a binding molecule of the invention together with a pharma- ceutically acceptable carrier and optionally one or more further active ingredients.

[0180] For intravitreal injection of binding molecules, generally, longer intervals between treatments are preferred.In one embodiment, the binding molecules of the present invention are administered every 6 weeks, preferably every 7 weeks, also preferably every 8 weeks, more preferably every 9 weeks, more preferably every 10 weeks, more preferably every 11 weeks, more preferably every 12 weeks.In a more preferred embodiment, the TrkB antibody is administered once every 3 months.Depending on the specific binding molecules of the present invention and their specific pharmacokinetic and other properties, the administration frequency may be longer, for example, every 13 weeks, every 14 weeks, every 15 weeks, every 16 weeks, every 17 weeks, every 18 weeks, every 19 weeks, or every 20 weeks.

[0181] Alternatively, other dosing regimens can be applied, including, for example, a loading dose, where the binding molecule of the invention can be injected monthly for three loading doses, then every 12 weeks. Similarly, the dosing frequency can be extended in such cases, and may be even longer, for example, every 13, 14, 15, 16, 17, 18, 19 or 20 weeks.

[0182] The anticipated estimated human dose of approximately 2.5 mg / eye corresponds to a 50 mg / mL formulation of which 50 μL would be injected intraocularly.

[0183] For therapeutic use, the binding molecules of the invention are formulated into pharmaceutical compositions suitable for easy administration to animals or humans. Exemplary formulations of the binding molecules or antibody molecules described herein can be prepared by mixing the binding molecules with physiologically acceptable carriers, excipients or stabilizers in the form of lyophilized or any dry formulation or aqueous or non-aqueous suspension. The carriers, excipients, modifiers or stabilizers are non-toxic at the dosages and concentrations employed. They may contain buffer systems such as phosphate, citrate, acetate and other inorganic or organic acids and their salts; antioxidants including ascorbic acid and methionine; preservatives such as octadecyldimethylbenzylammonium chloride, hexamethonium chloride, benzalkonium chloride, benzethonium chloride, phenol, butyl alcohol or benzyl alcohol; alkylparabens such as methylparaben or propylparaben; catechol, resorcinol, cyclohexanol, 3-pentanol and m-cresol; proteins such as serum albumin, gelatin or immunoglobulins; hydrophilic polymers such as polyvinylpyrrolidone or polyethylene glycol. The excipients may include: polysaccharides, such as PEG; amino acids, such as glycine, glutamine, asparagine, histidine, arginine or lysine; monosaccharides, disaccharides, oligosaccharides or polysaccharides, including glucose, mannose, sucrose, trehalose, dextrin or dextran, and other carbohydrates; chelating agents, such as EDTA; sugar alcohols, such as mannitol or sorbitol; salt-forming counterions, such as sodium; metal complexes (e.g., Zn-protein complexes) and / or ionic or non-ionic surfactants, such as TWEEN™ (polysorbates), PLURONICS™ or fatty acid esters, fatty acid ethers or sugar esters. Organic solvents, such as ethanol or isopropanol, may also be included in the formulation. Excipients may also have release-modifying or absorption-modifying functions.

[0184] Typically, an aqueous solution or suspension will be preferred. In general, a suitable formulation for a therapeutic protein, e.g., a binding molecule of the invention, will be a buffered protein solution, e.g., a solution containing the protein at an appropriate concentration (e.g., 0.001-400 mg / ml, preferably 0.005-200 mg / ml, more preferably 0.01-200 mg / ml, more preferably 1.0-100 mg / ml).

[0185] However, it will be apparent to one skilled in the art that the above given ingredients and amounts thereof represent only one preferred option, alternatives and variations of which will be immediately apparent to one skilled in the art or can be easily envisaged starting from the above disclosure.

[0186] TrkB binding molecule Based on the surprising finding that the design of the TrkB binding site (two scFvs) in the binding molecule of the invention clearly favors an optimal conformation for TrkB binding and activation (which may be independent of additional antigen binding sites, e.g., a VEGF binding site or a VEGF-induced clustering mechanism), the present invention further relates to a TrkB binding molecule comprising or consisting of two scFvs, wherein each scFv specifically binds to TrkB and both act as TrkB agonists.

[0187] Without wishing to be bound by theory, the inventors believe that the combination of two scFvs in the TrkB binding molecule achieves an optimal conformation that results in the full TrkB agonist activity observed. Thus, in one embodiment, the TrkB binding molecule is a full TrkB agonist, i.e., it is as effective at activating TrkB as the natural ligand, BDNF.

[0188] It is understood that two scFvs in a TrkB-binding molecule are bound to each other to form a TrkB-binding molecule. Two scFvs in a TrkB-binding molecule may be fused or otherwise covalently bound to each other. Alternatively, two scFvs may be linked via a peptide linker, preferably a peptide linker having a length of, for example, about 4 to 20 amino acids, more preferably a flexible peptide linker.

[0189] Each scFv in a TrkB binding molecule can bind to the same or different epitopes within the TrkB protein. Preferably, both scFvs bind to the same TrkB epitope.

[0190] The scFv in the TrkB binding molecule may comprise, from N-terminus to C-terminus, a VH domain (e.g., a murine, humanized or human VH domain), a linker and a VL domain (e.g., a murine, humanized or human VL domain), or vice versa, a VL domain, a linker and a VH domain. Preferably, the scFv in the TrkB binding molecule has a VL-VH orientation from N-terminus to C-terminus.

[0191] In addition, single chain Fv fragments can be further stabilized by incorporation of disulfide bonds between the VH and VL domains, within the VH domain or within the VL domain through incorporation of cysteine ​​residues. The term N-terminus refers to the first amino acid of the polypeptide chain, while the term C-terminus refers to the last amino acid at the C-terminus of the polypeptide chain. Thus, in one embodiment, one or both scFvs contain an additional cysteine ​​residue for disulfide bond formation.

[0192] In further related aspects, the TrkB binding molecule can further comprise an Ig molecule. In this regard, the Ig molecule can be a monoclonal antibody, a human monoclonal antibody, a humanized monoclonal antibody, a chimeric antibody, a fragment of an antibody, such as an Fv, Fab, Fab' or F(ab')2 fragment, a single chain antibody, such as a single chain variable fragment (scFv), a small modular immunopharmaceutical (SMIP), a domain antibody, a nanobody or a diabody.

[0193] The addition of additional Ig molecules supports the optimal conformation of the two scFvs in the TrkB binding molecule to achieve a full TrkB agonist effect. This effect is based on the conformation of the TrkB binding molecule, not on the specificity of the Ig molecule, and therefore is not expected to depend on the specificity of the Ig molecule. Thus, the Ig molecule can be designed to bind to any target. Since the observed full agonist effect is likely based mainly on the conformation of the scFvs in the TrkB binding molecule, it follows that the Ig molecule may not specifically bind to any target.

[0194] In a further preferred embodiment, the Ig molecule comprises or consists of an Fc region. Preferably, in this embodiment, each scFv is fused to the C-terminus of the heavy chain of the Fc region. In a further embodiment of this embodiment, each scFv is fused to the Fc region by a peptide linker, preferably a peptide linker having a length of about 4-20 amino acids.

[0195] It is also understood here that the two scFvs and the Ig molecule in the TrkB-binding molecule are bound to each other to form the TrkB-binding molecule. The scFvs may be directly fused or otherwise covalently bound to the Ig molecule, or they may be linked to the Ig molecule via a linker, preferably a peptide linker having a length of, for example, about 4 to 20 amino acids, more preferably a flexible peptide linker.

[0196] Preferably, the scFv is fused to the C-terminus of the heavy chain of the Ig molecule. Preferably, the Ig molecule is an IgG, F(ab) or F(ab')2. Thus, in a preferred embodiment, the TrkB binding molecule comprises or consists of two scFvs and an IgG, F(ab) or F(ab')2, where each scFv specifically binds to TrkB. In a related preferred embodiment, the TrkB binding molecule is bispecific and tetravalent and comprises or consists of two scFvs and an IgG or F(ab')2, where each scFv specifically binds to TrkB. In a preferred embodiment, the Ig molecule, more preferably an IgG, F(ab) or F(ab')2, specifically binds to VEGF.

[0197] Methods for linking scFv molecules to the C-terminus of the heavy chain of an Ig molecule, such as an IgG molecule, or linking a variable domain to an scFv molecule are well known in the art. Typically, a small linker sequence of glycine and serine amino acids (called a GS mini-linker) is used. The number of amino acids in the linker can vary from 4 (GGGS) (SEQ ID NO: 227), 6 (GGSGGS) (SEQ ID NO: 217), 10 (GGGGSGGGGS) (SEQ ID NO: 218), 15 (GGGGSGGGGSGGGGS) (SEQ ID NO: 219), 20 (GGGGSGGGGSGGGGSGGGGS) (SEQ ID NO: 220) or more. In practice, the linker is typically formed by combining a nucleic acid molecule encoding an Ig of interest with a nucleic acid encoding the desired scFv (which in this case would include nucleic acids encoding the heavy and light chain variable domains in either a VL-VH or VH-VL orientation for the TrkB binding site) spaced by a nucleic acid molecule encoding a linker sequence (e.g., a GS minilinker of any one of 5, 10, 15 or 20 amino acids, preferably the linker of SEQ ID NO: 218). This complete HC-scFv encoding nucleic acid molecule is then placed into an expression vector and introduced into a suitable host cell to form the complete Ig heavy chain-scFv single polypeptide, as described above.

[0198] Preferably, the GS minilinker between the scFv molecule and the C-terminus of the heavy chain of the Ig molecule is 10L1 (SEQ ID NO:218).

[0199] In one embodiment, the TrkB binding molecules induce activation of TrkB downstream signaling pathways more potently than the natural TrkB ligand, BDNF, hi a further embodiment, the TrkB binding molecules regulate gene expression through TrkB-mediated signaling pathways in a pattern comparable to that of BDNF.

[0200] In a further embodiment, the TrkB binding molecule is specific for TrkB phosphorylation and / or activation and does not non-specifically phosphorylate / activate TrkA or TrkC.

[0201] In one embodiment, the TrkB binding molecule comprises: (i) From the N-terminus to the C-terminus, - (optionally) a heavy chain variable domain (e.g., a murine, humanized or human VH domain) - preferably a constant domain of IgG (e.g. human IgG1 or IgG4) - (optionally) a peptide linker (e.g., a GS minilinker); and - an scFv specific for TrkB (e.g., an scFv comprising from N-terminus to C-terminus a VH domain (e.g., a murine, humanized or human VH domain), a linker and a VL domain (e.g., a murine, humanized or human VL domain) or vice versa, a VL domain, a linker and a VH domain and two heavy chains comprising (ii) each from the N-terminus to the C-terminus (optionally) a light chain variable domain (e.g., a murine, humanized or human VL domain); Preferably, the light chain constant domain of IgG (e.g., the human kappa chain) and two light chains containing It comprises or consists of:

[0202] The invention will now be illustrated by the following non-limiting examples.

[0203] Working Example method Cultivation of CHO cells expressing Trk receptors CHO cells expressing human TrkB receptor (ThermoFisher Scientific, #K1491) and customized CHO cells expressing human TrkA or TrkC receptors were cultured in DMEM (Lonza, #BE12-604F) supplemented with 10% fetal bovine serum, glutamax, non-essential amino acids, 20 mM HEPES, 5 μg / mL blasticidin, and 200 μg / mL zeocin. Customized CHO cells expressing cynomolgus monkey, rabbit, or rat TrkB receptors were cultured in a 1:1 mixture of Hams F12 (Lonza #BE12-615F) and DMEM (Lonza #BE12-604F) containing 5% fetal bovine serum, 8 mM glutamine, and 0.5 mg / mL G418. Customized CHO cells expressing the mouse TrkB receptor were cultured in DMEM (Lonza #BE12-604F) supplemented with 10% fetal bovine serum, glutamax, 10 mM HEPES and 0.8 mg / mL G418.

[0204] Analysis of TrkA / B / C and ERK1 / 2 phosphorylation in CHO cells expressing human, cynomolgus monkey, rabbit, rat or mouse Trk receptors Five thousand CHO cells expressing each Trk receptor were seeded into each cavity of a 384-well clear tissue culture plate (BD Falcon, #353963) and incubated in a humidified incubator at 37°C and 5% CO2. 24 hours after seeding, cell supernatants were replaced with room temperature starvation medium (DMEM with 0.1% BSA (Sigma, #A-3059) but no other supplements). After 15 minutes, starvation medium containing increasing concentrations of human BDNF (R&D #248-BD or Bachem #H-5594), human NGF (Biovision #4303R-20), human NT-3 (Sigma #N1905), agonistic antibodies or isotype controls were added in triplicate for 45 minutes at room temperature to stimulate TrkB and ERK1 / 2 phosphorylation. Starvation medium alone served as control.

[0205] In some experiments, BDNF or agonistic antibodies were preincubated for 1 hour with or without 2, 10, 50 or 200 ng / mL human VEGF (R&D Systems #293-VE-050) prior to cell stimulation. In these experiments, incubation with human VEGF alone served as a control.

[0206] To analyze whether agonistic TrkB antibodies limit BDNF-induced phosphorylation of TrkB and / or downstream ERK1 / 2 phosphorylation, CHO cells expressing the human TrkB receptor were incubated with increasing concentrations of antibodies without BDNF or at constant concentrations of 0.3 nM, 1 nM, or 3 nM BDNF.

[0207] After stimulation, cell supernatants were removed and cells were lysed for 20 min on wet ice in lysis buffer (1× Triton lysis buffer (Cell Signaling Technology #9803-S) supplemented with Complete Mini Protease Inhibitor Tablets (Roche #04693124001) and phosphatase inhibitor cocktails 2 (Sigma #P5726) and 3 (Sigma #P0044) and 1 mM PMSF (Sigma #93482). The resulting lysates were analyzed using a commercial assay (Perkin Elmer) according to the manufacturer's instructions. The assay was used to quantify TrkA phosphorylation at Y680 / 681, TrkB phosphorylation at Y706 / 707 or TrkC phosphorylation at Y709 / 710 using Perkin Elmer #ALSU-PTRKAB-A10K. Quantification of ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2) was similarly performed using another commercial assay (Perkin Elmer #TGRES10K or ALSU-PERK-A10K) following the manufacturer's instructions. Emission of the acceptor beads, reflecting phosphorylation events, was recorded at 570 nm in a Perkin Elmer EnVision microplate reader. Data including raw data (mean ± SEM) and nonlinear regression (log(agonist) vs. response (three parameters)) were prepared for presentation with GraphPad Prism (version 8).

[0208] Functional characterization of the TrkB extracellular domain (TrkB-ECD) CHO cells stably expressing human TrkB were incubated with increasing concentrations of the natural ligand BDNF or 10 nM BDNF with increasing concentrations of TrkB-ECD (R&D Systems #1494-TB). TrkB activation was assessed by measuring TrkB phosphorylation at Y706 / 707 as outlined above. Data represent the mean + / - SEM.

[0209] TrkB receptor internalization assay 24 hours prior to stimulation, 25,000 CHO cells expressing human TrkB receptors were seeded into each cavity of a 96-well black clear bottom tissue culture plate (PerkinElmer, #6055300) and incubated in a humidified incubator at 37°C and 5% CO2. The supernatant of CHO / hTrkB cells was replaced with starvation medium (DMEM with 0.1% BSA (Sigma, #A-3059) and 20 mM Hepes (Lonza, #BE17-737F)) and the cells were incubated for 30 minutes at 37°C. The cells were then stimulated with increasing concentrations of BDNF (R&D #248-BD or Bachem #H-5594) or agonistic TrkB antibodies alone or in combination with 1 nM BDNF and increasing concentrations of agonistic antibodies in starvation medium for 50 minutes at 37°C. Cells were fixed in 4% paraformaldehyde for 20 min, washed, blocked in 5% normal donkey serum in PBS for 1 h, and incubated with 1 μg / mL goat anti-TrkB antibody (R&D Systems, #AF-397) overnight at room temperature, followed by extensive washing and incubation in 2 μg / mL AlexaFluor 647 donkey anti-goat antibody (ThermoFisher Scientific, #A21447) and 1 μg / mL Hoechst #H3570 for 2 h at room temperature. After extensive washing, cells were stained with 2 μg / mL HCS-Cellmask green (ThermoFisher Scientific, #H32714) in PBS / 0.05% Tween-20 for 1 h. Cell surface receptors were imaged with a PerkinElmer Opera Phenix High Content Screening System equipped with a 20x water objective and analyzed with PerkinElmer Harmony High-Content Imaging and Analysis software. Data are presented either as heat maps with dark and bright fields of the heat map representing high and low percentages of cells above a fluorescence threshold, respectively, or as a diagram representing the percentage of cells with surface TrkB staining intensity above a threshold.In the latter case, data including raw data (mean±SEM) and nonlinear regression (log(agonist) vs. response (three parameters)) were prepared for presentation with GraphPad Prism (version 8).

[0210] Inhibition of VEGF-induced VEGF receptor 2 phosphorylation and downstream signaling Human retinal microvascular endothelial cells (HRMEC; Cell Systems #ACBRI181) were cultured on gelatin (Millipore #ES-006B) coated plates in endothelial cell basal medium (Promocell #C-22210) with supplements (Promocell #C-39210) and 10 U / mL penicillin / streptomycin, respectively. For analysis of VEGF capture, 12000 cells were seeded into each cavity of a 96-well clear tissue culture plate using regular growth medium and incubated in a humidified incubator at 37°C and 5% CO2. After 24 hours, the medium was replaced with starvation medium (Promocell #C-22210 supplemented with 0.1% BSA (Sigma, #A-3059) and 10 U / mL penicillin / streptomycin) and the cells were starved for 20 hours in a humidified incubator at 37°C and 5% CO2. Prior to stimulation of VEGFR2 signaling, 50ng / mL human VEGF (R&D Systems #293-VE-050) was pre-incubated with or without antagonistic antibodies or Eylea for 30-60 minutes at room temperature in starvation medium, then pre-warmed to 37°C for 15 minutes. Cell stimulation was achieved by adding preformed VEGF-antibody / Eylea complexes or VEGF alone to the cells for 5 minutes on plates warmed to 37°C. Starvation medium alone served as a control. Cells were lysed in lysis buffer (Perkin Elmer #ALSU-PVGFR-A500) for 10 minutes at room temperature, then incubated on ice for an additional 10 minutes.Commercially available assays were used to quantitate VEGFR2 phosphorylation at Y1175 (Perkin Elmer #ALSU-A500 or #ALSU-PVGFR-A10K), VEGFR2 phosphorylation at Y951 (Perkin Elmer #ALSU-PVGFR-B500), ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2) (Perkin Elmer #TGRES10K or ALSU-PERK-A10K), Src phosphorylation at Y419 (Perkin Elmer #ALSU-PSRC-A10K), p38-MAPK phosphorylation at Thr180 / Tyr182 (Perkin Elmer #ALSU-PP38-B500) according to the manufacturer's instructions. Emission of acceptor beads reflecting the phosphorylation events mentioned above was recorded at 570 nm in a Perking Elmer EnVision microplate reader and data including raw data (mean ± SEM) and nonlinear regression (log(agonist) vs. response (three parameters)) were prepared for presentation by GraphPad Prism (version 8).

[0211] To examine the effect of TrkB binding on VEGF uptake by bispecific and tetravalent Doppelmab, inhibition of VEGF-induced VEGFR2 phosphorylation was assessed in starved HRMECs in the absence or presence of TrkB-ECD. Prior to stimulation, increasing concentrations of each Doppelmab were incubated with 100 nM TrkB-ECD (R&D Systems #1494-TB) in starvation medium for 1 h at room temperature, followed by an additional 1 h incubation with 50 ng / mL human VEGF. As controls, HRMECs were incubated with (i) starvation medium alone, (ii) 50 ng / mL human VEGF, (iii) 50 ng / mL human VEGF preformed complexes with increasing concentrations of TrkB-ECD (1 h at room temperature), (iv) 50 ng / mL human VEGF preformed complexes with increasing concentrations of each Doppelmab (1 h at room temperature), and (v) increasing concentrations of TrkB-ECD alone. VEGF-A uptake was assessed by measuring VEGF receptor 2 (VEGFR2) phosphorylation at Y1175 as outlined above.

[0212] Inhibition of VEGF-induced HRMEC proliferation 3000 HRMECs were seeded into each cavity of a clear flat-bottom 96-well plate using endothelial cell basal medium (Promocell #C-22210) with supplements (Promocell #C-39210) and 10 U / mL penicillin / streptomycin and incubated in a humidified incubator at 37°C and 5% CO2. After 16 hours, the growth medium was replaced with starvation medium (endothelial cell basal medium supplemented with 2% fetal bovine serum (Promocell #C-22210)) and cells were starved for 8 hours in a humidified incubator at 37°C and 5% CO2. Prior to stimulation, human VEGF (R&D Systems #293-VE-050) was pre-incubated for 60 minutes in starvation medium at room temperature with or without antagonistic antibodies or Eylea. Cell stimulation was performed by adding preformed VEGF-antibody / Eylea complexes or VEGF alone to the cells. Starvation medium alone served as control (basal proliferation). Cell proliferation was assessed by automated phase contrast image-based quantification of total HRMEC nuclear area (Essen Bioscience, IncuCyte S3). Total HRMEC nuclear area was considered proportional to HRMEC number. Four images per well were recorded every 4 hours for a total of 96 hours using a 10x objective. Data represent relative cell number (cell number at time t / cell number at t=0) as a function of time, where the cell number at t=0 was set to 1. In some experiments, the area between each growth curve and the basal growth curve (growth in starvation medium) was plotted against the base 10 logarithm of each compound concentration. Where appropriate, data including raw data (mean ± SEM) and nonlinear regression (log(agonist) vs. response (three parameters)) were prepared for presentation with GraphPad Prism (version 8).

[0213] Inhibition of VEGF-induced HRMEC sprouting Inhibition of VEGF-induced sprouting was evaluated in an HRMEC-based spheroid assay. HRMEC were resuspended in regular growth medium containing 20% ​​Methocoel (1.2% methylcellulose and 10% fetal bovine serum in endothelial basal medium) and 25 μl drops containing 500 HRMEC were applied to square Petri dishes. The plates were turned upside down and the cells were cultured in hanging drops. This allowed the spontaneous formation of spheroids. After 24 hours, spheroids were collected and embedded in a Methocoel-collagen mixture (80% Methocoel and 20% FCS 1:1 mixed with 3 mg / mL rat tail collagen I (Corning #354236) in M199 medium) in a humidified incubator at 37° C. and 5% CO2 for 30 minutes to achieve collagen polymerization. Prior to stimulation, human VEGF (R&D Systems #293-VE-050) was pre-incubated in basal endothelial growth medium supplemented with 2% fetal bovine serum with or without antagonistic antibodies or Eylea for 60 min at room temperature. Sprouting was stimulated by adding preformed VEGF-antibody / Eylea complexes or VEGF alone to the spheroids for 24 h. Basal endothelial growth medium supplemented with 2% fetal bovine serum and without VEGF served as a control. After fixation in 4% paraformaldehyde, cells were washed extensively and stained overnight with 2 μg / mL HCS-Cellmask green (ThermoFisher Scientific, #H32714) in PBS containing 0.2% Triton X-100. Spheroids were analyzed on a ZEISS LSM 780 confocal microscope. Three-dimensional sprouting of HRMEC was quantified from maximum projection images of Z-stacks either manually and expressed as cumulative sprout length per spheroid (mm) or semi-automated with Zeiss ZEN imaging software and expressed as spheroid circumference (pixels). Data were prepared for presentation with GraphPad Prism (version 8) and represent the mean + / - SEM.

[0214] Electroretinogram Common methods Electroretinogram (ERG) is a non-invasive electrophysiological technique for assessing light-evoked electrical activity of various retinal neurons, allowing for the quantification of various aspects of retinal function, such as dim light or color vision. ERG was measured as the change in potential between the cornea and a reference electrode using an Espion E3 ERG recording system (Diagnosys LLC). Prior to ERG recording, animals were dark-adapted for at least 2 hours and anesthetized with an ip injection of ketamine (Ketanest, approx. 100 mg / kg) and xylazine (Rompun, approx. 7 mg / kg). Animals were placed on a heated stage to maintain a constant body temperature at 37°C. Pupils were dilated with topical tropicamide and 10% phenylephrine. A drop of 2% Methocel solution (OmniVision) was placed on the cornea to prevent the corneal eye from drying out during recording. Recordings were made simultaneously in both eyes using gold loop electrodes. The reference electrode was a toothless alligator clip moistened with Methocel and attached to the animal's cheek. For electrical ground, a clip was attached to the animal's tail. ERG signals were sampled at 1 kHz and recorded using a low frequency of 0.15 Hz and a high frequency cutoff of 500 Hz. Light stimuli were delivered by a set of light-emitting diodes or a xenon bulb (1 cd s / m 2 The study consisted of full-field flashes (approximately 4 ms duration) delivered by a Ganzfeld stimulator (ColorDome; Diagnosys) in either darkness or background light.

[0215] ERG Protocol First, ERG responses were recorded from dark-adapted animals (to isolate rod-driven responses) followed by recording in red background light (50 cd / m 2 , to isolate ultraviolet cone-driven ERG responses) and finally, green-blue background light (25.5 cd / m 2 , to isolate M-cone-driven ERG responses).

[0216] For dark-adapted ERGs, we scaled the response to 1·10 -5~100 cd·s / m 2 The trigger was a series of flashes in the range of 1·10 -5 and 3·10 -5 cd·s / m 2 Responses were averaged over 20 trials for a flash with a luminance of 0.05 cd s / m 2 1 to 10 -4 For the inter-flash, responses were averaged over 10 trials and were measured at 0.1 cd s / m 2 For each flash, responses were averaged over eight trials and expressed as 1 cd s / m 2 Average of 5 trials for a flash of 10 cd s / m 2 The responses of three trials were recorded for the final flash of 100 cd s / m 2 For the last flush of, one flash was recorded.

[0217] The interval between individual flashes was chosen to ensure that the retina fully recovered from each flash (no signs of flash-induced reduction in response amplitude or shortening of implicit time). Based on these criteria, the interval between flashes was chosen to be 1 10 -5 and 3·10 -5 cd·s / m 2 Flash for about 2 seconds, 0.05 cd·s / m 2 1 to 10 -4 Between flashes about 5 seconds, 0.1cd·s / m 2 Flash for 10 seconds and 1 cd·s / m 2 The flash duration was 20 seconds. 10cd·s / m 2 and 100 cd·s / m 2 The single flush was followed by a recovery period of 30 and 120 seconds, respectively.

[0218] For recording of UV-cone-driven and M-cone-driven responses, animals were first photoadapted to red background light and then to green background light for 2 min. Light responses were recorded at 0.02, 0.04, 0.08, 0.17, 0.35, 0.83, 1.66, 2.90 and 4.15 μW / m 2Each UV flash and 0.1 to max. 110 cd·s / m 2 The responses were elicited by an M-cone flash of 10 flashes. All responses from 10 trials were averaged with an inter-flash interval of 3 seconds.

[0219] Animals and STZ treatment Male Brown Norway rats (BN rats) were obtained from Charles River (Germany). Hyperglycemia was induced by ip injection of STZ (65 mg / kg body weight). Non-responders or poor responders, i.e., animals with blood glucose concentrations below 20 mM 7 days after STZ application, were not included in the study. Body weight and blood glucose levels were monitored regularly. STZ was administered approximately 3 weeks before intravitreal administration.

[0220] Dosage and intravitreal injection For ivt injections, rats were anesthetized with 2.5–3% isoflurane (Forene; Abbvie). For local anesthesia, a drop of 4 mg / mL oxybuprocaine hydrochloride (Novesine; Omnivision) was administered. 5 μL was injected intravitreally just posterior to the limbus in each eye with a 34-gauge needle (attached to a 10 μL Hamilton glass syringe).

[0221] Data analysis ERG was measured as the potential change between the cornea and a reference electrode using an Espion E3 ERG recording system (Diagnosys LLC).

[0222] Each ERG flash from the different sets was verified to be consistent before calculations using Matlab software, after which a mean curve was calculated from each flash.

[0223] To determine ERG a / b-wave amplitudes, ERG data were processed and analyzed using MATLAB software (version R2014a; MathWorks), where data were sorted with individual macros to create files for Matlab routines.

[0224] The b-wave amplitude was calculated from the trough of the a-wave response to the peak of the b-wave peak. The b-wave implicit time was measured as the time after the flash stimulus required to reach the b-wave peak.

[0225] The b-wave amplitude as a function of stimulus intensity was fitted by using a least-squares fitting method (GraphPad Prism, version 6.01 and later versions of GraphPad Prism). The a-wave amplitude was calculated from the baseline (zero line) to the negative a-wave response.

[0226] Statistical analysis was performed by one-way analysis of variance.

[0227] Example 1 Global design of a binding molecule that recognizes human VEGF and human TrkB Figure 1 The present inventors have developed binding molecules that bind to VEGF and TrkB and act as VEGF antagonists and TrkB agonists, respectively. The molecular design used has an IgG antibody with specificity for one target antigen (called a "master antibody") with scFvs of different specificity attached to the C-terminus of the heavy chain. The design scheme is shown in Figure 1. Preferably, the binding molecule is bispecific and tetravalent.

[0228] The bispecific molecule contains a flexible peptide sequence between the variable heavy (VH) and variable light (VL) domains of the scFv, which is linked to a master IgG antibody via an additional series of linkers. In one configuration, the scFv is oriented such that the VL domain forms the "N-terminus" of the scFv and is thus fused to the C-terminus of the heavy chain of the master antibody, while the VH forms the C-terminus of the scFv, and indeed the entire heavy chain polypeptide. However, it can be understood that this "N-VL-VH-C" structure can be reversed, i.e. "N-VH-VL-C".

[0229] To test the feasibility of this concept, a number of different bispecific molecules based on the format depicted in Figure 1 were prepared and further optimized over several cycles of molecular design starting from series 1 to series 4.

[0230] The Examples below describe a diverse series of bispecific molecules that bind VEGF and TrkB and the methods used to generate variations of this format, as well as the biological activity of these molecules.

[0231] Example 2 Preparation of binding domains that recognize VEGF and TrkB Figure 1 As can be seen, in order to prepare a bispecific molecule that binds human VEGF and TrkB, it is necessary to obtain the variable domains that bind to their respective target antigens.

[0232] For this purpose, immunoglobulin (Ig) VH and VL genes were obtained from various VEGF and TrkB binders and formatted into the bispecific molecules of the invention. Overall, three VEGF binding domains were obtained from individual VEGF binders called B20, G6 and Ranibizumab. For binding to TrkB, a binding domain was obtained from an individual TrkB binder called C2 (WO2010086828).

[0233] Formatting of individual binders into bispecific molecules of the invention is performed by routine methods known to those skilled in the art. Briefly, to construct gene segments encoding scFvs, pairs of VL and VH genes encoding the variable domains were linked by a gene segment encoding a flexible linker of the peptide sequence GGSEGKSSGSGSESKSTGGS (SEQ ID NO: 221). The resulting gene segment encoding scFvs was then cloned in frame at the 3' end of a gene encoding the heavy chain of a human IgG antibody. These coding segments were synthesized by overlap PCR and cloned into the expression vector pTT5.

[0234] Next, Fab (2) The VL and VH gene pairs encoding the moieties were formatted into the bispecific format outlined in Example 1. The VH gene was cloned into the pTT5 expression vector as an in-frame fusion at the 5' end of the gene encoding human Igγ. The gene encoding the scFv binder was cloned in frame at the 3' end of the same Igγ-encoding segment. Similarly, the VL gene was cloned into the pTT5 expression vector as an in-frame fusion with the gene encoding the human IgG kappa light chain.

[0235] The In-Fusion® HD Cloning Kit (Clonetech, USA) was used in the above procedure for directional cloning of the VH and VL genes. PCR primers for VL / VH were synthesized with 15 bp extensions complementary to the ends of the linearized vector. PCR was performed using the manufacturer's standard protocol and the amplicons were purified or treated with cloning enhancers and then cloned into the appropriate vector. E. coli was then transformed according to the manufacturer's instructions (Clonetech, USA). DNA minipreps were sequenced.

[0236] Each expression vector contains a eukaryotic promoter element for the gene encoding the chain, a signal sequence and a gene encoding the heavy or light chain, an expression cassette for a prokaryotic selection marker gene, e.g., ampicillin, and an origin of replication. These DNA plasmids were propagated in ampicillin-resistant E. coli colonies and purified.

[0237] The expression vector was transfected into CHO-E cells. The transfected CHO-E cells, growing in suspension in serum-free medium, were cultured in shake flasks with agitation at 140 rpm at 37°C and 5% CO2 and maintained in exponential growth conditions. On the day of transfection, the cells were chemically transfected with 1 mg light chain plasmid and 0.5 mg heavy chain plasmid. They were then cultured at 1-2 x 10 6 cells / mL in 1 L of Gibco® FreeStyle™ CHO Expression Medium (LifeTechnologies, NY, US). 6 Cells were then incubated under orbital shaking for 10-12 days and fed once with 150 ml of commercial feeding solution to allow protein expression. Binding molecule titers in cell culture supernatants were determined using an Octet® instrument (Pall ForteBio, CA, US) and a protA biosensor chip according to the manufacturer's instructions.

[0238] Recombinant binding molecules were purified from culture supernatants by Protein A affinity chromatography using MabSelect™ SuRe™ (Cytiva) followed by either size exclusion chromatography (Superdex 200, Cytiva) or cation exchange chromatography (POROS™ 50 HS, ThermoFisher) and stored in 60 mM NaOAc buffer, 100 mM NaCl, pH 5.0. Sample purity and degree of heterogeneity were assessed by mass spectrometry and analytical size exclusion chromatography. All samples were confirmed to have a monomer content of ≥90% and contain <10% impurities prior to functional testing.

[0239] This resulted in the generation of bispecific and tetravalent binding molecules with different series 1 to series 4, hereafter also referred to as Doppelmab (DMab), as shown in Table 3 below (see also Vekataramani et al., Biochemical and Biophysical Research Communications, Volume 504, Issue 1, 26 September 2018, Pages 19-24).

[0240] [Table 3] TIFF0007633397000042.tif154169

[0241] In brief, the following design activities were carried out:

[0242] Series 1 In series 1, we generated binding molecules that had an antigen-binding fragment (Fab) as the TrkB-binding portion in the bispecific binding molecule. (2) and various VEGF binders, such as B20, G6 or Ranibizumab, were evaluated. Additionally, various VL-VH or VH-VL orientations were tested and multiple linkers, such as 20L3, 20L1, 15L1, 10L1, 6GS, were used in the design of the binding molecules.

[0243] Series 2 In series 2, we generated binding molecules with an antigen-binding fragment (Fab) as the VEGF-binding moiety in the bispecific binding molecule and again evaluated various VEGF binders. This time, we replaced the TrkB-binding moiety with an scFv (2) and various VL-VH or VH-VL orientations were tested.

[0244] Series 3 In series 3, binding molecules were generated based on the antigen-binding fragment (Fab) and ranibizumab as the VEGF-binding moiety in a bispecific binding molecule. Again, the TrkB-binding moiety was replaced by an scFv (2) and based on the C2 TrkB binder. Permutations in VL-VH or VH-VL orientations and various linkers were tested.

[0245] Series 4 In the final series 4, various mutations, such as VHE1Q, VHE6Q and / or VLD70G, ​​were introduced into the binding portion of Ranibizumab and the stabilizing effect of CC bridges was analyzed on the TrkB-binding scFv portion.

[0246] The results of these design activities and their impact on properties and bioactivity are now described in more detail in the Examples below.

[0247] Example 3 Comparison of human TrkB activation by C2, BDNF, and the four first series doppelmab TPP-11735, 736, 737, and 738 Figures 2A to 2B CHO cells stably expressing human TrkB were incubated with increasing concentrations of the natural TrkB ligands BDNF, C2 antibody, IgG1 isotype control or the four Doppelmab TPP-11735, TPP-11736, TPP-11737, TPP-11738. TrkB activation was assessed by measuring (A) TrkB phosphorylation at Y706 / 707 or (B) ERK1 / 2 phosphorylation downstream of TrkB at T202 / Y204 (ERK1) and T185 / Y187 (ERK2), respectively. The lowest compound concentration was solvent alone.

[0248] result TrkB activation by the DMab was virtually identical to the parent C2 molecule. The DMab also exhibited the same properties as the parent C2 molecule, acting as a partial TrkB agonist (pTrkB) with approximately 40-50% of the efficacy of BDNF in activating TrkB. However, the DMab incorporated the entire (Fab)2 portion of the parent C2 antibody without altering the sequence or orientation / layout, a very encouraging result, as we demonstrated that formatting into this specific format did not negatively affect the ability of the TrkB binder to activate TrkB and TrkB downstream signaling. As expected, the isotype control did not activate TrkB.

[0249] Example 4 Comparison of TrkB activation in cynomolgus monkeys, rabbits, rats and mice by C2, BDNF and the first series of doppelmab TPP-11735 / 736 Figures 3A to 3D CHO cells stably expressing (A) Cynomolgus TrkB, (B) Rabbit TrkB, (C) Rat TrkB or (D) Mouse TrkB were incubated with increasing concentrations of the natural TrkB ligands BDNF, C2 antibody or Doppelmab TPP-11735 or TPP-11736. TrkB activation was assessed by measuring TrkB phosphorylation at Y706 / 707. The lowest compound concentration was solvent alone. Data represent the mean + / - SEM.

[0250] result Again, TrkB activation by DMab was virtually identical to that of the parent C2 molecule, and activation was comparable to that of the human TrkB receptor. In this assay, DMab again displayed the same properties as the parent C2 molecule, acting as a partial TrkB agonist (pTrkB) with approximately 40-50% of the efficacy of BDNF in activating TrkB.

[0251] Example 5 Selectivity of TPP-11735 / 736 / 737 / 738-mediated TrkB activation Figures 4A to 4C CHO cells stably expressing (A) human TrkA, (B) human TrkB or (C) human TrkC were incubated with increasing concentrations of C2 antibody or the four Doppelmab TPP-11735, TPP-11736, TPP-11737, TPP-11738. Trk receptor activation was assessed by measuring receptor phosphorylation at Y706 / 707. Incubation with increasing concentrations of the natural ligands for TrkA (NGF), TrkB (BDNF) and TrkC (NT-3) were used as controls. The lowest compound concentration represents solvent alone. Data represent the mean + / - SEM.

[0252] result None of the Doppelmas tested activated TrkA or TrkC. All Doppelmas were highly specific / selective for TrkB.

[0253] Example 6 Comparison of C2-, TPP-11736- and / or BDNF-induced internalization of human TrkB receptors FIG. 5A to FIG. 5B (A) CHO cells stably expressing human TrkB were incubated with increasing concentrations of the natural TrkB ligand BDNF or 1 nM BDNF with increasing concentrations of C2 antibody or Doppelmab TPP-11736. (B) TrkB internalization was assessed by immunofluorescence staining of surface TrkB receptors without cell permeabilization, followed by confocal microscopy analysis. Dark and bright fields in the heatmap represent high and low percentages of cells above the fluorescence threshold, respectively.

[0254] result BDNF induced TrkB receptor internalization (lanes 1+2; heatmap fields darken from top to bottom).

[0255] Antibodies C2 and TPP-11736 reduced BDNF-induced internalization of TrkB receptors (lanes 3-5 and 6-8; note that the heatmap fields are darkened from bottom to top).

[0256] Example 7 Comparison of human TrkB activation by C2 and TPP-11736 (first series) in the presence or absence of human VEGF Figure 6 CHO cells stably expressing human TrkB were incubated with increasing concentrations of C2 tool antibody or Doppelmab TPP-11736 with or without preincubation with 200 ng / mL human VEGF-A (hVEGF). TrkB activation was assessed by measuring ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2), downstream of TrkB. Incubation with increasing concentrations of human VEGF-A without antibody served as control. The lowest compound concentration was solvent alone. Data represent the mean. Error bars are omitted for clarity.

[0257] result Incubation with human VEGF alone did not alter ERK1 / 2 phosphorylation.

[0258] Furthermore, preincubation with human VEGF did not substantially alter the potency or efficacy of ERK1 / 2 phosphorylation by C2 or TPP-11736.

[0259] Example 8 Neuroprotective efficacy of DMab 11736 and C2 in STZ-induced diabetic rats Figure 7 Neuroprotective function of TrkB activation in a rat model of diabetes-induced retinal neurodegeneration using IVT injection of an agonistic TrkB tool antibody (C2) and Doppelmab TPP-11736. Animals were treated with STZ to induce hyperglycemia. Retinal function was assessed by electroretinogram (ERG) before and after treatment. Diabetes induction resulted in a delayed implicit time within 3 weeks after STZ treatment. At this time, animals were intravitreally administered isotype control antibody (anti-TNP) or C2 (19 μg / 5 μL, respectively) or an equimolar amount of TPP-11736 (25 μg / 5 μL). After 2 weeks of treatment, ERG recordings were repeated and analyzed. Rod-driven B-wave implicit substitution latencies immediately before and 2 weeks after intravitreal application of the antibodies are shown in Figure 7. Mean + / - SEM; n.s. p>0.05, no significant difference; ** p < 0.01; *** p<0.001; one-way ANOVA with Tukey's multiple comparison test.

[0260] result Two weeks after dosing, anti-TNP antibody treatment did not shorten diabetes-induced rod-driven b-wave implicit time compared to the time point before anti-TNP treatment (15.4 ms at t = 2 weeks vs. 13.3 ms at t = 0; p > 0.05, not significant).

[0261] After 2 weeks of treatment, TPP-11736- and C2-treated animals showed a significant shortening of diabetes-induced rod-driven b-wave implicit time latency compared to the time point before antibody treatment (C2: 8.62 ms at t = 2 weeks vs. 14.4 ms at t = 0; *** p<0.01;TPP-11736 t=10.3ms at 2 weeks vs. 16.1ms at t=0; ** p<0.01).

[0262] Consistent with earlier measurements in in vitro assays of TrkB activation, C2 and TPP-11736 demonstrated similar neuroprotective effects in vivo.

[0263] Example 9 Comparison of human VEGF-A uptake by Eylea and the first series of four Doppelmab TPP-11735, 736, 737 and 738-Inhibition of VEGF-induced VEGFR2 phosphorylation FIG. 8A to FIG. 8B Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated binding molecules or Eylea.

[0264] VEGF-A uptake was assessed by measuring VEGF receptor 2 (VEGFR2) phosphorylation at Y1175. (A) Comparison of Doppelmab TPP-11735, -736, -737 and -738. (B) Comparison of TPP-11736 and TPP-11738 with Eylea. Unstimulated cells (basal) and 50ng / ml human VEGF without antibody treatment served as controls. Data represent mean + / - SEM.

[0265] result TPP-11737 vs. TPP-11738: Both of these molecules had the G6 anti-VEGF molecule as the scFv. 737 was linked in a VH-VL orientation, while 738 had a VL-VH orientation. It turned out that the VL-VH orientation worked much better, based on the G6 anti-VEGF antibody.

[0266] TPP-11735 vs. TPP-11736: Both of these molecules had the B20 anti-VEGF molecule as the scFv. There was little performance difference between the two orientations for the B20 VEGF antibody in this assay.

[0267] TPP-11736 vs. TPP-11738: These molecules had either the B20 (VL-VH) or G6 (VL-VH) anti-VEGF molecule as scFv. Both binding molecules showed similar potency (IC50 approx. 4 nM), but TPP-11736 was more effective.

[0268] Eylea (IC50=0.7 nM) is more potent than either TPP-11736 or TPP-11738 (approximately 4 nM).

[0269] Example 10 Comparison of capture of human VEGF-A by Eylea and the first series of four Doppelmab TPP-11735, 736, 737 and 738-Inhibition of VEGF-induced ERK1 / 2 phosphorylation FIG. 9A to FIG. 9B Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated antibodies or Eylea.

[0270] VEGF-A uptake was assessed by TrkB activation assessed by measuring ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2). (A) Comparison of Doppelmab TPP-11735, -736, -737 and -738. (B) Comparison of TPP-11736, -738 with Eylea. Unstimulated cells (basal) and 50ng / ml human VEGF without antibody treatment served as controls. Data represent mean + / - SEM.

[0271] result TPP-11737 vs. TPP-11738: Both of these molecules had the G6 anti-VEGF molecule as the scFv. 737 was linked in a VH-VL orientation, while 738 had a VL-VH orientation. It turned out that the VL-VH orientation worked much better, based on the G6 anti-VEGF antibody.

[0272] TPP-11735 vs. TPP-11736: Both of these molecules had the B20 anti-VEGF molecule as the scFv. There was little performance difference between the two orientations for the B20 VEGF antibody in this assay.

[0273] TPP-11736 vs. TPP-11738: These molecules had either the B20 (VL-VH) or G6 (VL-VH) anti-VEGF molecule as scFv. Both binding molecules showed similar potency (IC50 approx. 4 nM), but TPP-11736 was more effective.

[0274] Eylea (IC50=2 nM) is more potent than either TPP-11736 or TPP-11738 (approximately 10 nM).

[0275] Example 11 Comparison of human VEGF-A uptake by Eylea and the first series of four Doppelmab TPP-11736-Inhibition of VEGF-Induced p38 MAPK Phosphorylation Figure 10 Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50ng / mL human VEGF with or without preincubation with increasing concentrations of TPP-11736 or Eylea. VEGF-A uptake was assessed by measuring p38 MAPK phosphorylation at T180 / Y182. Unstimulated cells (basal) and 50ng / ml human VEGF without antibody treatment served as controls. Data represent the mean + / - SEM.

[0276] result Eylea (IC 50 = 0.74 nM) was approximately 10 times more potent than TPP-11736 (approximately 7.4 nM).

[0277] Example 12 Comparison of capture of human VEGF-A by the first series of four Doppelmab TPP-11735, 736, 737 and 738 and Eylea - inhibition of VEGF-induced HRMEC proliferation Figures 11A to 11E Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 10 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated binding molecules or Eylea. Molecule concentrations are shown in mol / L. VEGF-A uptake was assessed by automated image-based quantification of HRMEC cell number (IncuCyte). Images were recorded every 4 hours for a total of 96 hours. Relative cell numbers are shown. Cell number at t=0 was set to 1. Unstimulated cells served as control (basal). Data represent mean + / - SEM.

[0278] result TPP-11737 vs. TPP-11738: Both of these molecules had the G6 anti-VEGF molecule as scFv. TPP-11737 was linked in a VH-VL orientation, while TPP-11738 had a VL-VH orientation. Apparently, the VL-VH orientation worked much better for G6. TPP-11737 had virtually no effect.

[0279] TPP-11735 (VH-VL) vs. TPP-11736 (VL-VH): Both of these molecules have the B20 anti-VEGF molecule as the scFv. Again, the VL-VH orientation performed better.

[0280] 736 vs. 738: B20 (VL-VH) vs. G6 (VL-VH): TPP-11736 (B20) was clearly more effective in this assay.

[0281] Finally, Eylea was more effective than TPP-11736.

[0282] Series 1 - Summary of Findings for TPP-11735, TPP-11736, TPP-11737, and TPP-11738 TrkB activation TrkB activation by DMabs was virtually identical to the parent C2 molecule, which was unexpected because the entire (Fab)2 portion of the C2 TrkB binder was incorporated into the DMab without altering sequence or orientation / layout, and all Doppelmab (and C2) were only partial TrkB agonists.

[0283] VEGF capture Apparently, the VL-VH orientation worked better for G6 and B20 as VEGF binders, with B20 (VL-VH) being more effective than G6 (VL-VH), and both Doppelmab (B20 and G6) were inferior to Eylea in vitro.

[0284] Based on these initial findings, the inventors set out to generate and test various linkers of B20 in the VL-VH orientation to improve VEGF capture of Doppelmab (TPP-16061 to TPP-16064), as described in the Examples below.

[0285] Example 13 Comparison of human TrkB activation by C2, BDNF and the first series of four Doppelmab TPP-16061, 16062, 16063 and 16064; TrkB phosphorylation Figure 12

[0286] Human CHO cells stably expressing human TrkB were incubated with increasing concentrations of the natural TrkB ligands BDNF, C2 antibody, TPP-11736 (B20, scFv, 20L3, VL-VH) or four Doppelmab with different linkers: TPP-16061 (B20, scFv, 20L1, VL-VH), TPP-16062 (B20, scFv, 15L1, VL-VH), TPP-16063 (B20, scFv, 10L1, VL-VH), TPP-16064 (B20, scFv, 6GS, VL-VH). TrkB activation was assessed by measuring TrkB phosphorylation at Y706 / 707. The lowest compound concentration was solvent alone. Data represent the mean. Error bars have been omitted for clarity.

[0287] result TrkB activation by DMab TPP-16061-16064 was virtually identical to the parent C2 molecule and TPP-11736. TrkB activation seemed largely independent of differences in the linker between the Fc and scFv (anti-VEGF) parts of the protein. This was expected, since the linker was located away from the TrkB(Fab)2 fragment between the Fc and scFv (anti-VEGF) parts of the protein. Nevertheless, it was good confirmation that different linkers can be utilized without affecting the activity of the binding molecules. Finally, all presented binding molecules showed only partial TrkB agonist activity.

[0288] Example 14 Comparison of human TrkB activation by BDNF, TPP-11736, and two first series doppelmabs TPP-16061 and 16062; ERK1 / 2 phosphorylation Figure 13 CHO cells stably expressing human TrkB were incubated with increasing concentrations of the natural TrkB ligand BDNF, TPP-11736 (B20, scFv, 20L3, VL-VH) or two Doppelmab with different linkers TPP-16061 (B20, scFv, 20L1, VL-VH) and TPP-16062 (B20, scFv, 15L1, VL-VH). TrkB activation was assessed by measuring ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2). The lowest compound concentration was solvent alone. Data represent the mean + / - SEM.

[0289] result TrkB activation by DMabs TPP-16061 and 16062 was virtually identical to TPP-11736, and here TrkB activation appeared largely independent of linker differences, with the presented binding molecules exhibiting only partial TrkB agonist activity.

[0290] Example 15 Comparison of human VEGF-A uptake by the first series of four Doppelmab TPP-16061, 062, 063, and 064 with TPP-11736 - (A) Inhibition of VEGF-induced VEGFR2 phosphorylation, (B) ERK1 / 2 phosphorylation FIG. 14A to FIG. 14B Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated antibodies; TPP-11736 (B20, scFv, 20L3, VL-VH) or four Doppelmab antibodies with different linkers: TPP-16061 (B20, scFv, 20L1, VL-VH), TPP-16062 (B20, scFv, 15L1, VL-VH), TPP-16063 (B20, scFv, 10L1, VL-VH), TPP-16064 (B20, scFv, 6GS, VL-VH). VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 (VEGFR2) phosphorylation at Y1175 or (B) ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2), respectively. Unstimulated cells (basal) and 50 ng / ml human VEGF without antibody treatment served as controls. Data represent the mean + / - SEM.

[0291] result The differences in the linker did not substantially affect the potency or efficacy of VEGF capture, which was somewhat unexpected given that the linkers were located in close proximity to the VEGF binding site, but further confirmed that a variety of linkers can be utilized without affecting the activity of the binding molecules.

[0292] Example 16 Comparison of human VEGF-A capture by the first series of four Doppelmab TPP-16 061, 062, 063 and 064 - inhibition of VEGF-induced HRMEC proliferation Figure 15 Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 10 ng / mL human VEGF with or without preincubation with 0.5 nM Doppelmab or 1 nM Eylea. VEGF-A uptake was assessed by automated image-based quantification of HRMEC cell number (IncuCyte). Images were recorded every 4 hours for a total of 96 hours. Relative cell numbers are shown. Cell number at t=0 was set to 1. Unstimulated cells served as control (base). Data represent mean + / - SEM.

[0293] result Similar to the phosphorylation assay shown in Example 15, the differences in linkers did not perform better (or worse) than the parent molecule TPP-11736. Eylea was used at 1 nM because it is considered to be a monovalent molecule, whereas Doppelmab is bivalent. Under these conditions, Eylea performed better in inhibiting VEGF-induced proliferation of HRMEC.

[0294] Example 17 Comparison of human VEGF-A capture by Eylea, TPP-11736 and the first series of four Doppelmabs TPP-16061, 062, 063 and 064 - inhibition of VEGF-induced HRMEC sprouting Figure 16 Human retinal microvascular endothelial cell (HRMEC) spheroids were embedded in collagen matrix. Endothelial sprouting was induced by incubation with 50ng / mL human VEGF for 24 hours with or without 24 hours preincubation with 2.5nM indicated Doppelmab or 5nM Eylea. Endothelial sprouting was assessed by confocal microscopy and spheroid circumferences obtained from maximum projections of Z-stacks are shown. Unstimulated cells served as control (base). Data represent mean + / - SEM. ns p>0.05 not significant vs. 50ng / mL hVEGF+2.5nM TPP-11736.

[0295] result Similar to the phosphorylation assay shown in Example 15, the differences in linkers did not perform better (or worse) than the parent molecule TPP-11736. Eylea was used at 1 nM because it is considered to be a monovalent molecule, whereas Doppelmab is bivalent. Under these conditions, Eylea performed better in inhibiting VEGF-induced proliferation of HRMEC.

[0296] Series 1 - Summary of Findings for TPP-16061, TPP-16062, TPP-16063, TPP-16064 TrkB activation: No difference / improvement was observed compared to the parent molecule, TPP-11736.

[0297] VEGF Capture: No difference / improvement compared to parent molecule, TPP-11736.

[0298] Based on these findings, we set out to compare the VEGF capture of TPP-11736 (B20 as scFv) with TPP-13788 (B20 IgG) to test whether reformatting B20 as a Fab would improve VEGF capture.

[0299] Example 18 Comparison of human VEGF-A capture by TPP-11736 and TPP-13788 (B20 IgG) - inhibition of VEGF-induced phosphorylation of VEGFR2, ERK1 / 2, and p38 MAPK Figures 17A to 17C Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50ng / mL human VEGF with or without preincubation with increasing concentrations of Doppelmab TPP-11736 (B20 anti-VEGF as scFv) or TPP-13788 (B20 IgG). Capture of VEGF-A was assessed by measuring (A) VEGF receptor 2 (VEGFR2) phosphorylation at Y1175, (B) ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2) or (C) p38 MAPK phosphorylation at T180 / Y182. 50ng / ml human VEGF without molecular treatment served as control. Data represent mean + / - SEM. The corresponding best-fit ICs are shown in Table 4 below. 50 (Nonlinear regression; log(agonist) vs. response (three parameters)) is reported.

[0300] [Table 4]

[0301] result The B20 IgG tested, TPP-13788, exhibits approximately 3-fold improved VEGF-A capture capacity versus TPP-11736.

[0302] Example 19 Comparison of human VEGF-A capture by TPP-11736 and TPP-13788 (B20 IgG) - inhibition of VEGF-induced HRMEC sprouting Figure 18 Human retinal microvascular endothelial cell (HRMEC) spheroids were embedded in a collagen matrix. Endothelial sprouting was induced by incubation with 50ng / mL human VEGF for 24 hours with or without preincubation with 2.5nM Doppelmab TPP-11736 or 2.5nM TPP-13788 (B20 IgG). Endothelial sprouting was assessed by confocal microscopy and spheroid circumferences obtained from maximum projections of Z-stacks are shown. Unstimulated cells served as control (basal). Data represent mean + / - SEM. ns p>0.05 not significant vs. 50ng / mL hVEGF+2.5nM TPP-11736.

[0303] result Despite improved VEGF capture in VEGFR2, ERK1 / 2 and p38 MAPK phosphorylation assays, TPP-13788 (B20 IgG)-mediated inhibition of VEGF-induced sprouting was not better than that of TPP-11736.

[0304] Example 20 Comparison of human TrkB activation by BDNF, TPP-11736, and two first series doppelmabs TPP-14936 and TPP-14937; phosphorylation of TrkB and ERK1 / 2 FIG. 19A to FIG. 19C CHO cells stably expressing human TrkB were incubated with increasing concentrations of the natural TrkB ligands BDNF, TPP-11736 (B20, scFv, 20L3, VL-VH) or the two Doppelmab TPP-13936 (Ranibizumab, scFv, 20L3, VH-VL) or TPP-14937 (Ranibizumab, scFv, 20L3, VL-VH). TrkB activation was assessed by measuring (A) TrkB phosphorylation at Y706 / 707, (B and C) ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2), respectively. The lowest compound concentration was solvent alone. Data represent the mean + / - SEM.

[0305] result TrkB activation by DMabs TPP-14936 and TPP-14937 was virtually identical to that of TPP-11736, which was expected since the TrkB-binding component of the binding molecules was always C2 as a Fab fragment.

[0306] Example 21 Comparison of TrkB activation by BDNF, TPP-11736 and Doppelmab TPP-14936 in cynomolgus monkeys and rats; TrkB phosphorylation FIG. 20A to FIG. 20B CHO cells stably expressing (A) cynomolgus monkey TrkB or (B) rat TrkB were incubated with increasing concentrations of the natural TrkB ligands BDNF, TPP-11736 (B20, scFv, 20L3, VL-VH) or TPP-14936 (Ranibizumab, scFv, 20L3, VH-VL). TrkB activation was assessed by measuring TrkB phosphorylation at Y706 / 707. The lowest compound concentration was solvent alone. Data represent the mean + / - SEM.

[0307] result TrkB activation by DMab TPP-14936 was virtually identical to that of TPP-11736. Again, this was expected since the TrkB-binding component of the two molecules is C2 as a Fab fragment in both cases.

[0308] Example 22 Comparison of human VEGF-A capture by TPP-11736, TPP-14936 or TPP-14937 - Inhibition of VEGF-induced phosphorylation of VEGFR2, ERK1 / 2 and Src Figures 21A to 21C Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of Doppelmab TPP-11736 (B20, scFv, 20L3, VL-VH), TPP-14936 (Ranibizumab, scFv, 20L3, VH-VL) or TPP-14937 (Ranibizumab, scFv, 20L3, VL-VH). VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 (VEGFR2) phosphorylation at Y1175, (B) ERK1 / 2 phosphorylation at T202 / Y204 and T185 / Y187, or (C) Src phosphorylation at Y419. 50ng / ml human VEGF without antibody treatment served as a control. Data represent the mean + / - SEM. The corresponding best fit ICs are shown in Table 5 below. 50 Values ​​and absolute efficacy values ​​(bottom plateau) are reported respectively (nonlinear regression; log(agonist) vs. response (three parameters)).

[0309] [Table 5]

[0310] result Efficacy of VEGF capture (IC 50 ) is similar between TPP-11736 and TPP-14936 / TPP-14937.

[0311] Efficacy of VEGF capture by TPP-14936 (IC 50 ) is somewhat better than TPP-14937 → better VH-VL than VL-VH orientation of Ranibizumab scFv.

[0312] VEGF sequestration by TPP-14936 and TPP-14937 was more effective (with smaller trough values) than TPP-11736.

[0313] Example 23 Comparison of human VEGF-A capture by TPP-11736 and Doppelmab TPP-14936 and TPP-14937 - Inhibition of VEGF-induced HRMEC sprouting Figure 22 Human retinal microvascular endothelial cell (HRMEC) spheroids were embedded in a collagen matrix. Endothelial sprouting was induced by incubation with 50ng / mL human VEGF for 24 hours with or without preincubation with 2.5nM Doppelmab TPP-11736 (B20, scFv, 20L3, VL-VH), TPP-14936 (Ranibizumab, scFv, 20L3, VH-VL) or TPP-14937 (Ranibizumab, scFv, 20L3, VL-VH). Endothelial sprouting was assessed by confocal microscopy and spheroid circumferences obtained from maximum projections of Z-stacks are shown. Unstimulated cells served as control (basal). Data represent mean + / - SEM. ns p>0.05 no significant difference, **** p<0.0001.

[0314] result Although VEGF sequestering by TPP-14936 and TPP-14937 was similar in the VEGFR2 / ERK1_2 / Src phosphorylation assay, there was a dramatic difference in the inhibition of VEGF-induced sprouting. There was no significant difference between TPP-11736 and TPP-14937, but TPP-14936 was much more effective at inhibiting VEGF-induced HRMEC sprouting than either TPP-14937 or TPP-11736.

[0315] This unexpected finding indicates that the VH-VL orientation of the anti-VEGF scFv is important for performance in this assay. VH-VL is much better than VL-VH. Interestingly, the VL-VH orientation was better than VH-VL in the cases of TPP-11735 vs. 736 (B20) and also in the case of TPP-11737 vs. 738 (G6).

[0316] Example 24 Comparison of human VEGF-A uptake by Doppelmab TPP-11736, TPP-14936 and TPP-14937 and Eylea - Inhibition of VEGF-induced HRMEC proliferation Figures 23A to 23D Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 10 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated binding molecules or Eylea. Binding molecule / Eylea concentrations are shown in mol / L. VEGF-A capture was assessed by automated image-based quantification of HRMEC cell number (IncuCyte). Images were recorded every 4 hours for a total of 84 hours. Relative cell numbers are shown. Cell number at t=0 was set to 1. Unstimulated cells served as control (base). Data represent mean + / - SEM.

[0317] result TPP-14936 and TPP-14937 showed better efficacy compared to TPP-11736, especially in inhibiting VEGF-induced HRMEC proliferation. TPP-14936 was more potent than TPP-14937.

[0318] Very surprisingly, TPP-14936 and TPP-14937 were more effective than Eylea. Both Doppelmabs reduced proliferation even below baseline. In contrast to Eylea, they showed complete inhibition of VEGF-induced proliferation. However, Eylea was even more potent than TPP-14936 and 14937.

[0319] Series 1 - Summary of findings of TPP-14936 / 14937 TrkB activation: No difference / improvement was observed compared to TPP-11736. This type of thing was expected since both binding molecules had the C2 binder as a Fab fragment.

[0320] VEGF capture Phosphorylation assay: potency of VEGF capture (IC 50) were similar between TPP-11736 and TPP-14936 / TPP-14937. VEGF sequestration by TPP-14936 and TPP-14937 was more effective than TPP-11736 (bottom values).

[0321] Sprouting assay: There was no significant difference between TPP-11736 and TPP-14937. Inhibition of VEGF-induced HRMEC sprouting by TPP-14936 was much more effective than TPP-14937 or TPP-11736. Unexpectedly, the VH-VL orientation of the anti-VEGF Ranibizumab was much better than VL-VH.

[0322] Proliferation assay: TPP-14936 and TPP-14937 showed better efficacy compared to TPP-11736, especially in inhibiting VEGF-induced HRMEC proliferation. TPP-14936 was more potent than TPP-14937, and both TPP-14936 and TPP-14937 were more effective than Eylea.

[0323] Based on these findings, we set out to better understand the inverse layout of Doppelmab and Ranibizumab / B20 as Fabs to improve the VEGF capture and efficacy of C2 TrkB binder as scFv. From this, the following binding molecules were generated and tested in the examples below (series 2): TPP-14938 (B20 Fab), TPP-14939 (B20 Fab), TPP-14940 (Ranibizumab Fab), TPP-14941 (Ranibizumab Fab).

[0324] Example 25 Comparison of human TrkB activation by BDNF, TPP-11736 and a second series of four doppelmabs TPP-14938, TPP-14939, TPP-14940 and TPP-14941; phosphorylation of TrkB and ERK1 / 2 24A to 24B and FIG. CHO cells stably expressing human TrkB were incubated with increasing concentrations of the natural TrkB ligand BDNF, TPP-11736 (C2 as Fab) or a second series of four Doppelmab TPP-14938 (C2, 20L3, VH-VL as scFv), TPP-14939 (C2, 20L3, VL-VH as scFv), TPP-14940 (C2, 20L3, VH-VL as scFv) and TPP-14941 (C2, 20L3, VL-VH as scFv). TrkB activation was assessed by (A) measuring TrkB phosphorylation at Y706 / 707 or (B) measuring ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2), respectively. The lowest compound concentration was solvent alone. Data represent the mean + / - SEM.

[0325] result TPP-11736 contained the C2 CDRs as a Fab. The four Doppelmab antibodies TPP-14938, TPP-14939, TPP-14940 and TPP-14941 contained C2 as an scFv.

[0326] In stark contrast to TPP-11736 or the original C2 antibody, both of which are partial TrkB receptor agonists, all four novel Doppelmabs here showed full TrkB receptor agonist activity. Here, TrkB activation by these Doppelmabs was as effective as the natural ligand BDNF. This was completely unexpected, since the original C2 antibody alone showed a partial TrkB receptor agonist. Without wishing to be bound by theory, it is believed that VEGF-induced clustering and conformation of the binding molecule with a single binding molecule may be responsible for the observed improvement in the efficacy and potency of TrkB activation (see also FIG. 25).

[0327] The potency of TrkB activation was also somewhat better for Doppelmab TPP-14938 and TPP-14940 compared to TPP-14939 and TPP-14941. Thus, the "VH-VL" orientation of C2 scFv demonstrated better potency of TrkB activation.

[0328] Example 26 Comparison of Cynomolgus / Rat TrkB Activation by BDNF, TPP-11736, and a Second Series of Four Doppelmabs TPP-14938, TPP-14939, TPP-14940, and TPP-14941; TrkB and ERK1 / 2 Phosphorylation FIG. 26A to FIG. 26B CHO cells stably expressing (A) cynomolgus monkey TrkB or (B) rat TrkB were incubated with increasing concentrations of the natural TrkB ligand BDNF, TPP-11736 (C2 as Fab) or a second series of two Doppelmab TPP-14940 (C2 as scFv, 20L3, VH-VL) and TPP-14941 (C2 as scFv, 20L3, VL-VH). TrkB activation was assessed by measuring TrkB phosphorylation at Y706 / 707. The lowest compound concentration was solvent alone. Data represent the mean + / - SEM.

[0329] result The results of Example 25 for human TrkB could be reproduced in cynomolgus monkey and rat models, and here, in stark contrast to TPP-11736, the novel Doppelmab was a full TrkB receptor agonist. TrkB activation by these Doppelmab was as effective as the natural ligand BDNF.

[0330] Similarly, the potency of TrkB activation was somewhat better with Doppelmab TPP-14940 compared to TPP-14941. Thus, the VH-VL orientation of C2 scFv demonstrated better potency of TrkB activation.

[0331] Example 27 Comparison of human TrkB activation (TrkB phosphorylation) by C2, TPP-14940 and TPP-14941 (second series) in the presence or absence of human VEGF - Synergistic effect Figures 27A to 27C CHO cells stably expressing human TrkB were incubated with increasing concentrations of Doppelmab (A) TPP-14940, (B) TPP-14941, or (C) C2, with or without preincubation with 200 ng / mL human VEGF-A (hVEGF). TrkB activation was assessed by measuring TrkB phosphorylation at Y706 / 707. The lowest compound concentration was solvent alone. Data represent the mean + / - SEM.

[0332] result Surprisingly, preincubation with human VEGF dramatically improved the potency of TrkB phosphorylation (TrkB activation) by TPP-14941 (C2 as scFv, 20L3, VL-VH).

[0333] The effect of VEGF on the potency of TrkB phosphorylation (TrkB activation) was greater for TPP-14941 compared to TPP-14940. Thus, while both molecules showed synergistic effects on the potency of TrkB phosphorylation, the orientation / geometry of the TrkB activation portion of Doppelmab (scFv VL-VH vs. scFv VH-VL vs. Fab) appears to have an effect on synergy, with scFvs with a VL-VH orientation being preferred.

[0334] VEGF preincubation did not affect the efficacy of TrkB phosphorylation (TrkB activation) by C2 (control).

[0335] Example 28 Comparison of human TrkB activation (ERK1 / 2 phosphorylation) by C2, TPP-14940 and TPP-14941 (second series) in the presence or absence of human VEGF - Synergistic effect Figures 28A to 28C CHO cells stably expressing human TrkB were incubated with increasing concentrations of Doppelmab (A) TPP-14940, (B) TPP-14941, or (C) C2, with or without preincubation with 200 ng / mL human VEGF-A (hVEGF). TrkB activation was assessed by measuring ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2), downstream of TrkB. The lowest compound concentration represents vehicle alone. Data represent the mean + / - SEM.

[0336] result Again, surprisingly, pre-incubation with human VEGF dramatically improved the potency of ERK1 / 2 phosphorylation (TrkB activation) by TPP-14941 (C2 as scFv, 20L3, VL-VH).

[0337] Also here, the effect of VEGF on the potency of ERK1 / 2 phosphorylation (TrkB activation) was greater for TPP-14941 compared to TPP-14940. Again, both molecules showed synergistic effects on the potency of TrkB phosphorylation, but the orientation / geometry of the TrkB activation portion of Doppelmab (scFv VL-VH vs. scFv VH-VL vs. Fab) appears to have an effect on synergy, with scFv with a VL-VH orientation being preferred.

[0338] As shown previously, VEGF preincubation also did not affect the potency of ERK1 / 2 phosphorylation (TrkB activation) by TPP-11736 (series 1, C2 as Fab).Finally, VEGF preincubation did not affect the potency of ERK1 / 2 phosphorylation (TrkB activation) by C2 (control).

[0339] Example 29 Comparison of human TrkB activation (TrkB / ERK phosphorylation) by VEGF alone and in combination with BDNF - A control study to examine synergistic effects FIG. 29A to FIG. 29B CHO cells stably expressing human TrkB were incubated with increasing concentrations of human VEGF-A (hVEGF) alone or increasing concentrations of BDNF, with or without a fixed concentration of 200 ng / mL hVEGF. Activation of TrkB was assessed by measuring (A) TrkB phosphorylation at Y706 / 707 or (B) ERK1 / 2 phosphorylation downstream of TrkB at T202 / Y204 (ERK1) and T185 / Y187 (ERK2), respectively. The lowest compound concentration was vehicle alone. Data represent the mean + / - SEM.

[0340] result hVEGF alone did not induce phosphorylation of TrkB or ERK1 / 2.

[0341] The dose-response curve for BDNF is nearly independent of hVEGF.

[0342] Example 30 Comparison of Cynomolgus TrkB activation (TrkB phosphorylation) by C2, TPP-14940 and TPP-14941 (second series) in the presence or absence of human VEGF - Synergistic effects Figures 30A to 30C CHO cells stably expressing cynomolgus TrkB were incubated with increasing concentrations of Doppelmab (A) TPP-14940, (B) TPP-14941, or (C) C2, with or without preincubation with 200 ng / mL human VEGF-A (hVEGF). TrkB activation was assessed by measuring TrkB phosphorylation at Y706 / 707. The lowest compound concentration was vehicle alone. Data represent the mean + / - SEM.

[0343] result Also in this model, preincubation with human VEGF dramatically improved the potency of TrkB phosphorylation (TrkB activation) by TPP-14941 (C2, 20L3, VL-VH as scFv), whereas VEGF had less effect on the potency of TrkB phosphorylation (TrkB activation) by TPP-14940 (C2, 20L3, HL as scFv).

[0344] Example 31 Comparison of cynomolgus monkey TrkB activation (ERK1 / 2 phosphorylation) by C2, TPP-14940 and TPP-14941 (second series) in the presence or absence of human VEGF - Synergistic effect Figures 31A to 31C CHO cells stably expressing cynomolgus TrkB were incubated with increasing concentrations of Doppelmab (A) TPP-14940, (B) TPP-14941, or (C) C2, with or without preincubation with 200 ng / mL human VEGF-A (hVEGF). TrkB activation was assessed by measuring ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2), downstream of TrkB. The lowest compound concentration was vehicle alone. Data represent the mean + / - SEM.

[0345] result Also in this model, preincubation with human VEGF dramatically improved the potency of TPP-14941 (C2, 20L3, VL-VH as scFv)-induced ERK1 / 2 phosphorylation (TrkB activation), whereas VEGF had less effect on the potency of TPP-14940 (C2, 20L3, HL as scFv)-induced ERK1 / 2 phosphorylation (TrkB activation).

[0346] Example 32 Comparison of human VEGF-A capture by TPP-11736 (B20 as scFv), TPP-14938 (B20 as Fab) or TPP-14939 (B20 as Fab) - Inhibition of VEGF-induced phosphorylation of VEGFR2, ERK1 / 2 and Src Figures 32A to 32C Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50ng / mL human VEGF with or without preincubation with increasing concentrations of Doppelmab TPP-11736 (B20 as scFv), TPP-14938 (B20 as Fab) or TPP-14939 (B20 as Fab). VEGF-A capture was assessed by measuring (A) VEGF receptor 2 (VEGFR2) phosphorylation at Y1175, (B) ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2) or (C) Src phosphorylation at Y419. 50ng / ml human VEGF without antibody treatment served as control. Data represent mean + / - SEM. Table 6 below shows the corresponding best fit IC 50 Values ​​and absolute efficacy values ​​(bottom plateau) are reported respectively (nonlinear regression; log(agonist) vs. response (three parameters)).

[0347] [Table 6]

[0348] result Efficacy of VEGF capture by TPP-14938 / 39 (IC 50 ) was somewhat better than TPP-11736.

[0349] VEGF sequestration by TPP-14938 and TPP-14939 was more effective (with a smaller trough) than TPP-11736.

[0350] Overall, the differences between B20 scFv and B20 Fab are rather small and correspond well to previous results shown in Figures 17 / 18: TPP-11736 vs. TPP-13788 (B20 IgG). In conclusion, B20 apparently tolerates reformatting from Fab to scFv relatively well.

[0351] Example 33 Comparison of human VEGF-A capture by TPP-11736 (B20 scFv), TPP-14938 (B20 Fab) or TPP-14939 (B20 Fab) - inhibition of VEGF-induced proliferation of HRMEC Figures 33A to 33C Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 10 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated binding molecules. Molecule concentrations are shown in mol / L. VEGF-A capture was assessed by automated image-based quantification of HRMEC cell number (IncuCyte). Images were recorded every 4 hours for a total of 84 hours. Relative cell numbers are shown. Cell number at t=0 was set to 1. Unstimulated cells served as control (basal). Data represent mean + / - SEM.

[0352] result Reformatting the B20 scFv (TPP-11736) into a B20 Fab (TPP-14938 and TPP-14939) did not improve potency or efficacy of VEGF-A capture in proliferation assays, and indeed the B20 Fab-based antibodies appeared to be somewhat less potent / effective than TPP-11736.

[0353] Example 34 Comparison of human VEGF-A capture by TPP-14936 (Ranibizumab, scFv, HL) or TPP-14937 (Ranibizumab, scFv, LH) with TPP-14940 and TPP-14941 (Ranibizumab Fab) - Inhibition of VEGF-induced phosphorylation of VEGFR2, ERK1 / 2 and Src Figures 34A to 34C Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of Doppelmab TPP-11736 (B20 as scFv), TPP-14936 (Ranibizumab VH-VL as scFv), TPP-14937 (Ranibizumab as scFv, VL-VH), TPP-14940 (Ranibizumab as Fab) or TPP-14941 (Ranibizumab as Fab). VEGF-A capture was assessed by measuring (A) VEGF receptor 2 (VEGFR2) phosphorylation at Y1175, (B) ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2) or (C) Src phosphorylation at Y419. 50 ng / ml human VEGF without antibody treatment served as a control. Data represent the mean + / - SEM. The corresponding best fit ICs are shown in Table 7 below. 50 Values ​​and absolute efficacy values ​​(bottom plateau) are reported respectively (nonlinear regression; log(agonist) vs. response (three parameters)).

[0354] [Table 7]

[0355] result Reformatting Ranibizumab from scFv to Fab dramatically improved its potency for VEGF capture in all three assays (IC10-fold better than Ranibizumab scFv or the reference molecule TPP-11736). 50 ).

[0356] The efficacy of VEGF capture by Ranibizumab-based molecules was much better than that of TPP-11736 (bottom value).

[0357] However, Ranibizumab did not tolerate reformatting from Fab to scFv very well (in contrast to B20, see previous example).

[0358] Example 35 Comparison of human VEGF-A capture by TPP-14936 (Ranibizumab as scFv, VH-VL) or TPP-14937 (Ranibizumab as scFv, VL-VH) with TPP-14940 and TPP-14941 (Ranibizumab Fab) - Inhibition of VEGF-induced proliferation of HRMEC Figures 35A to 35D Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 10 ng / mL human VEGF with or without preincubation of increasing concentrations of the indicated binding molecules. Molecule concentrations are shown in mol / L. VEGF-A capture was assessed by automated image-based quantification of HRMEC cell number (IncuCyte). Images were recorded every 4 hours for a total of 84 hours. Relative cell numbers are shown. Cell number at t=0 was set to 1. Unstimulated cells served as control (basal). Data represent mean + / - SEM.

[0359] result Reformatting Ranibizumab (TPP-11736 / TPP-14937) as an scFv into Ranibizumab Fab (TPP-14940 and TPP-14941) significantly improved the potency of VEGF-A capture in proliferation assays.

[0360] For TPP-14940 and TPP-14941, complete VEGF inhibition was seen already at a concentration of 0.25 nM, whereas complete inhibition was required for 1 nM TPP-14936 and 4 nM TPP-14937, respectively.

[0361] The efficacy of VEGF-A inhibition was similar among all four molecules (all full inhibitors).

[0362] Example 36 Comparison of human VEGF-A capture by TPP-14938 or TPP-14939 (B20 Fab) with TPP-14940 and TPP-14941 (Ranibizumab Fab) - Inhibition of VEGF-induced phosphorylation of VEGFR2, ERK1 / 2 and Src Figures 36A to 36C Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of Doppelmab TPP-11736 (B20 as scFv), TPP-13938 (B20 as Fab) or TPP-13939 (B20 as Fab), TPP-14940 (Ranibizumab as Fab) or TPP-14941 (Ranibizumab as Fab). VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 (VEGFR2) phosphorylation at Y1175, (B) ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2) or (C) Src phosphorylation at Y419. 50ng / ml human VEGF without any molecule treatment served as a control. Data represent the mean + / - SEM. The corresponding best fit ICs are shown in Table 8 below. 50 Values ​​and absolute efficacy values ​​(bottom plateau) are reported respectively (nonlinear regression; log(agonist) vs. response (three parameters)).

[0363] [Table 8]

[0364] result Doppelmab with Ranibizumab Fab demonstrated clearly better efficacy in capturing VEGF than B20 Fab-based Doppelmab in all three assays.

[0365] Doppelmab with Ranobizumab Fab demonstrated clearly better efficacy in VEGF capture than B20 Fab-based Doppelmab in all three assays (nadir values).

[0366] Example 37 Comparison of human VEGF-A capture by TPP-14938 or TPP-14939 (B20 as Fab) with TPP-14940 and TPP-14941 (Ranibizumab as Fab) - Inhibition of VEGF-induced proliferation of HRMEC Figures 37A to 37D Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 10 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated binding molecules. Molecule concentrations are shown in mol / L. VEGF-A capture was assessed by automated image-based quantification of HRMEC cell number (IncuCyte). Images were recorded every 4 hours for a total of 84 hours. Relative cell numbers are shown. Cell number at t=0 was set to 1. Unstimulated cells served as control (basal). Data represent mean + / - SEM.

[0367] result The potency and efficacy of the Ranibizumab Fab-based molecules TPP-14940 and TPP-14941 in inhibiting VEGF-induced proliferation was clearly superior to that of the B20 Fab-based molecules TPP-14938 and TPP-14939.

[0368] Even at the highest concentration (16 nM), TPP-14938 and TPP-14939 were unable to completely inhibit VEGF-induced sprouting.

[0369] In contrast, TPP-14940 and TPP-14941 completely inhibited VEGF-induced proliferation already at a 64-fold lower concentration of 0.25 nM.

[0370] Example 38 Comparison of human VEGF-A capture by Doppelmab TPP-11736 (B20 as scFv), TPP-14936 (Ranibizumab as scFv, VH-VL), TPP-14937 (Ranibizumab as scFv, VL-VH), TPP-14938 (B20 as Fab) or TPP-14939 (B20 Fab), TPP-14940 (Ranibizumab as Fab), TPP-14941 (Ranibizumab as Fab) or Eylea - Inhibition of VEGF-induced HRMEC sprouting Fig. 38 Human retinal microvascular endothelial cell (HRMEC) spheroids were embedded in a collagen matrix. Endothelial sprouting was induced for 24 h by incubation with 50 ng / mL human VEGF with or without preincubation with 2.5 nM Doppelmab TPP-11736 (B20 scFv), TPP-14936 (Ranibizumab scFv, VH-VL), TPP-14937 (Ranibizumab scFv, VL-VH), TPP-14938 (B20 Fab), TPP-14939 (B20 Fab), TPP-14940 (Ranibizumab Fab), TPP-14941 (Ranibizumab Fab) or 5 nM Eylea. Endothelial sprouting was assessed by confocal microscopy and spheroid circumferences obtained from maximum projections of Z-stacks are shown. Unstimulated cells served as control (baseline). Data represent mean + / - SEM. ns p>0.05 not significant; compared to 50 ng / mL hVEGF § p<0.0001; compared with 50ng / mL hVEGF, #p>0.05 not significant. *** p<0.001; one-way ANOVA with Tukey's multiple comparison test.

[0371] result The Ranibizumab Fab-based Doppelmab TPP-14940 and TPP-14941 were clearly the best VEGF capture agents in this assay. In contrast to Eylea, TPP-14940 and TPP-14941 completely inhibited VEGF-induced sprouting. No statistically significant difference was observed between endothelial sprouting in the absence of VEGF (basal) and 50 ng / mL hVEGF preincubated with 2.5 nM TPP-14940 or TPP-14941.

[0372] Inhibition of VEGF-induced EC sprouting by TPP-14940 and TPP-14941 was significantly better than that by any of the other Doppelmab inhibitors.

[0373] Inhibition of VEGF-induced EC sprouting by TPP-14940 and TPP-14941 was significantly better than that by Eylea.

[0374] Furthermore, there was no statistically significant difference between endothelial sprouting stimulated with 50 ng / mL hVEGF and 50 ng / mL hVEGF preincubated with Eylea.

[0375] Example 39 Comparison of human VEGF-A uptake by Doppelmab TPP-11940 and Eylea - Inhibition of human VEGF-A-induced retinal hyperpermeability in Brown Norway rats Figures 39A to 39B TPP-14940 (but not Eylea) prevented human VEGF-A-induced hyperpermeability in rat retina. (A) Time protocol showing experimental procedure. 15 min after intravitreal (ivt) administration of anti-VEGF compounds (13 or 26 pmol / eye Eylea or TPP-14940) or control (26 pmol TPP-11737), 13 pmol human VEGF-A / eye was administered by ivt injection. PBS injection served as control. 24 h later, 1 mL / kg Evans Blue (EB) solution (45 mg / mL in 0.9% saline) was administered by intravenous (iv) injection for 30 min, after which the eyes were isolated and fixed. At the same time points, plasma samples were collected to ensure equal systemic EB exposure. (B) Quantification of VEGF-A-induced hyperpermeability in Brown Norway rat retinas was performed by measuring EB extravasation in retinal flat mounts by confocal microscopy. The eyes were cut along the ora serrata, the lens and vitreous were removed, and the eyecups were fixed in paraformaldehyde (4%) for 1 h at 4°C, then transferred to PBS overnight at 4°C. The retina was separated from the outer segment (sclera and choroid), transferred to a glass slide, and cut four times to achieve a flat cloverleaf-like structure. The tissue was covered with mounting medium (Vectashield H-1200 containing the DNA stain DAPI) and a coverslip was placed on top to obtain a retinal flat mount. The samples were excited at a wavelength of 639 nm, and the emission of Evans blue at 669 nm was recorded with a LSM 700 confocal laser scanning microscope (Carl Zeiss, Jena; gain 800, laser intensity 2%, 5 stacks 60 μm) to obtain images of the retinal flat mount with maximum intensity projection. Analysis of the total fluorescence intensity was performed after opening the images in the program ImageJ with a threshold of 30. *** p<0.001; * p<0.05;ns p>0.05;#p>0.05 Not significantly different from TPP-11737+PBS. One-way ANOVA with Tukey's multiple comparison test. n=9-17.

[0376] result Under control conditions, intravitreal VEGF-A injection increased vascular permeability by approximately 60%.TPP-11737 was used as a control antibody because it has the same molecular format (Doppelmab) as TPP-14940, but in initial in vitro assays, this compound did not capture VEGF.

[0377] At 1:1 and 2:1 molecular ratios of binding molecule to VEGF, TPP-14940 completely blocked vascular hyperpermeability, whereas, in sharp contrast, Eylea failed to significantly reduce vascular leakage under the same conditions and even at twice the concentration.

[0378] Series 2 - Summary of Findings for TPP-14938, TPP-14939, TPP-14940, and TPP-14941 TrkB activation In sharp contrast to TPP-11736, all four novel doppelmas were full TrkB receptor agonists. TrkB activation by these doppelmas was as effective as that by the natural ligand, BDNF.

[0379] The efficacy of TrkB activation was somewhat better in Doppelmab TPP-14938 / 40 versus TPP-14939 and TPP-14941 with the "VH-VL" orientation.

[0380] Surprisingly, preincubation with human VEGF dramatically improved the efficacy of TrkB phosphorylation (TrkB activation) by TPP-14941 [C2, scFv, 20L3, VL-VH], exhibiting a synergistic effect.

[0381] In contrast to TPP-14941, the effect of VEGF on the potency of TrkB phosphorylation (TrkB activation) by TPP-14940 [C2, scFv, 20L3, VH-VL] was relatively small.

[0382] VEGF capture Reformatting the B20 scFv (TPP-11736) into a B20 Fab (TPP-14938 and TPP-14939) did not improve potency or efficacy of VEGF-A capture in proliferation assays.

[0383] Reformatting Ranibizumab scFv (TPP-11736 / TPP-14937) into Ranibizumab Fab (TPP-14940 and TPP-14941) significantly improved the efficacy of VEGF-A capture in proliferation assays.

[0384] Complete VEGF inhibition was seen already at concentrations as low as 0.25 nM using TPP-14940 and TPP-14941.

[0385] The potency and efficacy of the ranibizumab Fab-based molecules TPP-14940 and TPP-14941 in inhibiting VEGF-induced proliferation was superior to that of the B20 Fab-based molecules TPP-14938 and TPP-14939.

[0386] Since the inventors observed linker clipping in the binding molecules in the examples below (data not shown), they derived further molecules based on TPP-14940: TPP-19986 (VH-VL, 10L1), TPP-19987 (VH-VL, 20L1) and based on TPP-14941: TPP-19988 (VL-VH, 10L1), TPP-19989 (VL-VH, 20L1), all based on Ranibizumab as Fab and C2 as scFv.

[0387] Example 40 Comparison of human TrkB activation by BDNF, TPP-14940 [C2, scFv, 20L3, VH-VL] or its linker variants TPP-19986 [C2, scFv, 10L1, VH-VL] or TPP-19987 [C2, scFv, 20L1, VH-VL] or TPP-14941 [C2, scFv, 20L3, VL-VH] and its linker variants TPP-19988 [C2, scFv, 10L1, VL-VH] or TPP-19989 [C2, scFv, 20L1, VL-VH]; phosphorylation of TrkB and ERK1 / 2 - series 3 Figures 40A to 40D Human CHO cells stably expressing human TrkB were incubated with increasing concentrations of the natural TrkB ligand BDNF, TPP-14940 [C2, scFv, 20L3, VH-VL] or its linker variants TPP-19986 [C2, scFv, 10L1, VH-VL] or TPP-19987 [C2, scFv, 20L1, VH-VL] or TPP-14941 [C2, scFv, 20L3, V-VH] and its linker variants TPP-19988 [C2, scFv, 10L1, VL-VH] or TPP-19989 [C2, scFv, 20L1, VL-VH]. TrkB activation was assessed by measuring (A and B) TrkB phosphorylation at Y706 / 707 or (C and D) ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2), respectively. The lowest compound concentration was solvent alone. Data represent the mean + / - SEM. The corresponding best fit ICs for TrkB and ERK1 / 2 phosphorylation are shown below in Tables 9 and 10. 50 Values ​​and absolute efficacy values ​​(bottom plateau) are reported respectively (nonlinear regression; log(agonist) vs. response (three parameters)).

[0388] [Table 9]

[0389] [Table 10]

[0390] result There were no significant differences between the parent molecule and each linker variant. The potency and efficacy of phosphorylation of TrkB and ERK1 / 2 were virtually identical. This is noteworthy because the linker was located close to the TrkB binding site but did not affect the biological activity of the binding molecule.

[0391] All tested molecules were full TrkB agonists.

[0392] Again, as shown above, the efficacy of TrkB / ERK phosphorylation was better for the TPP-14940 molecule and its linker variants (VH-VL orientation) compared to the TPP-14941 molecule and its linker variants (VL-VH orientation).

[0393] Example 41 Comparison of human TrkB internalization by BDNF, TPP-14940 [C2, scFv, 20L3, VH-VL] or TPP-14941 [C2, scFv, 20L3, VL-VH] and linker variants TPP-19988 [C2, scFv, 10L1, VL-VH] or TPP-19989 [C2, scFv, 20L1, VL-VH] - Series 3 Figures 41A to 41B CHO cells stably expressing cynomolgus monkey TrkB were incubated with increasing concentrations of the natural TrkB ligand BDNF or 1 nM BDNF with increasing concentrations of the first series of Doppelmab TPP-14940 [C2, scFv, 20L3, VH-VL] or TPP-14941 [C2, scFv, 20L3, VL-VH] and linker variants TPP-19988 [C2, scFv, 10L1, VL-VH] or TPP-19989 [C2, scFv, 20L1, VL-VH]. TrkB internalization was assessed by immunofluorescence staining of surface TrkB receptors without cell permeabilization, followed by confocal microscopy analysis. Dark and bright fields in the heatmap represent high and low percentages of cells above the fluorescence threshold, respectively.

[0394] result BDNF induced TrkB receptor internalization. TPP-14940, TPP-14941, TPP-19988, and TPP-19989 were full TrkB receptor agonists, but none of them, Doppelmab, increased BDNF-induced receptor internalization, but rather reduced BDNF-induced internalization of TrkB receptors (Note: heatmap fields are darkened from bottom to top).

[0395] Example 42 Comparison of human VEGF-A capture by TPP-14940 and its linker variants TPP-19986 and TPP-14941 and its linker variants TPP-19988 - Inhibition of VEGF-induced phosphorylation of VEGFR2 - Series 3 Fig. 42 Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50ng / mL human VEGF with or without preincubation with increasing concentrations of Doppelmab TPP-14940 and its linker variants TPP-19986 and TPP-14941 and its linker variants TPP-19988. VEGF-A capture was assessed by measuring VEGF receptor 2 (VEGFR2) phosphorylation at Y1175. 50ng / ml human VEGF without any molecular treatment served as a control. Data represent the mean + / - SEM. Table 11 below shows the corresponding best fit IC for inhibition of VEGFR2 phosphorylation. 50 Values ​​(nonlinear regression; log(agonist) vs. response (three parameters)) are reported.

[0396] [Table 11]

[0397] result There were no significant differences between the parent molecules TPP-14940 / TPP-14941 and each linker variant. The potency and efficacy of inhibition of VEGF-induced VEGFR2 phosphorylation were virtually identical.

[0398] Example 43 Comparison of Human VEGF-A Capture by Doppelmab TPP-14940 [C2, scFv, 20L3, VH-VL] or its linker variants TPP-19986 [C2, scFv, 10L1, VH-VL] or TPP-19987 [C2, scFv, 20L1, VH-VL] or TPP-14941 [C2, scFv, 20L3, VL-VH] and its linker variants TPP-19988 [C2, scFv, 10L1, VL-VH] or TPP-19989 [C2, scFv, 20L1, VL-VH] - Inhibition of VEGF-induced HRMEC Sprouting - Series 3 Fig. 43 Human retinal microvascular endothelial cell (HRMEC) spheroids were embedded in a collagen matrix. Endothelial sprouting was induced by incubation with 50 ng / mL human VEGF for 24 h with or without preincubation with 2.5 nM Doppelmab TPP-11736 (B20 scFv), TPP-14940 [C2, scFv, 20L3, VH-VL] or its linker variants TPP-19986 [C2, scFv, 10L1, VH-VL] or TPP-19987 [C2, scFv, 20L1, VH-VL] or TPP-14941 [C2, scFv, 20L3, VL-VH] and its linker variants TPP-19988 [C2, scFv, 20L1, VL-VH] or TPP-19989 [C2, scFv, 20L1, VL-VH]. Endothelial sprouting was assessed by confocal microscopy and spheroid periphery was shown from maximum projection of Z-stacks. Unstimulated cells served as control (base). Data represent mean + / - SEM. ns p>0.05 not significant; compared to 50ng / mL hVEGF § p<0.0001; compared with 50ng / mL hVEGF, #p>0.05 not significant. *** p<0.001; one-way ANOVA with Tukey's multiple comparison test.

[0399] result No significant differences were observed between the parent molecules TPP-14940 / TPP-14941 and each linker variant. All molecules completely inhibited VEGF-induced sprouting of HRMEC.

[0400] Series 3 - Summary of findings of TPP 19986, 19987, 19988, 19989 TrkB activation No significant differences were observed between the parent molecule (TPP-14940 / 14941) and each linker variant.

[0401] All tested binding molecules were full TrkB receptor agonists. The VH-VL orientation and its linker variants were even more potent.

[0402] No changes in receptor internalization were observed, which was quite surprising given that the novel Doppelmab was similar to BDNF, a full TrkB receptor agonist.

[0403] VEGF capture No significant differences were observed between the parent molecule (TPP-14940 / 14941) and each linker variant.

[0404] CMC characteristics With the use of the 10L1 and 20L1 linkers, linker clipping was no longer observed (data not shown).

[0405] As a next step, a final series of binding molecules was designed (series 4) in order to improve the CMC and biophysical properties of the binding molecules, while always trying to preserve the excellent biological function of the binding molecules of the invention.

[0406] Example 44 Comparison of human TrkB activation by Doppelmab with and without disulfide bridge (CC) in the TrkB scFv portion; TrkB phosphorylation-series 4 Figures 44A to 44D CHO cells stably expressing human TrkB were incubated with increasing concentrations of the indicated Doppelmab with or without CC cross-linking in the anti-TrkB scFv portion. (A) TPP-22180 vs. TPP-22204; (B) TPP-22192 vs. TPP-22216; (C) TPP-22190 vs. TPP-22214; (D) TPP-22191 vs. TPP-22215. TrkB activation was assessed by measuring TrkB phosphorylation at Y706 / 707. The lowest compound concentration was solvent alone. Data represent the mean + / - SEM. The corresponding best-fit IC for TrkB activation is shown below in Table 12. 50 Values ​​and absolute or relative efficacy values ​​(upper plateau), respectively, are reported (nonlinear regression; log(agonist) vs. response (three parameters)).

[0407] [Table 12]

[0408] result Disulfide bridges were introduced primarily to improve the CMC properties of the binding molecules. Surprisingly, the inclusion of disulfide bridges in the scFv anti-TrkB moiety further improved both the potency and efficacy of TrkB activation. A greater effect was observed in binding molecules with a VL-VH orientation.

[0409] Example 45 Comparison of rat TrkB activation by Doppelmab with and without disulfide bridge (CC) in the TrkB scFv portion; TrkB phosphorylation-series 4 Figures 45A to 45B CHO cells stably expressing rat TrkB were incubated with increasing concentrations of the indicated Doppelmab with or without CC cross-linking in the anti-TrkB scFv moiety. (A) TPP-22180 vs. TPP-22204; (B) TPP-22192 vs. TPP-22216. TrkB activation was assessed by measuring TrkB phosphorylation at Y706 / 707. The lowest compound concentration was solvent alone. Data represent the mean + / - SEM. The corresponding best-fit IC for TrkB activation is shown below in Table 13. 50 Values ​​and absolute or relative efficacy values ​​(upper plateau), respectively, are reported (nonlinear regression; log(agonist) vs. response (three parameters)).

[0410] [Table 13]

[0411] result Inclusion of disulfide bridges in the scFv anti-TrkB component improved both the potency and efficacy of TrkB activation. Again, the effect of the VL-VH orientation on the binding molecules appeared to be greater. Overall, the data were in good agreement with those obtained for human TrkB.

[0412] Example 46 Selectivity of TPP-22204 / 22214-mediated TrkB activation Fig. 46 CHO cells stably expressing (A) human TrkA, (B) human TrkB or (C) human TrkC were incubated with increasing concentrations of C2 antibody or Doppelmab TPP-22204 or TPP-22214. Trk receptor activation was assessed by measuring receptor phosphorylation at Y706 / 707. Incubation with increasing concentrations of the natural ligands for TrkA (NGF), TrkB (BDNF) and TrkC (NT-3) were used as controls. The lowest compound concentration was solvent alone. Data represent the mean + / - SEM.

[0413] result Both Doppelmabs were highly specific / selective for TrkB - they did not activate either TrkA or TrkC.

[0414] Example 47 Effect of TPP-22214 on BDNF-induced TrkB activation Figures 47A to 47B CHO cells stably expressing the human TrkB receptor were incubated with increasing concentrations of (A) C2 antibody or (B) Doppelmab TPP-22214 in the presence or absence of constant concentrations of 0.3 nM, 1 nM or 3 nM BDNF. Activation of TrkB was assessed by measuring receptor phosphorylation at Y706 / 707. Data represent the mean + / - SEM.

[0415] result In sharp contrast to C2, TPP-22214 did not limit BDNF-induced TrkB activation.

[0416] Example 48 Comparison of human TrkB activation (TrkB phosphorylation) by TPP-22214 in the presence or absence of human VEGF Figures 48A to 48D CHO cells stably expressing human TrkB were incubated with increasing concentrations of Doppelmab TPP-22214 with or without preincubation with (A) 200 ng / mL human VEGF-A (hVEGF), (B) 50 ng / mL hVEGF, (C) 10 ng / mL hVEGF, or (D) 2 ng / mL hVEGF. TrkB activation was assessed by measuring TrkB phosphorylation at Y706 / 707. The lowest compound concentration was vehicle alone. Data represent the mean + / - SEM.

[0417] result As shown previously, preincubation with 200ng / mL human VEGF improved the potency of TrkB phosphorylation (TrkB activation) by TPP-22214. More importantly, this synergistic effect was nearly preserved with 50, 10, and even 2ng / mL hVEGF. Similar effects were also obtained in other species, e.g., cynomolgus monkeys (data not shown).

[0418] Example 49 Comparison of human TrkB activation (ERK phosphorylation) by TPP-22214 in the presence or absence of human VEGF Figures 49A to 49D CHO cells stably expressing human TrkB were incubated with increasing concentrations of Doppelmab TPP-22214 with or without preincubation with (A) 200 ng / mL human VEGF-A (hVEGF), (B) 50 ng / mL hVEGF, (C) 10 ng / mL hVEGF, or (D) 2 ng / mL hVEGF. Activation of TrkB was assessed by measuring ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2), downstream of TrkB. The lowest compound concentration was vehicle alone. Data represent the mean + / - SEM.

[0419] result Similarly, preincubation with 200 ng / mL human VEGF improved the potency of ERK phosphorylation downstream of TrkB activation by TPP-22214. Also, the synergistic effect was largely preserved at concentrations as low as 2 ng / mL hVEGF in this assay. Similar effects were obtained in other species, e.g., cynomolgus monkeys (data not shown).

[0420] Example 50 VEGF binding enhances TrkB activation-a study of the proposed mechanism Fig. 50 Complexes of TPP-22214 and hVEGF were prepared in 1× Dulbecco's PBS by mixing samples together at the indicated molar ratios (1:1, 4:1 or 20:1). Samples were incubated at room temperature for 1 hour before analysis. An Agilent 1200 was used in conjunction with a mini DAWN TREOS and an Optilab REX. Samples were run at a flow rate of 0.6 ml / min in 1× Dulbecco's PBS as the mobile phase. 100 μl of the complex or TPP-22214 alone were injected onto a Superose 6 Increase column (30 cm×10 mm) where the molecules were subjected to hydrodynamic volume separation. Data were then analyzed using Astra 6.1.1.17.

[0421] result TPP-22214 forms complexes in the presence of VEGF-A. The largest complexes are formed at a 1:1 molar ratio. These complexes suggest complex formation between Doppelmab and VEGF in greater than 1:1 ratios. These larger complexes may result in clustering of TrkB receptors on the cell surface. Experimental data support the proposed mechanism.

[0422] Example 51 Comparison of human TrkB internalization by increasing concentrations of BDNF or TPP-22214 and by increasing concentrations of TPP-22214 in the presence of 1 nM BDNF Figures 51A to 51C CHO cells stably expressing human TrkB were incubated with (A) increasing concentrations of the natural TrkB ligand BDNF, (B) increasing concentrations of TPP-22214, or (C) 1 nM BDNF with increasing concentrations of TPP-22214. TrkB internalization was assessed by immunofluorescence staining of surface TrkB receptors without cell permeabilization followed by confocal microscopy analysis. Data represent the percentage of cells with surface TrkB staining intensity above threshold; mean + / - SEM.

[0423] result (A) Incubation with BDNF induced TrkB receptor internalization.

[0424] (B) TPP-22214 did not induce TrkB receptor internalization.

[0425] (C) Increasing concentrations of TPP-22214 reduced / abolished BDNF-induced receptor internalization.

[0426] Example 52 Neuroprotective effects of TPP-22214 and C2 in STZ-induced diabetic rats Fig. 52 Neuroprotective function of TrkB activation in a rat model of diabetes-induced retinal neurodegeneration using IVT injection of an agonistic TrkB antibody (C2) and Doppelmab TPP-22214. Animals were treated with STZ to induce hyperglycemia. Retinal function was assessed by electroretinogram (ERG) before and after treatment. Diabetes induction resulted in a delay in implicit time within 3 weeks after STZ treatment. At this time, animals were intravitreally administered isotype control antibody (anti-TNP) or C2 (19 μg / 5 μL, respectively) or an equimolar amount of TPP-22214 (25 μg / 5 μL). Rod-driven B-wave implicit time was shown to be delayed immediately before and 2 weeks after intravitreal application of the molecules; mean + / - SEM; n.s. p>0.05; no significant difference (ns), **** p<0.0001; one-way ANOVA with Tukey's multiple comparison test.

[0427] result Two weeks after administration, anti-TNP antibody treatment did not shorten the diabetes-induced rod-driven b-wave implicit time compared to the time point before anti-TNP treatment. In fact, the implicit latency was even significantly prolonged compared to the time point before anti-TNP antibody application (17.1 ms at t = 2 weeks vs. 11.2 ms at t = 0 weeks; **** p<0.0001), indicating that STZ-induced retinal damage was not fully established at t=0 weeks and that anti-TNP treatment was unable to halt the further increase in implicit time delay.

[0428] Two weeks after administration, C2 antibody treatment prevented further increases in the implicit time delay. However, C2 antibody treatment did not significantly shorten diabetes-induced rod-driven b-wave implicit time compared to the time point before C2 treatment (10.3 ms at t = 2 weeks vs. 12.8 ms at t = 0 weeks; n.s. p>0.05; not significant).

[0429] After 2 weeks of dosing, TPP-22214 antibody treatment prevented further increases in implicit time delays compared to pre-TPP-22214 treatment time points and even significantly shortened diabetes-induced rod-driven b-wave implicit time by more than 40% (8.3 ms at t = 2 weeks vs. 14.2 ms at t = 0 weeks;**** p<0.0001).

[0430] Example 53 Comparison of human VEGF-A capture by Doppelmab and parent molecules TPP-14940 / TPP-14941 with and without disulfide bridge (CC)-Inhibition of VEGF-induced phosphorylation of VEGFR2 Figures 53A to 53B VEGF-A uptake was assessed by measuring VEGF receptor 2 (VEGFR2) phosphorylation at Y1175. 50 ng / ml human VEGF without molecular treatment served as control. Data represent mean + / - SEM. The corresponding best fit ICs are shown in Table 14 below. 50 Values ​​are reported (nonlinear regression; log(agonist) vs. response (three parameters)).

[0431] [Table 14]

[0432] result There were no significant differences between the parent molecules TPP-14940 / TPP-14941 and the Doppelmab variants with and without disulfide bridges. The potency and efficacy of inhibition of VEGF-induced VEGFR2 phosphorylation were virtually identical.

[0433] Example 54 Comparison of human VEGF-A uptake by TPP-22214 or Eylea-Y1175 inhibits VEGF-induced phosphorylation of VEGFR2 and ERK1 / 2, both associated with EC proliferation Figures 54A to 54B Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50ng / mL human VEGF with or without preincubation with increasing concentrations of TPP-22214 or Eylea. VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 (VEGFR2) phosphorylation at Y1175 or (B) ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2). 50ng / ml human VEGF without molecular treatment served as control. Data represent mean + / - SEM. The corresponding best fit values ​​are reported in Table 15 below (nonlinear regression (log(agonist) vs. response (3 parameters)). Best fit values ​​that are significantly different between Eylea and TPP-22214 (P<0.05) are shown in bold. P values ​​are also reported in Table 15.

[0434] [Table 15]

[0435] result Compared to Eylea, TPP-22214 completely inhibited VEGF-induced phosphorylation of VEGFR2 and ERK1 / 2 at Y1175. The improved efficacy of TPP-22214 in VEGF sequestering can also be seen in the highly significant difference in the bottom plateau values ​​(see Table 15). Finally, TPP-22214 was significantly more potent in VEGF sequestering than Eylea.

[0436] Example 55 Comparison of human VEGF-A uptake by TPP-22204, TPP-22214 or TPP-22216 with Eylea - inhibition of VEGF-induced proliferation of HRMEC Figures 55A to 55E Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 10 ng / mL human VEGF with or without preincubation with increasing concentrations of the indicated binding molecules: (A) TPPP-22204, (B) TPP-22214, (C) TPP-22216 or (D) Eylea. VEGF-A capture was assessed by automated image-based quantification of HRMEC cell number (IncuCyte). Images were recorded every 4 hours for a total of 84 hours. Relative cell numbers are shown. Cell number at t=0 was set to 1. Unstimulated cells served as control (basal). Data represent mean + / - SEM. (E) Plot of the difference between the area under the growth curve and the area under the basal curve against the concentration of Eylea or TPP-22214. The corresponding best fit values ​​are reported in Table 16 below (nonlinear regression (log(agonist) vs. response (three parameters)). Best fit values ​​that are significantly different between Eylea and TPP-22214 (P<0.05) are shown in bold. P values ​​are also reported in Table 16.

[0437] [Table 16]

[0438] result The potency and efficacy of Doppelmab in inhibiting VEGF-induced proliferation was clearly superior to Eylea (LogIC 50 and large significant differences in the values ​​of the bottom plateau).

[0439] Even at the highest concentration (16 nM), Eylea was not able to completely inhibit VEGF-induced sprouting. In contrast, all Doppelmab completely inhibited VEGF-induced proliferation already at a 64-fold lower concentration of 0.25 nM.

[0440] Example 56 Comparison of human VEGF-A uptake by TPP-2214 or Eylea - Inhibition of VEGF-induced phosphorylation of VEGFR2 and p38-MAPK with Y1214 Figures 56A to 56B Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50ng / mL human VEGF with or without preincubation with increasing concentrations of TPP-22214 or Eylea. VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 (VEGFR2) phosphorylation at Y1214 or (B) p38-MAPK phosphorylation at T180 / Y182. 50ng / ml human VEGF without molecular treatment served as control. Data represent mean + / - SEM. The corresponding best fit ICs are shown in Table 17 below. 50 Values ​​(nonlinear regression; (log(agonist) vs. response (three parameters)) are reported. Best fit values ​​significantly different between Eylea and TPP-22214 (P<0.05) are shown in bold. P values ​​are also reported in Table 17.

[0441] [Table 17]

[0442] result In sharp contrast to Eylea, TPP-22214 completely inhibited VEGF-induced phosphorylation of VEGFR2 and p38-MAPK at Y1214. The more effective VEGF capture of TPP-22214 can be seen in the highly significant difference in the bottom plateau values ​​in Table 17. Furthermore, a trend was observed in which the efficacy of VEGF capture by TPP-22214 was better than that of Eylea.

[0443] Example 57 Comparison of human VEGF-A uptake by TPP-22204, TPP-22214, TPP-22215 or TPP-22216 with Eylea - inhibition of VEGF-induced HRMEC sprouting Figures 57A to 57B Human retinal microvascular endothelial cell (HRMEC) spheroids were embedded in a collagen matrix. Endothelial sprouting was induced by incubation with 50 ng / mL human VEGF for 24 h with or without preincubation with 2.5 nM TPP-22204, TPP-22214, TPP-22215 or TPP-22216 or 5 nM Eylea. (A) Endothelial sprouting was assessed by confocal microscopy and spheroid circumferences obtained from maximum projections of Z-stacks are shown. Unstimulated cells served as control (basal). Data represent mean + / - SEM. ns p>0.05 no significant difference, **** p<0.0001. (B) Representative maximum projection images from spheroids after 24 h of sprouting under basal conditions or after stimulation with 50 ng / mL human VEGF with or without preincubation with 2.5 nM TPP-22214 or 5 nM Eylea. Bar = 100 μm.

[0444] result Similar to the Y1214 VEGFR2 phosphorylation assay, Doppelmab was consistently effective in inhibiting VEGF-induced HRMEC sprouting. Moreover, Doppelmab was able to completely inhibit VEGF-induced endothelial cell sprouting. Of note, Eylea was used at twice the molecular concentration compared to Doppelmab, but still failed to completely inhibit cell sprouting.

[0445] Example 58 Comparison of human VEGF-A uptake by TPP-22214 or Eylea-Y951 inhibits VEGF-induced phosphorylation of VEGFR2 and Src Figures 58A to 58B Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50ng / mL human VEGF with or without preincubation with increasing concentrations of TPP-22214 or Eylea. VEGF-A capture was assessed by measuring (A) VEGF receptor 2 (VEGFR2) phosphorylation at Y951 or (B) Src phosphorylation at Y419. 50ng / ml human VEGF without antibody treatment served as control. Data represent mean + / - SEM. The corresponding best fit values ​​are reported in Table 18 below (nonlinear regression; (log(agonist) vs. response (three parameters)). Best fit values ​​significantly different (P<0.05) between Eylea and TPP-22214 are shown in bold. P values ​​are also reported in Table 18.

[0446] [Table 18]

[0447] result In sharp contrast to Eylea, TPP-22214 completely inhibited VEGF-induced phosphorylation of VEGFR2 at Y951 and Src at Y419. The more effective VEGF-sequestering of TPP-22214 can be seen in the highly significant difference in the bottom plateau values ​​(Table 18).

[0448] Example 59 Comparison of human VEGF-A uptake by Doppelmab TPP-22214 and Eylea - Inhibition of human VEGF-A-induced retinal hyperpermeability in Brown Norway rats Figures 59A to 59B TPP-22214 prevented human VEGF-A-induced hyperpermeability in rat retina. (A) Time protocol showing experimental procedure. 15 min after intravitreal (ivt) administration of anti-VEGF compounds (13 or 26 pmol / eye Eylea or TPP-14940) or control (26 pmol TPP-11737), 13 pmol / eye human VEGF-A was administered by ivt injection. PBS injection served as control. 24 h later, 1 mL / kg Evans Blue (EB) solution (45 mg / mL in 0.9% saline) was administered by intravenous (iv) injection for 30 min, after which the eyes were isolated and fixed. At the same time points, plasma samples were collected to ensure equal systemic EB exposure. (B) Quantification of VEGF-A-induced hyperpermeability in Brown Norway rat retinas was performed by measuring EB extravasation in retinal flat mounts by confocal microscopy. The eyes were cut along the ora serrata, the lens and vitreous were removed, and the eyecups were fixed in paraformaldehyde (4%) for 1 h at 4°C, then transferred to PBS overnight at 4°C. The retina was separated from the outer segment (sclera and choroid), transferred to a glass slide, and cut four times to achieve a flat cloverleaf-like structure. The tissue was covered with mounting medium (Vectashield H-1200 containing the DNA stain DAPI) and a coverslip was placed on top to obtain a retinal flat mount. The samples were excited at a wavelength of 639 nm, and the emission of Evans blue at 669 nm was recorded with a LSM 700 confocal laser scanning microscope (Carl Zeiss, Jena; gain 800, laser intensity 2%, 5 stacks 60 μm) to obtain images of the retinal flat mount with maximum intensity projection. Analysis of the total fluorescence intensity was performed after opening the images in the program ImageJ with a threshold of 30. *** p<0.001; * *p<0.05; ns p>0.05. One-way ANOVA with Tukey's multiple comparison test. n=9–17. Incubation with Eylea at a molar ratio of 67:1. VEGF is shown for comparison.

[0449] result Under control conditions, intravitreal VEGF-A injection increased vascular permeability by approximately 60%. Note: TPP-11737 was used as a control molecule because it has the same molecular format (Doppelmab) as TPP-22214 but did not capture VEGF in initial in vitro assays (see above).

[0450] At a 1:1 molecular ratio of binding molecule to VEGF, TPP-22214 completely blocked vascular hyperpermeability, whereas, in sharp contrast, Eylea failed to significantly reduce vascular leakage under the same conditions and even at twice the concentration. Only at a very high molecular ratio of 67:1, Eylea was able to inhibit human VEGF-induced vascular leakage.

[0451] Example 60 Effect of TrkB binding on VEGF uptake by TPP-22214 - Comparison of human VEGF-A uptake by TPP-22214 in the presence or absence of the TrkB extracellular domain Figures 60A to 60B (A) Functional characterization of the TrkB extracellular domain (TrkB-ECD). CHO cells stably expressing human TrkB were incubated with increasing concentrations of the natural ligand BDNF or 10 nM BDNF with increasing concentrations of TrkB-ECD. TrkB activation was assessed by measuring TrkB phosphorylation at Y706 / 707. The lowest compound concentration was solvent alone. Data represent the mean + / - SEM.

[0452] (B) Effect of TPP-22214 binding to TrkB-ECD on inhibition of VEGF-induced VEGFR2 phosphorylation. Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF, 50 ng / mL human VEGF with increasing concentrations of TPP-22214, or 50 ng / mL human VEGF with increasing concentrations of TPP-22214 and 100 nM TrkB-ECD. HRMEC incubation with increasing concentrations of TrkB-ECD in the absence or presence of 50 ng / mL VEGF and unstimulated cells (basal) served as controls. VEGF-A uptake was assessed by measuring VEGF receptor 2 (VEGFR2) phosphorylation at Y1175. Data represent the mean + / - SEM.

[0453] result Figure (A): Preincubation of BDNF with increasing concentrations of TrkB-ECD dramatically reduced TrkB receptor activation, indicating that it is able to bind and capture BDNF, as would be expected from a functional TrkB-ECD.

[0454] Figure (B): VEGF uptake by TPP-22214 was independent of the presence or absence of TrkB-ECD. Both curves were nearly identical. This indicated that TrkB binding of TPP-22214 itself did not limit VEGF uptake by Doppelmab. This suggests that Doppelmab can simultaneously activate the TrkB receptor and capture VEGF.

[0455] Example 61 Manipulation efforts to improve CMC properties Series 3~Series 4 The biophysical properties of the molecules from series 3 and series 4 were evaluated. During the production of the molecules, the quality after Protein A was measured by the percentage of monomer was tested using analytical size exclusion chromatography. Additionally, the thermal stability and onset of aggregation of each molecule was evaluated in 10 mM histidine pH 6.0. Thermal stability was measured using a thermal shift assay, which measures the unfolding of a protein with temperature using Sypro-orange dye. Each T m was calculated as the peak maximum of the first derivative of the fluorescence signal over temperature. The onset of aggregation was measured using dynamic light scattering to measure the hydrodynamic radius as a function of temperature. Finally, the storage stability of the various Doppelmab formulations was tested by measuring aggregation after 2 weeks at either 40° C. or 5° C. Aggregation was measured using analytical size exclusion chromatography.

[0456] Molecular modeling was used to identify single point mutations that could potentially increase the pI of the Fv portion of Ranibizumab and therefore potentially improve its solubility without affecting its conformational stability. This computational design exercise suggested three framework mutations, namely VH E1→Q, VH E6→Q and VL D70G, ​​where VH and VL refer to the variable portions of the heavy and light chains in Ranibizumab Fv, respectively. The mutations in the heavy chain, VH E1→Q and VH E6→Q, correspond to the human germline residues at these positions. The mutation in the light chain, VL D70→G, is predicted to disrupt the solvent exposed negatively charged patch of Ranibizumab Fv.

[0457] Engineering efforts improved the overall biophysical properties of Doppelmab, translating into improved CMC properties for the entire molecule. An improvement in the proportion of monomers after Protein A was seen for most molecules tested, improving the overall manufacturability of the molecule. m 1 and T aggThe improvements seen in indicate that the manipulations improved the conformation and colloidal stability of the molecule, which further results in a better stability profile at both 40° C. and 5° C.

[0458] [Table 19]

[0459] Example 62 Measurement of vitreous half-life of TPP-22214 New Zealand White female rabbits were administered bilateral intravitreal injections of TPP-22214 or Bevacizumab. Vitreous samples were collected at various time points and concentrations of TPP-22214 or Bevacizumab were measured by ELISA. Animals were examined ophthalmologically before and at regular intervals after intravitreal injection.

[0460] [Table 20]

[0461] Table 20 shows the experimentally determined vitreous half-life results of TPP-22214 and Bevacizumab, comparing the in-house experimentally determined values ​​with the known literature values ​​of Bevacizumab, Ranibizumab and Aflibercept. The reported half-life of Bevacizumab is identical to the in-house / experimentally derived value, which confirms the accuracy of the applied method and the comparability of the data sets. TPP-22214 has been shown to have an approximately 50% higher in vivo half-life in rabbits compared to Bevacizumab, for example (6.7 days vs. 4.3 days).

[0462] [Table 21]

[0463] Table 21 reports literature values ​​for ocular VEGF concentrations in patients with wet age-related macular degeneration (wAMD).

[0464] [Table 22]

[0465] Based on the values ​​shown in Table 21, the human vitreous half-life (t 1 / 2 ) (9.7 days; Hutton-Smith, L., Mol. Pharmaceutics, 2016, 13, 2941-2950), a conservative estimate of the human dosing frequency of TPP-22214 was calculated (Table 21). At a dose of 2 mg / eye, a dosing interval of 4 months is reasonable.

[0466] Example 63 Comparison of human TrkB activation by BDNF, TPP-6830, and the two doppelmabs TPP-23457 and TPP-23459 Figures 61A to 61B CHO cells stably expressing human TrkB were incubated with increasing concentrations of the natural TrkB ligand BDNF, TPP-6830 (an additional monoclonal TrkB antibody) or the two Doppelmab TPP-23457 (TPP-6380 as scFv, 10L1, VH-VL; Ranibizumab as Fab) and TPP-23459 (TPP6380 as scFv, 10L1, VL-VH; Ranibizumab as Fab).

[0467] Here, the novel Doppelmab molecule was based on a different TrkB binder (TPP-6830) to assess whether the effects observed with Doppelmab based on the C2 TrkB binder could also be reproduced with a different TrkB binder.

[0468] result Again, in stark contrast to TPP-6830 (the original TrkB antibody, which is a partial TrkB receptor agonist), the two novel Doppelmab TPP-23457 and TPP-23459 Doppelmab now showed full TrkB receptor agonist activity, and TrkB activation by these Doppelmas was at least as effective as that of the natural ligand, BDNF.

[0469] This further supports the theory that the conformation of the binding molecule may be responsible for the observed increased efficacy, independent of the specific TrkB binder.

[0470] Example 64 Comparison of human TrkB activation (TrkB phosphorylation) by TPP-6830 (a TrkB monoclonal antibody) and the two doppelmab antibodies TPP-23457 and TPP-23459 in the presence or absence of human VEGF Figures 62A to 62C CHO cells stably expressing human TrkB were incubated with increasing concentrations of Doppelmab (A) TPP-23457, (B) TPP-23459, or (C) TPP-6830, with or without preincubation with 200 ng / mL human VEGF-A (hVEGF). TrkB activation was assessed by measuring TrkB phosphorylation at Y706 / 707. The lowest compound concentration was vehicle alone. Data represent the mean + / - SEM.

[0471] result Again, preincubation with human VEGF dramatically improved the efficacy of TrkB phosphorylation (TrkB activation) by TPP-23457 (TPP-6830 as scFv, 10L1, VH-VL; Ranibizumab as Fab) and TPP-23459 (TPP-6380 as scFv, 10L1, VL-VH; Ranibizumab as Fab).

[0472] VEGF preincubation did not affect the efficacy of TrkB phosphorylation (TrkB activation) by TPP-6830 (control).

[0473] This further supports the theory that the observed increased potency is independent of the specific TrkB binder.

[0474] Example 65 Comparison of human TrkB activation (ERK phosphorylation) by TPP-6830 (a TrkB monoclonal antibody) and the two doppelmabs TPP-23457 and TPP-23459 in the presence or absence of human VEGF Figures 63A to 63C CHO cells stably expressing human TrkB were incubated with increasing concentrations of Doppelmab (A) TPP-23457, (B) TPP-23459, or (C) TPP-6830, with or without preincubation with 200 ng / mL human VEGF-A (hVEGF). TrkB activation was assessed by measuring ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2), downstream of TrkB. The lowest compound concentration was vehicle alone. Data represent the mean + / - SEM.

[0475] result Again, preincubation with human VEGF dramatically improved the potency of ERK1 / 2 phosphorylation (TrkB activation) by TPP-23457 (TPP-6830 as scFv, 10L1, VH-VL; Ranibizumab as Fab) and TPP-23459 (TPP-6380 as scFv, 10L1, VL-VH; Ranibizumab as Fab).

[0476] As shown previously, VEGF preincubation did not affect the potency of ERK1 / 2 phosphorylation (TrkB activation) by TPP-6830 (control), further supporting the theory that the observed enhancement in potency is independent of a specific TrkB binder.

[0477] Example 66 Effect of TPP-23457 on BDNF-induced TrkB activation (TrkB phosphorylation) Figures 64A to 64B CHO cells stably expressing the human TrkB receptor were incubated with increasing concentrations of (A) C2 antibody or (B) Doppelmab TPP-23457 in the presence or absence of constant concentrations of 0.3 nM, 1 nM or 3 nM BDNF. Activation of TrkB was assessed by measuring receptor phosphorylation at Y706 / 707. Data represent the mean + / - SEM.

[0478] result In sharp contrast to C2, TPP-23457 did not limit BDNF-induced TrkB activation.

[0479] Example 67 Effect of TPP-23457 on BDNF-induced TrkB activation (Erk phosphorylation) Figures 65A to 65B CHO cells stably expressing the human TrkB receptor were incubated with increasing concentrations of (A) C2 antibody or (B) Doppelmab TPP-23457 in the presence or absence of constant concentrations of 0.3 nM, 1 nM or 3 nM BDNF. Activation of TrkB was assessed by measuring ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2), downstream of TrkB. Data represent the mean + / - SEM.

[0480] result In sharp contrast to C2, TPP-23457 did not limit BDNF-induced Erk phosphorylation, which is downstream of TrkB.

[0481] Example 68 Comparison of human VEGF-A uptake by TPP-22214, TPP-23457 or TPP-23459 inhibition of VEGF-induced phosphorylation of VEGFR2 and ERK1 / 2 at Y1175, both of which are associated with EC proliferation Figures 66A to 66B Human retinal microvascular endothelial cells (HRMEC) were starved and then incubated with 50 ng / mL human VEGF with or without preincubation with increasing concentrations of TPP-22214, TPP-23457, or TPP23459. VEGF-A uptake was assessed by measuring (A) VEGF receptor 2 (VEGFR2) phosphorylation at Y1175 or (B) ERK1 / 2 phosphorylation at T202 / Y204 (ERK1) and T185 / Y187 (ERK2). 50 ng / ml human VEGF without molecular treatment served as control. Data represent mean + / - SEM.

[0482] result VEGF uptake by Doppelmab TPP-23457 and TPP-23459 was virtually identical to that of TPP-22215.

[0483] As expected, Doppelmbab, based on a different TrkB scFv binder (but not TPP-6830 and C2), also did not affect the VEGF-capturing function of the Fab.

Claims

1. Two or more, including at least one antigen-binding site that specifically binds to vascular endothelial growth factor (VEGF) and at least one antigen-binding site that specifically binds to tropomyosin receptor kinase B (TrkB). It is a bispecific and tetravalent binding molecule, The antigen-binding site that specifically binds to VEGF is the antigen-binding site a)-c): a) an antigen binding site containing a light chain CDR containing the amino acid sequences of SEQ ID NO: 145 (CDR1), SEQ ID NO: 146 (CDR2) and SEQ ID NO: 147 (CDR3); or an antigen binding site containing a heavy chain CDR containing the amino acid sequences of SEQ ID NO: 148 (CDR1), SEQ ID NO: 149 (CDR2) and SEQ ID NO: 150 (CDR3). Antigen binding sites or containing heavy chain CDRs containing the amino acid sequences of SEQ ID NO: 151 (CDR1), SEQ ID NO: 152 (CDR2) and SEQ ID NO: 153 (CDR3). an antigen binding site comprising a heavy chain CDR containing the amino acid sequences of SEQ ID NO: 154 (CDR1), SEQ ID NO: 155 (CDR2) and SEQ ID NO: 156 (CDR3); b) an antigen binding site containing a light chain CDR containing the amino acid sequences of SEQ ID NO: 157 (CDR1), SEQ ID NO: 158 (CDR2) and SEQ ID NO: 159 (CDR3); or an antigen binding site containing a heavy chain CDR containing the amino acid sequences of SEQ ID NO: 160 (CDR1), SEQ ID NO: 161 (CDR2) and SEQ ID NO: 162 (CDR3). or an antigen binding site containing a heavy chain CDR containing the amino acid sequences of SEQ ID NO: 163 (CDR1), SEQ ID NO: 164 (CDR2) and SEQ ID NO: 165 (CDR3). an antigen binding site comprising a heavy chain CDR containing the amino acid sequences of SEQ ID NO: 166 (CDR1), SEQ ID NO: 167 (CDR2) and SEQ ID NO: 168 (CDR3); c) an antigen binding site comprising a light chain CDR containing the amino acid sequences of SEQ ID NO: 169 (CDR1), SEQ ID NO: 170 (CDR2) and SEQ ID NO: 171 (CDR3); or an antigen binding site containing a heavy chain CDR containing the amino acid sequences of SEQ ID NO: 172 (CDR1), SEQ ID NO: 173 (CDR2) and SEQ ID NO: 174 (CDR3). Antigen binding sites or containing heavy chain CDRs containing the amino acid sequences of SEQ ID NO: 175 (CDR1), SEQ ID NO: 176 (CDR2) and SEQ ID NO: 177 (CDR3). Antigen binding sites containing heavy chain CDRs containing the amino acid sequences of SEQ ID NO: 178 (CDR1), SEQ ID NO: 179 (CDR2) and SEQ ID NO: 180 (CDR3). is selected from the group consisting of The antigen binding site that specifically binds to TrkB is a) an antigen binding site containing a light chain CDR containing the amino acid sequences of SEQ ID NO: 201 (CDR1), SEQ ID NO: 202 (CDR2) and SEQ ID NO: 203 (CDR3); or antigen binding sites containing heavy chain CDRs containing the amino acid sequences of SEQ ID NO: 204 (CDR1), SEQ ID NO: 205 (CDR2) and SEQ ID NO: 206 (CDR3). or an antigen binding site containing a heavy chain CDR containing the amino acid sequences of SEQ ID NO: 207 (CDR1), SEQ ID NO: 208 (CDR2) and SEQ ID NO: 209 (CDR3). Antigen binding sites containing heavy chain CDRs containing the amino acid sequences of SEQ ID NO: 210 (CDR1), SEQ ID NO: 211 (CDR2) and SEQ ID NO: 212 (CDR3). is selected from The binding molecule is SEQ ID NO: 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29, 31, 33, 35, 37, 39, 41; 43, 45, 47, 49, 51, 53, 55, 57, 59, 61, 63, 65, 67, 69, 71, 73, 75, 77, 79, 81, 83, 85, 8 7, 89, 91, 93, 95, 97, 99, 101, 103, 105, 107, 109, 111, 113, 115, 117, 119, 121, 123, 125, 127, 129, 131, 133, 135, A light chain containing an amino acid sequence that is at least 90% identical to the amino acid sequence of 137, 139, 141 or 143 and SEQ ID NO: 2, 4, 6, 8, 10, 12, 14, 16, 18, 20, 22, 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 44, 46 ,48,50,52,54,56,58,60,62,64,66,68,70,72,74,76,78,80,82,84,86,88,90, 92, 94, 96, 98, 100, 102, 104, 106, 108, 110, 112, 114, 116, 118, 120, 122, 124, 126, 128, 130, 132, 134, 136, 138, 140, 142 or 144 amino acid sequences, including a heavy chain containing at least 90% identical amino acid sequences; Binding molecules.

2. Two or more, including at least one antigen-binding site that specifically binds to vascular endothelial growth factor (VEGF) and at least one antigen-binding site that specifically binds to tropomyosin receptor kinase B (TrkB). It is a bispecific and tetravalent binding molecule, (i) a light chain containing the amino acid sequence of SEQ ID NO: 41 and a heavy chain containing the amino acid sequence of SEQ ID NO: 42; (ii) a light chain containing the amino acid sequence of SEQ ID NO: 43; and an amino acid sequence of SEQ ID NO: 44; A heavy chain comprising (iii) a light chain comprising the amino acid sequence of SEQ ID NO: 45 and a heavy chain comprising the amino acid sequence of SEQ ID NO: 46 or (iv) an amino acid sequence of SEQ ID NO:

47. A heavy chain comprising a light chain comprising an amino acid sequence of SEQ ID NO: 48 or (v) a light chain comprising an amino acid sequence of SEQ ID NO: 49 and a heavy chain comprising an amino acid sequence of SEQ ID NO: 50 or (vi) a heavy chain comprising an amino acid sequence of SEQ ID NO: 51 a light chain containing the amino acid sequence of SEQ ID NO: 52 or a light chain containing the amino acid sequence of SEQ ID NO: 53 and a heavy chain containing the amino acid sequence of SEQ ID NO:

54. Heavy chain comprising a column or (viii) a light chain comprising an amino acid sequence of SEQ ID NO: 55 and a heavy chain comprising an amino acid sequence of SEQ ID NO: 56 or (ix) a light chain comprising an amino acid sequence of SEQ ID NO: 57 and a light chain comprising an amino acid sequence of SEQ ID NO: 57 : A heavy chain containing the amino acid sequence 58 or (x) a light chain containing the amino acid sequence 59 and a heavy chain containing the amino acid sequence 60 or (xi) a heavy chain containing the amino acid sequence 68 or (x) a light chain containing the amino acid sequence 69 A light chain containing the amino acid sequence of 1 and a heavy chain containing the amino acid sequence of SEQ ID NO: 62 or (xii) a light chain containing the amino acid sequence of SEQ ID NO: 63 and a heavy chain containing the amino acid sequence of SEQ ID NO: 64 or (xiii) a light chain containing the amino acid sequence of SEQ ID NO: 64 ) a light chain containing the amino acid sequence of SEQ ID NO: 65 and a heavy chain containing the amino acid sequence of SEQ ID NO: 66 or (xiv) a light chain containing the amino acid sequence of SEQ ID NO: 67 and a light chain containing the amino acid sequence of SEQ ID NO:

67. A heavy chain comprising the amino acid sequence of column number: 68 or a light chain comprising the amino acid sequence of SEQ ID NO: 69 and a heavy chain comprising the amino acid sequence of SEQ ID NO: 70 or (xvi) an amino acid sequence of SEQ ID NO:

71. Includes a light chain comprising a heavy chain comprising an amino acid sequence of SEQ ID NO: 72 or (xvii) a light chain comprising an amino acid sequence of SEQ ID NO: 73 and an amino acid sequence of SEQ ID NO: 74 Heavy chain or (xviii) a light chain comprising the amino acid sequence of SEQ ID NO: 75 and a heavy chain comprising the amino acid sequence of SEQ ID NO: 76 or (xix) a light chain comprising the amino acid sequence of SEQ ID NO: 77 and a light chain comprising the amino acid sequence of SEQ ID NO: 78 heavy chain comprising the amino acid sequence or (xx) light chain comprising the amino acid sequence of SEQ ID NO: 79 and heavy chain comprising the amino acid sequence of SEQ ID NO: 80 or (xxi) SEQ ID NO: 8 A light chain containing the amino acid sequence of 1 and a heavy chain containing the amino acid sequence of SEQ ID NO: 82 or (xxii) a light chain containing the amino acid sequence of SEQ ID NO: 83 and a heavy chain containing the amino acid sequence of SEQ ID NO: 84 or (xxiii) a light chain containing the amino acid sequence of SEQ ID NO: 84 ) a light chain containing the amino acid sequence of SEQ ID NO: 85 and a heavy chain containing the amino acid sequence of SEQ ID NO: 86 or (xxiv) a light chain containing the amino acid sequence of SEQ ID NO: 87 A heavy chain comprising the chain and the amino acid sequence of SEQ ID NO: 88 or (xxv) a light chain comprising the amino acid sequence of SEQ ID NO: 89 and a heavy chain comprising the amino acid sequence of SEQ ID NO: 90 or (xxvi) a amino acid of SEQ ID NO: 91 A light chain containing the sequence and a heavy chain containing the amino acid sequence of SEQ ID NO: 92 or a light chain containing (xxvii) an amino acid sequence of SEQ ID NO: 93 and an amino acid sequence of SEQ ID NO: 94 Heavy chain comprising an acid sequence or (xxviii) a light chain comprising an amino acid sequence of SEQ ID NO: 95 and a heavy chain comprising an amino acid sequence of SEQ ID NO: 96 or (xxix) a light chain comprising an amino acid sequence of SEQ ID NO: 97 A heavy chain comprising the amino acid sequence number: 98 or a light chain comprising the amino acid sequence number: 99 and a heavy chain comprising the amino acid sequence number: 100 or a light chain comprising the amino acid sequence number: 90 xxxi) a light chain containing the amino acid sequence of SEQ ID NO: 101 and a heavy chain containing the amino acid sequence of SEQ ID NO: 102 or (xxxii) a light chain containing the amino acid sequence of SEQ ID NO: 103 and an amino acid sequence of SEQ ID NO: 104 Heavy chain or (xxxiii) a light chain containing the amino acid sequence of SEQ ID NO: 105 and a heavy chain containing the amino acid sequence of SEQ ID NO: 106 or (xxxi) a light chain containing the amino acid sequence of SEQ ID NO: 106 v) A light chain containing the amino acid sequence of SEQ ID NO: 107 and a heavy chain containing the amino acid sequence of SEQ ID NO: 108 or (xxxv) a light chain containing the amino acid sequence of SEQ ID NO: 109 and a amino acid sequence of SEQ ID NO: 110 Heavy chain or (xxxvi) light chain containing the amino acid sequence of SEQ ID NO: 111 and heavy chain or (xxxvii) sequence containing the amino acid sequence of SEQ ID NO: 112 A light chain comprising the amino acid sequence number: 113 and a heavy chain comprising the amino acid sequence number: 114 or (xxxviii) a light chain comprising the amino acid sequence number: 115 and a heavy chain comprising the amino acid sequence number: 116 (xxxix) A light chain containing the amino acid sequence of SEQ ID NO: 117 and a heavy chain containing the amino acid sequence of SEQ ID NO: 118 or (xl) SEQ ID NO: 119 and a heavy chain comprising the amino acid sequence of SEQ ID NO: 120 or (xli) a light chain comprising the amino acid sequence of SEQ ID NO: 121 and a heavy chain comprising the amino acid sequence of SEQ ID NO: 122 or (xlii) a light chain comprising the amino acid sequence of SEQ ID NO: 122 A light chain containing the amino acid sequence of SEQ ID NO: 123 and a heavy chain containing the amino acid sequence of SEQ ID NO: 124 or (xliii) an amino acid sequence of SEQ ID NO: 125 and a heavy chain comprising the amino acid sequence of SEQ ID NO: 126 or (xliv) a light chain comprising the amino acid sequence of SEQ ID NO: 127 and a heavy chain comprising the amino acid sequence of SEQ ID NO: 128 or (xliv) a light chain comprising the amino acid sequence of SEQ ID NO: 128 or (xliv) a light chain comprising the amino acid sequence of SEQ ID NO: 128 Light chains containing 129 amino acid sequences and heavy chains containing the amino acid sequence of SEQ ID NO: 130 or (xlvi) light chains containing the amino acid sequence of SEQ ID NO: 131 A heavy chain comprising the amino acid sequence number: 132 or (xlvii) a light chain comprising the amino acid sequence number: 133 and a heavy chain comprising the amino acid sequence number: 134 or (xlviii) a amino acid sequence number: 135 Light chain and heavy chain comprising the amino acid sequence of SEQ ID NO: 136 or (xlix) light chain comprising the amino acid sequence of SEQ ID NO: 137 and SEQ ID NO: 1 A heavy chain comprising 38 amino acid sequences or (l) a light chain comprising the amino acid sequence of SEQ ID NO: 139 and a heavy chain comprising the amino acid sequence of SEQ ID NO: 140 or (li) a light chain comprising the amino acid sequence of SEQ ID NO: 141 and a heavy chain comprising the amino acid sequence of SEQ ID NO: 142 or (lii) a light chain comprising the amino acid sequence of SEQ ID NO: 143 and a heavy chain comprising the amino acid sequence of SEQ ID NO: 144 Including, Binding molecules.

3. encoding (i) a heavy chain or heavy chain variable domain and / or (ii) a light chain or light chain variable domain of the binding molecule according to claim 1 or 2; Isolated nucleic acid molecules.

4. 4. comprising the isolated nucleic acid molecule of claim 3, Viral vector.

5. 4. Contains the nucleic acid molecule of claim 3, Expression vector.

6. 6. Transfected with the expression vector of claim 5; Host cell.

7. (a) culturing the host cell according to claim 6 under conditions that allow expression of the following molecules; (b) recovering the molecule; 3. The method of producing a binding molecule according to claim 1 or 2.

8. 3. The binding molecule according to claim 1 or 2 for use in a pharmaceutical.

9. 9. The binding molecule for use according to claim 8, wherein the use is the treatment of an ocular or retinal disease or neurodegenerative disease.

10. 10. The binding molecule for use according to claim 9, wherein the use is for the treatment of neural / neuropean or retinal diseases.

11. The use includes macular degeneration, age-related macular degeneration, diabetic retinopathy, diabetic macular edema, retinal pigmentosa, hereditary retinal dystrophy, hereditary macular dystrophy, myopic degeneration, maple atrophy, retina Arterial occlusion, endophthalmitis, uveitis, cystic macular edema, choroidal neovascular membrane secondary to any retinal disease, optic neuropathy, glaucoma, retinal detachment, toxic retinopathy, radiation retinopathy and traumatic retinopathy , precursor and mild to moderate art Delayed disease progression in patients with Heimers' disease, Huntington's disease, Parkinson's disease, Treatment of cognitive disorders associated with major depressive disorder, schizophrenia, schizophrenia, and initial psychosis in individuals with attenuated psychosis syndrome 10. The binding molecule for use according to claim 9, which is for the prevention of, prevention of recurrence in patients with schizophrenia, for the treatment of treatment-resistant depression, bulimia, obesity or metabolic syndrome.

12. 10. The binding molecule for use according to claim 9, wherein the use is for the treatment of exudative age-related macular degeneration (wAMD).

13. 10. The binding molecule for use according to claim 9, wherein the use is for the treatment of retinal venous occlusion (RVO).

14. 10. The binding molecule for use according to claim 9, wherein the use is for the treatment and / or prevention of mapped atrophy.

15. 3. comprising a pharmaceutically acceptable carrier and a binding molecule according to claim 1 or 2; Pharmaceutical composition.

16. 3. Use of the binding molecule according to claim 1 or 2 for the manufacture of a pharmaceutical composition for treating or preventing an ocular or retinal disease or a neurodegenerative disease.

17. 3. The binding molecule of claim 1 or 2 for use in a method of treating or preventing an ocular or retinal disease or neurodegenerative disease.

Images