Single-domain antibodies targeting the heparin-binding EGF-like growth factor

Abstract

The invention relates to a single domain antibody that binds Heparin-binding EGF-like growth factor (HB-EGF) for use in preventing, treating and / or diagnosing e.g. a brain disease, wherein the single domain antibody is in combination with a therapeutic compound and / or a diagnostic / imaging compound. The single domain antibody may in particular be a Variable domain of a Heavy chain only (VHH) antibody of camelid origin or an engineered single-domain antibody (sdAb) of other mammalian and be capable of transporting cargo such as peptides or other molecules into the brain across the blood-brain barrier (BBB) by receptor-mediated transcytosis (RMT). The brain disease may for example be Alzheimer's Disease, Parkinson's Disease, Huntington's Disease, brain tumor, or Hunter syndrome.

Description

[0001] SINGLE-DOMAIN ANTIBODIES TARGETING THE HEPARIN-BINDING EGF-LIKE GROWTH FACTOR

[0002] Technical field

[0003] The present invention relates to the field of binding polypeptides, particularly for their use in the prevention, diagnosis, imaging and / or treatment of brain diseases such as brain tumors, Alzheimer’s Disease, Parkinson’s Disease, Huntington’s Disease, brain ischemia, metabolic diseases, neurodevelopmental disorders, or meningitis and encephalitis.

[0004] Background of the invention

[0005] The blood-brain barrier (BBB) represents a significant challenge for efficient drug delivery into the brain. The BBB maintains brain homeostasis via a tightly integrated multi-cellular structure and strictly selective transport mechanisms [1, 2], which usually keeps most molecules (such as pathogens) out of the brain.

[0006] One strategy to selectively transport drugs across the BBB is to harness the receptor-mediated transcytosis (RMT) mechanism. RMT is a selective vesicular transport mechanism of brain endothelial cells, which facilitates the transport of large cargoes, such as peptides and proteins, across the BBB [3, 4], RMT is initiated by ligand-binding to its specific receptor expressed on the apical side of polarized brain endothelial cells, followed by the trafficking of the ligandreceptor complex in vesicles through the cell, and the release of the ligand at the basolateral side of the endothelial cell for its subsequent transport into the cerebrospinal fluid (CSF) [4],

[0007] Hence, a therapeutic molecule conjugated to such a ligand can be delivered into the brain in a targeted manner. To date, several RMT-enabled receptors presented on the BBB have been explored for cerebral drug delivery, such as the transferrin receptor (TfR), insulin receptor (InsR), and low-density lipoprotein receptor (LDLR) [1, 3, 5, 6], However, these approaches share therapeutic limitations due to their binding and transport competition with endogenous ligands. This results in reduced drug delivery efficiency, or impediment of necessary cerebral nutrient transport, which leads to metabolic dysfunctions or cytotoxicity [7, 8],

[0008] ProHB-EGF is the trans-membrane precursor of the heparin-binding EGF-like growth factor (HB-EGF). Apart from the juxtracrine effect, proHB-EGF undergoes ectodomain shedding to generate its mature version of the soluble HB-EGF (sHB-EGF), which binds to its downstream receptors (such as EGFR) to impose different physiological functions with regards to cell proliferation, survival, migration, and differentiation [9, 10], HB-EGF is constitutively expressed in the brain, like on endothelial cells, glial cells, and neurons [9],

[0009] Cross-reacting material 197 (CRM197), a non-toxic derivative of the Diphtheria toxin (DT)

[0014] , was found to be capable of binding proHB-EGF (also called DT receptor) and to carry large proteins (serving as proof-of-concept for drugs) across the BBB via RMT

[0013] , While CRM 197 represented a promising BBB shuttle molecule, it was found to retain a certain level of toxicity, causing the inhibition of protein synthesis and endonucleolytic degradation of DNA [15-18], In addition, it was found that this approach shows a limited efficiency in people who have been vaccinated against Diphtheria.

[0010] It is an objective of the present disclosure to overcome one or more of the above-identified or other problems in the art and / or to provide a new or improved strategy for the prevention, diagnosis and / or treatment of brain diseases, such as brain tumors, Alzheimer’s Disease, Parkinson’s Disease, Huntington’s Disease, Hunter syndrome, neurodevelopmental disorders or meningitis and encephalitis.

[0011] Summary of the invention

[0012] The present inventors present a novel approach utilizing the proHB-EGF-mediated BBB transcytosis. Previously, others have demonstrated the employment of single-domain antibodies derived from camelid heavy-chain-only antibodies, also known as the variable heavy chain of the heavy-chain-only antibody (VHH) or single-domain antibody (sdAb), for the transcytosis by the human polylmmunoglobulin receptor (plgR)

[0021] ,

[0013] Being among the smallest antigen-binding units ever developed, sdAbs (-15 kDa) have gained particular attention over full-sized antibodies (-150 kDa) in the field of targeted drug delivery due to lower immunogenicity, better structural stability, and deeper tissue penetration [22, 23, 21], To date, several sdAbs have been generated that were able to transcytose the BBB by specifically targeting receptors such as TfR, a(2,3)-sialoglycoprotein receptor, insulin-like growth factor 1 receptor (IGF1R) and vascular cell adhesion molecule 1 (VCAM-1) [24, 25],

[0014] The present inventors have identified multiple sdAbs, which selectively bind to proHB-EGF and sHB-EGF. These sdAbs were generated by immunization of llamas with the ectodomain of human proHB-EGF and sHB-EGF, followed by phase display selections. The inventors confirmed that the identified sdAbs can undergo RMT using in vitro transwell BBB models comprising of either primary human brain microvascular endothelial cells (hBMECs) or induced human pluripotent stem cell (hiPSC)-derived BMEC-like cells. In addition, they showed RMT using a neurovascular unit model in an organopiate microfluidic culture platform. With promising features, these proHB-EGF targeted sdAbs can be used as BBB shuttle molecules for drug delivery into the brain.

[0015] The inventors considered that RMT-mediated drug delivery by utilizing receptors highly involved in important nutrient transport across the BBB may have safety concerns. ProHB-EGF is the transmembrane precursor of the heparin-binding EGF-like growth factor (HB-EGF) presented on the luminal membrane side of brain endothelial cells. This protein is characterized as an internalizing trans-membrane protein with so far not any known endogenous ligands. Hence, no foreseeable issues in terms of binding competition or blocking of the essential nutrients transport can be expected. Moreover, under diseased conditions such as ischemia or neuroinflammation, proHB-EGF gets highly enriched in the endothelium, which could offer a convenient window for improved drug delivery into the diseased brain [11, 12], Altogether, these make proHB-EGF an applicable candidate for RMT and transport across the BBB.

[0016] In this disclosure, the present inventors show the selection and characterization of multiple sdAbs with high binding affinity for proHB-EGF, and the testing of their BBB transcytosis in vitro. These sdAbs were generated by immunization of llamas with the ectodomain of human proHB-EGF and sHB-EGF, followed by phase display selections. The ability of the sdAbs to transcytose was demonstrated in a primary human cell-based BBB model and by a human induced pluripotent stem cell (hiPSC)-based iBBB model in a transwell setup. In addition, the successful transcytosis of the sdAbs was demonstrated in a neurovascular unit grown in an organopiate microfluidic culture platform. The present inventors have shown that two sdAbs candidates in particular, i.e. F12 and H7, targeting proHB-EGF are promising BBB shuttle molecules for delivering therapeutic molecules into the brain.

[0017] Detailed description of the invention

[0018] The present disclosure provides for a binding polypeptide, preferably an antibody or functional fragment thereof, most preferably a single domain antibody (sdAb) (e.g. variable domain of the heavy chain of a heavy-chain-only (VHH) antibody, that binds Heparin-binding EGF-like growth factor (HB-EGF), preferably in the form of proHB-EGF.

[0019] HB-EGF may exist in different forms depending on its cellular processing. These forms include pre-proHB-EGF, proHB-EGF, and sHB-EGF:

[0020] - Pre-proHB-EGF is the full-length, nascent precursor form of HB-EGF initially synthesized in the cell. It includes a signal peptide at its N-terminus, which directs the nascent protein to the endoplasmic reticulum (ER) for further processing; - ProHB-EGF is the transmembrane precursor form of HB-EGF after the signal peptide has been cleaved off. The extracellular domain can be released as an active growth factor (soluble HB-EGF);

[0021] - sHB-EGF (soluble HB-EGF) is the soluble form of HB-EGF that is released from the cell membrane following the proteolytic cleavage (shedding off) of the extracellular domain of proHB-EGF by enzymes like ADAM (a disintegrin and metalloprotease) family members.

[0022] In a particularly preferred embodiment, the binding polypeptide according to the present disclosure is a single domain antibody (sdAb). Herein, the term ‘antibody’ refers to a protein, preferably belonging to the immunoglobulin class of proteins, that may exhibit binding specificity to a target and induce molecular or cellular responses. Binding specificity as used herein refers to the ability of antibodies to discriminate between target (epitope) and (any) another molecule or protein in the given context. As used herein, the term ‘functional fragment’ refers to a portion of an antibody, which may be smaller in size and / or may lack certain parts of the full-sized antibody, but still retain the ability to bind its target. Single-domain antibody typically refers to a (functional) fragment of a (mammalian / human) antibody. In the present disclosure, a singledomain antibody can particularly refer to a single monomeric variable antibody domain derived from the variable domain of the heavy chain of a heavy-chain-only (VHH) antibody (which is devoid of light chains. VHHs can be found in camelidae or cartilaginous fish (also referred to as VHH and VNAR, respectively). The binding polypeptides according to the present disclosure preferably have a molecular weight between 1-200 kDa, 2-100 kDa, 3-90 kDa, 4-80 kDa, 5-70 kDa, 6-60 kDa, 8-50 kDa or 10-40 kDa. The binding peptide according to the present disclosure may be recombinantly engineered and / or humanized.

[0023] The binding polypeptide according to the disclosure is preferably capable of transporting any cargo across the blood-brain barrier (BBB) into a recipient’s brain. The transport of cargo such as large molecules into the brain is strictly regulated by the BBB. The present binding peptide can utilize receptor-mediated transcytosis (RMT), which can be seen as a targeted vesicular transport mechanism of the brain endothelial cells, to specifically transport (large) cargoes or molecules into the brain. The binding peptide, or single domain antibody or VHH according to the present disclosure is preferably capable of (active) transport of any cargo into the cerebrospinal fluid. The cargoes, e.g. compounds to be transported, can be used for therapeutic, imaging, and diagnostic purposes.

[0024] The binding polypeptide according to the present disclosure preferably is in combination with a compound, preferably a therapeutic compound and / or diagnostic or imaging compound, and / or can be used preferably for preventing, treating and / or diagnosing a brain disease. Also foreseen is use of the binding polypeptide according to the present disclosure, preferably in combination with an (imaging) compound) for imaging a brain disease, e.g. imaging the progression and / or state of brain disease.

[0025] In other words, the present disclosure provides for a therapeutic / diagnostic / imaging method (or method for transporting cargo across the BBB), which may comprise administering the binding polypeptide, e.g., in combination with a therapeutic, imaging and / or a diagnostic compound, to a subject in need thereof. Accordingly, the present disclosure provides for the use of the binding peptide, e.g. in combination with a therapeutic compound, imaging and / or a diagnostic compound, in the manufacturing of a medicament, imaging and / or diagnostic compound in the treatment / diagnosis of a brain disease. As disclosed herein, the term ‘subject’ may preferably refer to a mammal, preferably a human.

[0026] The therapeutic compound, imaging and / or diagnostic compound may be conjugated to the (individual) binding polypeptide according to the present disclosure. There are various ways to achieve this, for example in the case of a single-domain antibody this can be performed using: - genetic fusion, wherein for example the single-domain antibody and the therapeutic compound, imaging and / or diagnostic compound are comprised in a single polypeptide chain, and / or produced by expressing a vector comprising coding nucleic acid sequences of the singledomain antibody and the therapeutic compound, imaging and / or a diagnostic compound, preferably separated by a linker sequence such as a G4S-linker;

[0027] - Cys-Maleimide conjugation. For example, the thiol group (R-SH) of an introduced free cysteine residue in the C-terminal region (such as outside the sdAb sequence) reacts with a maleimide group introduced into the therapeutic, imaging and / or a diagnostic compound forming a stable thioether (C-S) bond;

[0028] - NHS-ester chemistry. For example, NHS (N-hydroxysuccinimide) esters react with primary amines (found on lysine residues or the N-terminus of proteins), forming stable amide bonds;

[0029] Carbodiimide crosslinking (EDC / NHS), specifically using EDC (1-ethyl-3-(3-dimethylaminopropyl)carbodiimide), activates carboxyl groups to react with primary amines, forming amide bonds. Often, NHS is used to stabilize the intermediate and increase efficiency; - Click chemistry, particularly strain-promoted azide-alkyne cycloaddition (SPAAC), typically involves the reaction between an azide and a cyclooctyne (strained alkyne) to form a stable triazole linkage;

[0030] - Protein ligation. For example by using sortase, intein, or transglutaminase.

[0031] Sortase: An enzyme that recognizes a specific peptide sequence (LPXTG) and facilitates the site-specific ligation of proteins. Intein: A protein splicing element that can ligate exteins (external protein sequences) through a trans-splicing reaction.

[0032] Transglutaminase: An enzyme that forms isopeptide bonds between glutamine and lysine residues.

[0033] Herein, the term ‘conjugate(d)’ may be used interchangeably with the term ‘couple(d)’ and refers to any means by which two or more components may be attached and / or combined and / or interact, for example by covalent bonds, ionic bonds, hydrogen bonds, van der Waals forces, polar bonds, non-polar bonds and / or hydrophobic interactions, preferably covalent bonds, or by (peptide) linker, (enzymatically) cleavable peptide linker.

[0034] In one example, the therapeutic compound, imaging and / or a diagnostic compound is conjugated to the (single-domain) antibody by coupling to a cysteine, preferably an unpaired cysteine, preferably C-terminally from the FR4. In a preferred embodiment, the therapeutic compound, imaging and / or a diagnostic compound is conjugated to the (single-domain) antibody by coupling to a thiol group in the (unpaired) cysteine. Said (unpaired) cysteine may be introduced using recombinant DNA technology, which the skilled person will be familiar with. Most preferably, the coupling to (an unpaired) cysteine is through a reaction with thiol-reactive groups, preferably a maleimide, comprised in the therapeutic compound and / or a diagnostic or imaging compound.

[0035] Preferably, the therapeutic / diagnostic / imaging compound according to the present disclosure has a size of not more than 500, 400, 300, 200, 100, 90, 80, 70, 60, 50, 40, 30, 20, 15, 10, 5 kDa and / or of at least 1, 2, 3, 4, 5 kDa. This can be measured by mass spectrometry or SDS-PAGE.

[0036] In an embodiment, a therapeutic compound according to the present disclosure may be used in the prevention and / or treatment of disease, preferably a brain disease, wherein the brain disease is preferably selected from a disease characterized by protein aggregation (e.g. amyloids and / or prions present in Alzheimer’s Disease, Parkinson’s Disease and Huntington’s Disease), brain tumor (e.g. glioma and metastasized tumor), metabolic diseases (e.g. Hunter syndrome, or Mucopolysaccharidosis type II (MPS II)), encephalitis / meningitis, brain ischemia, headache, cluster headache, migraine, or neurodevelopmental disorders, such as autism and epilepsy.

[0037] A disease characterized by undesired protein aggregation (amyloidosis) is a condition where misfolded proteins aggregate, impairing brain function, and promoting cell death, often leading to neurodegenerative diseases. Typical examples are Alzheimer’s disease, which is characterized by beta-amyloid plaque accumulation and tau tangles, leading to memory loss and cognitive decline; and prion disease, such as Creutzfeld-Jacob disease, which is characterized by the aggregation of prions resulting in death.

[0038] Parkinson’s disease is a neurodegenerative disorder caused by the loss of dopamine-producing neurons, leading to tremors, rigidity, and movement difficulties.

[0039] Huntington’s disease is a genetic neurodegenerative disorder characterized by involuntary movements, cognitive decline, and psychiatric symptoms due to mutated huntingtin protein. Brain tumors refer to an abnormal growth of cells in the brain, which can be benign or malignant, causing pressure on brain tissue and impairing function.

[0040] In metabolic diseases, such as Hunter Syndrome (or Mucopolysaccharidosis type II (MPS II) the body lacks certain enzymes to break down certain molecules, leading to tissue buildup, skeletal abnormalities, organ dysfunction, and developmental delays.

[0041] Encephalitis refers to inflammation of the brain, often caused by infection, resulting in headaches, fever, confusion, seizures, or neurological damage.

[0042] Brain ischemia refers to a condition, in which blood flow towards the brain is reduced, causing tissue damage due to a lack of oxygen and nutrients.

[0043] Headache is generally considered as pain in the head or neck area, which may be caused by various factors, including stress, tension, or underlying medical conditions.

[0044] Cluster headache refers to severe, recurrent headaches on one side of the head, often around the eye, occurring in cyclical patterns or clusters.

[0045] Migraine is a neurological condition characterized by intense headaches, often with nausea, sensitivity to light, and sometimes aura.

[0046] Neurodevelopmental disorder refers to a disorder affecting (or which has affected) brain development, leading to impaired social, cognitive, or motor functions, including conditions like autism and epilepsy.

[0047] Autism refers to a neurodevelopmental disorder characterized by difficulties in social interaction, and communication, and restricted, repetitive behaviors or interests.

[0048] Epilepsy is a neurological disorder marked by recurrent, unprovoked seizures due to abnormal electrical activity in the brain.

[0049] The binding polypeptide, preferably the VHH, of the present disclosure may have the following typical structure with complementarity determining regions (CDR) and framework regions (FR): (N-terminus) - (NTR) - FR1 - CDR1 - FR2 - CDR2 - FR3 - CDR3 - FR4 - C-terminal region (CTR) - (C-terminus), wherein FR1, FR2, FR3 and / or FR4 may be optional. Herein, the term is merely used to indicate the border of different CDR and FR regions and how these are positioned (i.e. are located) with respect to each other. The term N-terminal / C-terminal region is optional and may be used to comprise sequences used for recognition, conjugation and / or purification. The CTR (or NTR) may for example comprise a purification tag, or an epitope tag, or a protein ligation tag, or a cysteine residue as discussed elsewhere herein, preferably without affecting the binding properties of the VHH. The terms ‘(N-terminal)’ and ‘(C-terminal)’ are merely used to indicate the two ends of the (amino acid) sequence of the antibody, particularly the end with a free amino group (N-terminus) and the end with a free carboxyl group (C-terminus). The ‘N-terminal’ and ‘C-terminal’ are used to describe the structure of a typical antibody, preferably a typical VHH.

[0050] In a preferred embodiment, the binding polypeptide, e.g. single domain antibody or VHH of the present disclosure comprises:

[0051] - a CDR1 having an amino acid sequence with at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with any of SEQ ID NO: 16, 23, 30, 37, 44, 51, 58, 65, 72, 79, 86, 93, 100, and 107, preferably SEQ ID NO: 30 and / or SEQ ID NO:44;

[0052] - a CDR2 having an amino acid sequence with at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identify with any of SEQ ID NO: 18, 25, 32, 39, 46, 53, 60, 67, 74, 81, 88, 95, 102, and 109, preferably SEQ ID NO: 32 and / or SEQ ID NO:46;

[0053] - a CDR3 having an amino acid sequence with at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identify with any of SEQ ID NO:20, 27, 34, 41, 48, 55, 62, 69, 76, 83, 90, 97, 104, and 111, preferably SEQ ID NO: 34 and / or SEQ ID NO:48.

[0054] In addition or alternatively, the binding polypeptide, e.g. single domain antibody or VHH of the present disclosure preferably comprises:

[0055] - an FR1 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with any of SEQ ID NO: 15, 22, 29, 36,43, 50, 57, 64, 71, 78, 85, 92, 99, 106; - an FR2 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with any of SEQ ID NO: 17, 24, 31, 38, 45, 52, 59, 66, 73, 80, 87, 94, 101, 108; - an FR3 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with any of SEQ ID NO: 19, 26, 33, 40, 47, 54, 61, 68, 75, 82, 89, 96, 103, 110.

[0056] In a particularly preferred embodiment, the binding polypeptide, e.g. single domain antibody or VHH of the present disclosure preferably comprises:

[0057] - CDR1 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with SEQ ID NO: 44;

[0058] - CDR2 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with SEQ ID NO: 46;

[0059] - CDR3 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with SEQ ID NO: 48, preferably with an:

[0060] - FR1 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with SEQ ID NO: 43;

[0061] - FR2 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with SEQ ID NO: 45;

[0062] - FR3 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with SEQ ID NO:47.

[0063] The above characterization refers to single-domain antibody H7 as disclosed herein.

[0064] Even more preferably, the binding polypeptide, e.g. single-domain antibody or VHH of the present disclosure comprises a:

[0065] - CDR1 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with SEQ ID NO: 30;

[0066] - CDR2 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with SEQ ID NO: 32;

[0067] - CDR3 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with SEQ ID NO: 34, preferably with an:

[0068] - FR1 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with SEQ ID NO:29;

[0069] - FR2 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with SEQ ID NO:31;

[0070] - FR3 having an amino acid sequence with preferably at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with SEQ ID NO:33.

[0071] The above characterization refers to single domain antibody F12 as disclosed herein.

[0072] According to in vitro BBB models it was found that the sdAb F12 exhibits enhanced transcytosis efficiency compared to sdAb H7. However, sdAb H7 has shown higher affinity according to binding assay results. These two single-domain antibodies are most preferred in the present disclosure.

[0073] The binding polypeptide, e.g. single-domain antibody or VHH of the present disclosure preferably comprises an amino acid sequence with at least 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 64, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97, 98, 99 or 100% sequence identity with any of SEQ ID NO:1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, preferably SEQ ID NO: 3 and / or SEQ ID NO: 5, most preferably SEQ ID NO: 3. In a particularly preferred embodiment, the binding polypeptide, preferably an antibody, most preferably a single domain antibody (e.g. variable domain of the heavy chain of a heavy-chain-only (VHH) antibody) according to the invention is biparatopic. A biparatopic binding polypeptide, (single domain) antibody or VHH refers to a type of binding polypeptide that is engineered to bind to two different (non-overlapping) epitopes (binding sites) on the same target molecule, e.g. heparin-binding EGF-like growth factor (HB-EGF). A biparatopic VHH is typically engineered by linking two VHH domains, which specifically recognize two different epitopes on the same target antigen. This could be done via a peptide linker or other methods of fusion. By binding to two distinct sites on the same antigen, a biparatopic VHH can achieve a stronger and more specific binding and may elevate the internalization and transcytosis over the BBB as compared to a single-domain antibody. As a result, this dual binding can enhance transcytosis and consequently the therapeutic, imaging, or diagnostic efficacy of the cargo molecule.

[0074] In the present disclosure, it is preferred that the binding polypeptide, preferably an antibody, most preferably a single-domain antibody (e.g. variable domain of the heavy chain of a heavy chain only (VHH) antibody), that binds heparin-binding EGF-like growth factor (HB-EGF) according to the invention, is not or does not comprise the cross- reacting material 197 (CRM 197), or any non-toxic derivative of the Diphtheria toxin (DT).

[0075] In a preferred embodiment, at least one amino acid (e.g. amino acid other than histidine, such as arginine, and / or proline / tyrosine) in the single-domain antibody (or specifically in the VHH thereof) is substituted by histidine. The histidine amino acid is preferably located in the paratope and / or of CDR2 or CDR3. This strategy was found to increase the antibody-receptor dissociation in the acidified endosome and to improve the release of shuttle antibodies at the basolateral side (abluminal) of the endothelial cell into the brain.

[0076] As described earlier herein, the binding polypeptide, such as the single-domain antibody or VHH, as disclosed herein, may be combined with a therapeutic compound, an imaging and / or a diagnostic compound.

[0077] In a preferred embodiment, the therapeutic compound is one or more of

[0078] - a drug suitable for brain tumor treatment, such as a cytotoxic agent (such as Temozolomide, carboplatin); e.g. for treatment of neurodevelopmental disorders (such as Bumetanide) - RNA preferably siRNA, RNAi, miRNA, and preferably capable of modulation of protein synthesis (e.g. by brain cells);

[0079] - a growth factor, for example, epidermal growth factor (EGF) or insulin-like growth factor (IGF), and preferably suitable for modulation of neuronal cells; - a protein, preferably a binding peptide that binds one or more of the epidermal growth factor receptor variant III (EGFvIll), beta-amyloid, tau, alpha-synuclein, prion protein and / or wherein the binding peptide is chosen from a single-domain antibody, affibody, designed ankyrin repeat proteins (DARPins) or monoclonal antibody or fragment thereof; preferably to dissolve protein aggregates and / or to induce targeted protein degradation (TPD) via LYTAC (Lysosometargeting chimera);

[0080] - an enzyme, e.g. to replace deficient or malfunctioning extracellular enzymes, preferably chosen from iduronate-2-sulfatase (I2S) preferably to treat Hunter syndrome;

[0081] - a cytokine preferably chosen from interleukin-6, interleukin-10, tumor necrosis factor-alpha, transforming growth factor-beta;

[0082] - a neurotrophic factor, preferably chosen from nerve growth factor, brain-derived neurotrophic factor, glial cell line-derived neurotrophic factor, neurotrophin-3, ciliary neurotrophic factor, proteins such as cytokines and neurotrophic factors, to promote neuronal cell survival and repair; and / or

[0083] - a therapeutic compound comprised of (or in) a nanoparticle, preferably a liposome, micelle, nanoparticle aptamer (NANAP), or polymeric nanoparticle comprising a therapeutical compound.

[0084] As a therapeutic compound is also foreseen a binding peptide that binds a brain target, wherein the brain target is preferably selected from the group consisting of P-secretase 1, AP, epidermal growth factor, epidermal growth factor receptor 2, epidermal growth factor receptor variant III, tau, phosphorylated tau, apolipoprotein E4, alpha-synuclein, oligomeric fragments of alpha-synuclein, CD20, huntingtin, prion protein, leucine-rich repeat kinase 2, parkin, presenilin 2, gamma-secretase, death receptor 6, amyloid precursor protein, p75 neurotrophin receptor and caspase 6.

[0085] In addition or alternatively, the diagnostic compound according to the present disclosure may be one or more of

[0086] - peptide, or protein suitable for detection by binding of protein aggregates and / or tumor (optionally conjugated to any one of the following);

[0087] - (targeted) imaging agent, preferably chosen from gadolinium, ultra-small iron particles, or nuclides such as18F,68Ga,89Zr preferably bound to the above-mentioned peptide / protein either directly or via chelate rings as DOTA (1,4,7,10-Tetraazacyclododecane-1,4,7,10-tetraacetic acid), NOTA (1,4,7-Triazacyclononane-1,4,7-triacetic acid), HYNIC (Hydrazinonicotinic Acid). Use is foreseen with e.g. immuno-PET-CT or, immuno-SPECT, MRI and / or Scintigraphy.

[0088] In addition or alternatively, the compound for image-guided surgery may be one or more of peptide or protein suitable for detection of brain tumor conjugated to optical imaging agents, such as IRDye800.

[0089] The binding polypeptide or the single-domain antibody according to the present disclosure may also be used in combination with radiation therapy, chemotherapy and / or one or more immunomodulating agents, wherein the immune modulating agent is preferably an immune checkpoint inhibitor.

[0090] Optimal use of the therapeutic / diagnostic / imaging compound may be determined by those skilled in the art, and will vary depending on e.g., the strength of the preparation, the mode of administration, and the advancement of the disease condition. Additional factors relating to a particular subject being treated including subject age, weight, gender, diet, and time of administration, will result in a need to adjust the use. The therapeutic compound according to the present disclosure may be for use in combinational therapy.

[0091] What is also foreseen is to extend in vivo half-life of the binding peptide or single-domain antibody of the present disclosure, for example by fusing or coupling to a protein domain that extends in vivo half-life time of the binding peptide or single-domain antibody, preferably wherein the protein domain is chosen from albumin (binding) domain (including a VHH), and fragment crystallizable region (Fc-region) of conventional immunoglobulins.

[0092] In a preferred embodiment, the binding peptide or single-domain antibody for use according to the present disclosure may be comprised in a pharmaceutical composition. Accordingly, the present disclosure further relates to a pharmaceutical composition comprising the said. Herein a pharmaceutic composition refers to those compounds, materials, compositions and / or dosage forms which are, within the scope of sound medical judgment, suitable, e.g. non-toxic and / or (generally) do not lead to allergic response and / or irritation when in contact with tissues of human beings and animals. The pharmaceutical composition may also comprise any other therapeutically effective agents and / or other agents, such as a pharmaceutically acceptable solvent, diluent, carrier, buffer, excipient, adjuvant, carrier medium, antiseptic, filling, stabilizing or thickening agent.

[0093] The pharmaceutical composition according to the disclosure may be in any form suitable for administration, for example a solid, semisolid, or liquid form. A formulation can be selected from the list consisting of, but not limited to, powder, solutions, emulsions and suspensions. The pharmaceutical compositions may be produced by any conventional processes known in the art, which the skilled person will be aware of. Preferably, the pharmaceutical composition according to the disclosure is for administration by systemic route, for example by the intravenous route. More preferably, the composition comprising the binding peptide or single-domain antibody can be administered in separate doses that are spaced over time. Their administration routes, dosing schedules, and optimal galenic forms can be determined according to the criteria generally taken into account when establishing a treatment suited

[0094] Brief description of the Figures

[0095] Fig. 1: Isolation of proHB-EGF targeting sdAbs. a) Schematic structure of HB-EGF. b) Antibody titer determination of Llama sera collected at indicated days after immunization, c) Five indicated sdAbs clones were analyzed for binding to sHB-EGF or pre-proHB-EGF. d) ProHB-EGF targeting sdAbs (F12, H7) and negative control sdAb (R2) conjugated to Alexa-647 dye were analyzed by SDS-PAGE and tested for binding to sHB-EGF. Alexa-647 labeled CRM 197 was used as a positive control. The graph shows mean ±SD. e) COS-7 cells transfected with GFP-labeled full-length human HB-EGF were incubated with Alexa-647 labeled sdAbs or commercial anti-proHB-EGF antibody and imaged by confocal microscopy. Scale bars are 50 pm. Cell nuclei were stained with Hoechst dye.

[0096] Fig. 2: Effects of sdAbs on viability and integrity of hBMECs. a) Cell viability was determined after the hBMECs were incubated with 1000 nM sdAbs for 3 hrs or 48 hrs. b) Apparent permeability (Papp) for 3 sdAbs was determined using 20 kDa FITC-dextran after the hBMECs monolayer growing in the transwell was incubated with 1000 nM sdAbs or 500 nM CRM 197. All graphs show mean ± SD. Statistical analysis was performed using one-way ANOVA followed by Dunnett’s multiple comparisons tests; ns indicates no significant differences; *P < 0.05.

[0097] Fig. 3: Binding of proHB-EGF targeting sdAbs on hBMECs. a) WB quantification of proHB-EGF expression in primary hBMECs and hCMEC / D3 cell lines, b) Well plate cultured hBMECs were incubated with 1000 nM of one Alexa-647 conjugated sdAb or commercial anti-proHB-EGF antibody and imaged by confocal microscopy, c) Transwell-cultured hBMECs were incubated with the commercial anti-proHB-EGF antibody on the apical side and imaged by confocal microscopy. The fluorescence image was presented as an orthogonal view, d) The binding of Alexa-647 conjugated sdAbs was tested on the transwell cultured hBMECs monolayer. All scale bars are 50 pm. All cell nuclei were stained with Hoechst dye. All graphs show mean ± SD. Statistical analysis was performed using student’s t-test for panel a and one-way ANOVA followed by Dunnett’s multiple comparisons tests for panel d; ns indicates no significant differences; *P < 0.05, **P < 0.01. Fig. 4: Analysis of proHB-EGF expression in hBMECs by Western Blot (WB) quantification, a) WB quantification of proHB-EGF expression in hBMECs cultured in three indicated commercial endothelial media, b) WB quantification of proHB-EGF expression in hBMECs affected by serum level, serum starvation, and cAMP / forskolin stimulation, c) WB quantification of proHB-EGF expression in hBMECs affected by the co-culture of astrocytes or U87 cells. All graphs show mean ± SD. All graphs show mean ± SD. Statistical analysis was performed using oneway ANOVA. For multiple comparisons, T ukey’s multiple comparisons tests were performed for panels a and c; Dunnett’s multiple comparisons tests were performed for panel b; ns indicates no significant differences; *P < 0.05, **P < 0.01, ***P < 0.001.

[0098] Fig. 5: sdAb transcytosis in co-cultured hBMECs BBB model, a) Immune Fluorescence Microscopy (IFM) of the co-cultured BBB model consisting of a monolayer of the hBMECs growing in the insert of the transwell and astrocytes growing on the back of the insert, stained for VE-cadherin and ZO-1, s100p and GFAP, respectively, b) Analysis of sdAb (1000 nM) transcytosis in the co-cultured BBB model. Medium on the apical side was additionally supplied with 10000 nM 20 kDa dextran as a barrier leakage indicator, c) IFM shows co-localization of Alexa647 conjugated sdAbs and proHB-EGF during transcytosis. White arrow lines indicate the area for co-localization analysis, d) 3D IFM showing the proHB-EGF-sdAb complex during transcytosis. All cell nuclei were stained with Hoechst dye. All scale bars are 20 pm. The graph shows mean ±SD. Statistical analysis was performed using one-way ANOVA followed by Tukey’s multiple comparisons tests; ns indicates no significant differences; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

[0099] Fig. 6: sdAb transcytosis in hiPSC-derived iBBB model, a) IFM of the iBBB model consisting of a monolayer of the hiPSC-derived BMEC-like cells in the transwell, stained for VE-cadherin, ZO-1, Claudin-5, Occludin, GLUT-1, P-gp and proHB-EGF. Cell nuclei were stained with Hoechst dye. Scale bars are 50 pm. b) WB quantification of proHB-EGF expression in the hiPSC-derived BMEC-like cells. hBMECs were used as a positive control, c) Analysis of sdAb (1000 nM) transcytosis in the iBBB model. Medium on the apical side was additionally supplied with 10000 nM 20 kDa dextran as a barrier leakage indicator. All graphs show mean ±SD. Statistical analysis was performed using student’s t-test for panel b and one-way ANOVA followed by Tukey’s multiple comparisons tests for panel c; ns indicates no significant differences; *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

[0100] Fig. 7: Detection of released sdAbs at basolateral (abluminal) side; comparison of transcytosis efficiency of sdAbs directed to different targets. 14 VHH sequences: Framework region CDRl Framework region 2 CDR2 Framework region 3 CBR3 Framwork region 4 F7 ( SEQTDNO:!) EVQLVESGGGLVQDGGSLRLSGAASG WYRQTPGNQRELVATI QTGGSADYVMS VKGRFTI SRDNAKNTVYLHMNRLKPEDTAVYYC NGYDFGRPMWP WGQGTQVTVSS

[0101]

[0102] ( SEQIDNO: 15) (SEQIDNO: 16) (SEQIDNO: 17) (SEQIDNO: 18 ) (SEQIDNO: 19) (SEQIDNO: 20 ) (SEQIDNQ: 21) A10 ( SEQIDNO: 2) EVQLVESGGGLVQDGGSLRLSGAASG WYRQTPGNQRELVATI QTGGSADYVMS VKGRFTISRDNAKNTVYLHMNRLKPEDTAVYYC NGYDFGRPMWP WGQGTQVTVSS (SEQIDNO: 22) (SEQIDNO: 23) (SEQIDNO: 24 ) (SEQIDNQ: 25) (SEQIDNO: 26) (SEQIDNO: 27) (SEQIDNO: 28 )

[0103]

[0104] Fl 2 SEQIDNO: 3) EVQLVESGGGLVQPGGSLRLSCAASG WYRQTPGNQRELVATI OTGGSADYVNS VKGRFTI SRDNAKNTVYLHMNRLKPEDTAVYYC1G¥DFGRpMTOp

[0105] WGQGTQVTVSS

[0106] (SEQIDNO: 29) (SEQIDNO: 30 ) (SEQIDNO: 31) (SEQIDNO: 32) (SEQIDNO: 33) (SEQIDNO: 34 ) (SEQIDNO: 35) CIO (SEQIDNO:4 ) EVQLVESGGGLVQDGGSLRLSCAASG WYRQTPGNQRELVATI OTGGSADYVNSHGYDF6R PMJWp WGQGTQVTVSS ( SEQIDNO: 36) (SEQIDNO: 37 ) (SEQIDNO: 38 ) (SEQIDNO: 39) (SEQIDNO: 40 ) (SEQIDNO: 1 ) (SEQIDNO: 42 ) H7 ( SEQIDNO i S) EVQLVESGGGLVQDGGSLRESCAASG WYPQTFGNQRELVATI OTGGSADYVSS NGYDFGRPMWP WGRGTQVTVSS (SEQIDNO: 43) (SEQIDNO: 44) (SEQIDNO: 45) (SEQIDNO: 46) (SEQIDNO: 47 ) (SEQIDNO: 48) (SEQIDNO: 49) B4 ( SEQIDNO: 6) EVQLVESGGGLVQPGGSLRLSCAASG W 1 RQAPGKERELVAS I TSGGDTNYADS ISAE'TKTGSRW 1 STB lEDH WGQGTQVTVSS (SEQIDNO: 50) (SEQIDNO: 51) (SEQIDNO: 52) (SEQIDNO: 53) (SEQIDNO: 54 ) (SEQIDNO: 55 ) (SEQIDNO: 56) All (SEQIDNO: 7) EVQLVESGGGLVQPGGSLRLSCAASG FTFSNYAMS WRQAPGKGLEWVSAI GETLTTYADS VKGRFT I SRDNAKNLLYLQMNSLKPEDTAVYYC ARYRDNYL GRGTQVTVSS ( SEQIDNO: 57) (SEQIDNO: 58 ) (SEQIDNO: 59) (SEQIDNO: 60) (SEQIDNO: 61 ) (SEQIDNO: 62 ) (SEQIDNO: 63) Bl ( SEQIDNO: 8) EVQLVESGGGLVQPGGSLRLSCAASG WVRQAPGKGPEWVSTI SISGGSRTTYADS VKGRFT I SRDNAKNKLYLQMNSLKPEDTAVYYC MQAQFPVIERPP GQGTQVTVSS ( SEQIDNO: 64 ) (SEQIDNO: 65) (SEQIDNO: 66) (SEQIDNO: 67) (SEQIDNO: 68) (SEQIDNO: 69) (SEQIDNO: 70) Fll (SEQIDNO: 9) EVQLVESGGGLVQAGGSLRLSCAASG WFRQAPGKEREF ASI SRSDGYPDYffllS VKGRFT I SRDNAKNTVNLQMNSLKPEDTAVYYC WGQGTQVTVSS ( SEQIDNO: 71) (SEQIDNO: 72 ) (SEQIDNO: 73) (SEQIDNO: 74 ) (SEQIDNO: 75) (SEQIDNO: 76) (SEQIDNO: 77 ) A9 ( SEQ I DNO: 10 ) EVQLVESGGGLVQAGGSLRLS GA WYRQPS GKERE L AT I SRDGSAVFVDS VKDRFTSSRDNTKNTAYLQMNSLKPEDTAVYYC WDVLGRTY WGQGTQVTVSS ( 5EQTDNO: 78 ) (SEQIDNO: 79) ( SEQIDNO: 80) (SEQIDNO: 81) (SEQIDNO: 82) (SEQIDNO: 83 ) ( SEQIDNO: 84 ) A3 ( SEQ I DNO: 11 ) EVQLVESGGDLVQAGGTLRLS CAASG WYRQSDEKQRELVAT HRS GATE YGDS VKGRFT I SRDNAKNMVYLQMNSLKPEDTAVYYC NVHVLIIRNY WGQGTQVTVSS ( SEQIDNO: 85) (SEQIDNO: 86) (SEQIDNO: 87 ) (SEQIDNO: 88 ) (SEQIDNO: 8 9) (SEQIDNO: 90 ) (SEQIDNO: 91) Cl ( SEQIDNO: I2) EVQLVESGGDLVQAGGTLRLSCAASN > WYF. QPPEKQRELVA5V HRS GATE YGDS VKGRFT I SRDKAKNTVYLQMNNLKPEDTAVYYC NVHVLDRNY WGQGTQVTVSS ( SEQIDNO: 92) (SEQIDNO: 93 ) (SEQIDNO: 94) (SEQIDNO: 95) (SEQIDNO: 96) (SEQIDNO: 97 ) ( SEQIDNO: 98 ) DI 2 SEQ I DNO: 13 ) EVQLVESGGDLVQAGGTLRLS CAASR W i F QPPEKPREL ATV HRS GATE YGDS VKGRFT I SRDNAENMVYLQMNSLKPEDTAVYYC SvNVT, HRHY WGQGTQVTVSS ( SEQIDNO: 99) SEQIDNO: 100) (SEQIDNO: 101 ) (SEQIDNOl 102 ) ( SEQIDNO: 103) (SEQIDNO: 104 ) (SEQIDNO: 105 ) G10 ( SEQIDNO: 14 ) EVQLVESGGDLVQAGGTLRLS CAASG WYRQPPNKQRELVGTV HRS GATE YGDS VKGRFT I SRDIAKNMVYLQMDSLKPEDTAVYYC IIIWLHRNY WGQGTQVTVSS ( SEQIDNO: 106) (SEQIDNO: T 07 ) (SEQIDNO: 108 ) (SEQIDNO: 109 ) ( SEQIDNO: 110) (SEQIDNO: 111 ) (SEQIDNO: 112) In case of differences between the above-listed sequences and the sequences in the accompanying ST26 listing, the above sequences are preferred.

[0107] EXPERIMENTAL SECTION

[0108] Example 1

[0109] Here the inventors demonstrate the ability of proHB-EGF targeted single-domain antibodies to pass the brain endothelial cell barrier via RMT. The specific transport of these sdAbs was confirmed by both the primary human cell-based BBB model and the hiPSC-derived iBBB model using trans-well systems, and a neurovascular unit grown in an organopiate microfluidic culture platform. Evidence for RMT was demonstrated by the transport of anti-proHB-EGF-sdAbs complexes across the polarized endothelial cell layer and the detection of released sdAbs at the basolateral (abluminal) side. These proHB-EGF targeted sdAbs are aimed to be used as BBB shuttle tools that facilitate (targeted) brain therapy / imaging and / or diagnosis.

[0110] Methods

[0111] Cell culture

[0112] Primary human brain microvascular endothelial cells (hBMECs) (ACBRI 376) were purchased from Cell Systems. Cells were maintained in culture flasks coated with 50 pg / ml collagen-IV (C5533, Sigma) and 25 pg / ml fibronectin (F4759, Sigma) in MV-2 medium (C-2212, Promocell) supplemented with 1* Antibiotic Antimycotic reagent (A5955, Sigma) and in a 5% CO2 humidified incubator at 37°C. The medium was refreshed every two days. Upon reaching confluency around 70%, cells were detached with 0.025% trypsin EDTA (#25300054, Thermo) and subcultured at a ratio between 1:3-1:6 in newly coated flasks. Cells were used for experiments with no more than 9 passages. The primary human astrocytes (#1800, ScienCell) were purchased from ScienCell. Cells were maintained in culture flasks coated with hESC-qualified Matrigel (#354277, Corning) in a 5% CO2 humidified incubator at 37°C. AGM medium (#1801, ScienCell) supplemented with 1* Antibiotic Antimycotic reagent was used, which was refreshed every two days. Upon reaching 90% confluency, the astrocytes were detached with 0.025% trypsin EDTA and subcultured at a density of 5000 cells / cm2in new coated flasks. The astrocytes were used for experiments with no more than 10 passages. The hiPSCs were generated in the lab from the IMR90 fibroblast cell line by episomal plasmid transfection method, as previously reported by others

[0026] , HiPSCs were feeder-freely maintained in StemFlex medium (A3349401, Thermo) in 6-well plates coated with hESC-qualified Matrigel in a 5% CO2 humidified incubator at 37°C. The medium was refreshed daily. Upon reaching 70% confluency, cells were washed once with PBS and detached with ReleSR (#05872, STEMCELL) and subcultured as colonies at a ratio between 1:10-1:20 in a new coated well plate. The hiPSCs were used for experiments with no more than 50 passages. The COS-7 cell line (CRL-1651, ATCC) was maintained in culture flasks in low glucose DMEM (D6046, Sigma) containing 10% FBS (v / v) and 1* Antibiotic Antimycotic reagent. The U87 cell line (HTB-14, ATCC) was maintained in culture flasks in low glucose DMEM containing 10% FBS and 1* Antibiotic Antimycotic reagent. The b. End.3 cell line (CRL-2299, ATCC) was maintained in culture flasks in high glucose DMEM (D6429, Sigma) containing 10% FBS and 1* Antibiotic Antimycotic reagent. All cell lines were regularly tested for mycoplasma contamination and were all found to be continuously negative.

[0113] HB-EGF transfection of COS-7 cells

[0114] COS-7 cells were seeded in wells of a 6-well plate. Upon reaching 85% confluency, cells were transiently transfected with a plasmid expressing the full-length gene of human HB-EGF carrying a C-terminal GFP-tag (HG10325-ACG, SinoBio) using Lipofectamine 3000 reagent (L3000001, Thermo). Two days after transfection, when cells showed the most intense membranous GFP signal, they were detached with trypsin EDTA and seeded in well plates for experiments.

[0115] Llama immunization

[0116] Immunizations of two llamas were performed at Kaneka Eurogentec S. A (Liege, Belgium). Primarily, both animals received 4 injections of a purified protein mixture consisting of 15 pg human pre-proHB-EGF (ectodomain, 20-148aa) (#215635, Abeam) and 15 pg sHB-EGF (64-148aa) (#269156, Abeam) (Fig. 1a) at an interval of 2 weeks. For the final booster injection, 2 weeks after the final injection with the purified recombinant proteins, the two llamas were injected with isolated membrane vesicles prepared from 2x107COS-7 cells transiently transfected to express the full-length human HB-EGF, following a published method

[0027] ,

[0117] Immune response evaluation

[0118] To validate the immune response of the llamas towards HB-EGF, sera titers were evaluated via enzyme-linked immunosorbent assay (ELISA)-based protein binding assay. MaxiSorp 96-well plates (#44240421, Thermo) were overnight coated with 50 ng / well of either pre-proHB-EGF (20-148 aa) or sHB-EGF (64-148 aa) in 50 mM carbonate buffer (pH 9.4) at 4°C. The following day, the coating solution was removed, and the wells were blocked with 2% bovine serum albumin (BSA) (A2153, Sigma) in PBS for 2 hrs at RT. Serially diluted sera (1-106times) were added to the coated wells and incubated for 1.5 hrs at RT, followed by 3 times washing in PBS containing 1% BSA and 1 time washing with PBS. The wells were further incubated with rabbit anti-VHH antibody (QE19, QVQ) for 2 hrs at RT, followed by the same washing procedure mentioned above. The wells were then incubated with donkey anti-rabbit IRDye® 680RD antibody (#926-68073, LI-COR, 1:1000) for 1 hr at RT in the dark, followed by the same washing procedure as mentioned above. The wells were finally supplied with fresh PBS and the signal indicating binding was detected using the Odyssey® infrared imaging system at 700 nm. Data was analyzed using GraphPad Prism 9 software.

[0119] Library construction and phage selection

[0120] The total mRNA isolated from the peripheral blood lymphocytes (PBLs) of two llamas was transcribed into cDNA, and specific primers were used to amplify the VHH regions, which were ligated into the phagemid vector pUR8100-Myc-His, as described previously

[0028] , Transformation of electrocompetent TG1 cells resulted in the generation of two libraries named “SNL166 & SNL167”, which had a size of 2.5 x 106and 1.15 x 107transformants, respectively. The generated phage libraries were applied for phage selection on pre-proHB-EGF (20-148aa) and sHB-EGF (64-148aa). The SNL 166 phage library and the SNL 167 phage library were applied separately in each two selection rounds on either the pre-proHB-EGF or sHB-EGF ectodomain. The output phages (4x test groups) from both libraries were pooled together to be used for the final selection round on the murine b. End.3 cells, given the need to perform possible preclinical evaluation. After the final panning round, the resulting TG1 cells were plated on agar plates supplemented with 100 pg / ml kanamycin and grown overnight at 37°C. The screening of the monoclonal phages for their binding to pre-proHB-EGF and / or sHB-EGF ectodomain was evaluated by phage ELISA. Single colonies were picked and grown in 2x YT broth supplemented with kanamycin and 2% glucose in the U-bottom 96-well plates (#351177, Thermo) at 37°C. When OD (optical density) 600 of the cultures reached around 0.5 (equals around 4x 108cells / ml), the cultures were infected with an equal amount of helper phages. Cultures were grown for 30 min without shaking followed by overnight culture with shaking at 37°C. Five hundred nanograms of pre-proHB-EGF (20-148 aa) or sHB-EGF (64-148 aa) were coated per well of the MaxiSorp 96-well plates overnight at 4°C. The next day, wells were washed three times with PBS and blocked with 2% (w / v) skimmed milk in PBS for 1 hr at RT. Phage cultures in the U-bottom plates were centrifuged at 4700 rpm for 15 min at 4°C, and the supernatant was added to the assay plates in a blocking buffer. The assay plates were incubated for 2 hrs at RT on a shaker followed by five times washing with PBS. Mouse-anti-M13-HRP (#27-9421-01, G E-Healthcare) was diluted 1:10000 in blocking buffer and incubated in the assay plates for 1 hr at RT while shaking. The plates were washed again five times with PBS. TMB solution (N301, Thermo) was added to the wells, and the reaction was stopped with 2M H2SO4. Absorbance was measured at 450 nm with the FLUOstar plate reader (BMG TABTECH).

[0121] sdAb production and purification

[0122] Phagemids encoding the sdAbs were extracted from the TG1 cells. Sequences of the selected sdAb clones were ordered at Twist Bioscience Corporation, with codons optimized for expression in E. coli. In addition, the open reading frame included an N-terminal pelB signal sequence to facilitate the periplasmic secretion and a C-terminal cysteine followed by a 6* His-tag. The synthesized DNAs were cloned into pET-21 (+) expression vector followed by heatshock transfection into competent E. coli BL-21 Codon Plus (DE3)-RIL cells (Agilent Technologies). Single colonies of transformed bacteria were picked and pre-cultured overnight in LB medium supplemented with 2% glucose, 100 pg / ml ampicillin, and 35 pg / ml chloramphenicol. The culture was then upscaled by diluting 1:100 in TB medium supplemented with 0.1% glucose and ampicillin. sdAb expression was induced upon culture reaching log phase by adding 1 mM isopropyl p-D-1-thihogalactopyranoside (IPTG) (I6758, Sigma). Following overnight induction at 25°C and shaking at 400 rpm, the bacteria were harvested by centrifuging at 4800x g for 20 min at 4°C. The resulting pellet was resuspended in PBS, and was subjected to 3 freeze-thaw cycles at -20°C. The periplasmic fraction was separated by centrifugation at 10000* g for 20 min at 4°C, and the supernatant containing soluble His-tagged sdAbs was filtered through 0.45 pm filters before loading to Ni-NTA beads (#30230, Qiagen) for purification. The sdAbs were eluted, dialyzed in PBS overnight, and stored at -20°C. The R2 sdAb, which was raised against an irrelevant target, the copper-containing azo-dye Reactive Red 6 [29, 30], was in-house produced and used as a negative control.

[0123] For basic characterization, the sdAbs were separated by SDS-PAGE under reducing conditions, and the gel was stained by Coomassie blue. Binding affinities of sdAbs to pre-proH B-EGF or sHB-EGF were evaluated by protein binding assay, as described later. A non-linear regression curve was fitted for “one site-specific binding” using GraphPad Prism 9 software. Binding affinities of the sdAbs represented as KD (ligand concentration that occupies half of the maximum receptor binding sites at equilibrium) were automatically determined.

[0124] Fluorophore conjugation of sdAbs

[0125] Three sdAbs (F12, H7, R2) were conjugated to Alexa-647 fluorophore via the single unpaired cysteine added to the C-terminal of the sdAb, as previously described

[0031] , Before conjugation, the cysteine on the sdAb was first reduced with 2 M equivalents of Tris (2-Carboxyethyl) phosphine hydrochloride (TCEP) (C4706, Sigma) in borate buffer (250 mM sodium borate, 250 mM NaCI and 1 mM diethylenetriaminepentaacetic acid, pH: 8) at 37°C for 2 hrs. A protein labeling kit (A20347, Thermo) was used for the site-specific conjugation of maleimide-Alexa-647 to the sdAbs. Free dye was removed by 7K MWCO Zeba Spin desalting columns (#89882, Thermo). Degree of conjugation (DOC) was calculated according to the manufacturer’s instructions. The percentage of free dye in the sdAb preparation was determined by SDS-PAGE followed by gel imaging with the ChemiDoc MP imaging system (BioRad) with a 647 nm laser line. Band intensities were measured using Imaged software. In addition, CRM197 (D2189, Sigma), which was used as a control ligand for proHB-EGF, was randomly labeled with NHS-Alexa-647 on its lysine using a protein labeling kit (A20173, Thermo) and purified in the same way.

[0126] sdAb cell binding assay

[0127] COS-7 cells transfected with full-length human HB-EGF were seeded in a 96-well plate with a density of 40000 cells / cm2and cultured overnight. Alexa-647 labeled sdAbs (F12, H7, R2) were prepared at 1000 nM in cold assay medium (basal DMEM + 0.5% FBS + 25 mM HEPES (#15630080, Thermo)). In parallel, commercial proHB-EGF antibody (#66792, Abeam) was prepared at 1:200 in cold assay medium, as positive control. Cell medium was replaced with sdAb medium or commercial proHB-EGF antibody medium and the plate was placed on ice for 1 hr in the dark to allow for equilibrium sdAb binding. Cells were then washed twice with cold assay medium to remove unbound sdAbs. Additionally, for cells incubated with the commercial antibody, a fluorescently labeled secondary antibody was needed for detection (refer to the ‘Immunofluorescence microscopy method’ section). The cells were fixed in 4% paraformaldehyde (PFA) (P6148, Sigma) for 10 min and permeabilization in 0.1% Triton X100 in PBS for 10 min. Nuclei were stained with Hoechst (#62249, Thermo Fisher, 1:1000) in PBS for 10 min in the dark. Cells were finally washed three times in PBS and images were acquired by confocal microscopy (Yokogawa).

[0128] sdAbs binding to hBMECs was checked on both adherent culture in well plate and on monocultured BBB model in transwell insert. HBMECs at 40 000 cells / cm2seeding density were grown in collagen-IV and fibronectin-coated 96-well plate in MV-2 medium until confluency. Cells were cultured for 8 days before sdAb binding assay. For well plate cultured hBMECs, the method to access sdAb binding was identical as described above. For sdAb binding in the transwell BBB model, 500 pl sdAb medium was given in the inserts, whereas 1200 pl assay medium was added in the outer wells. After binding, the membrane of the inserts was cut off and placed in a well plate before fixation. For confocal imaging, the stained insert was inverted and mounted in the well plate. For quantification, 16 random fields per well were taken at 20* magnification and the images were analyzed using Imaged software. The acquired sdAb signal was then corrected with Alexa-647 labeling efficiency. Cell binding data was presented as sdAb signal normalized to the number of cell nuclei.

[0129] In-vitro BBB transwell model

[0130] Polycarbonate Cell Culture Inserts (#140652, Thermo) with a pore size of 0.4 pm and culture area of 1.13 cm2were used. For the co-culture BBB model, the basal side of the insert was first coated with 300 pl Matrigel overnight in the incubator with the insert inverted. Primary human astrocytes were harvested, pelleted, resuspended in 300 pl AGM medium, and seeded on the basal side of the coated insert at 150000 cells / cm2. The insert was inverted in the incubator for another night to allow astrocyte sedimentation. The next day, the endothelial cells were seeded. First, the apical side of the insert was coated with 200 pl collagen-IV (50 pg / ml) and fibronectin (25 pg / ml) (diluted in PBS) for 4 hrs in the incubator with the basal side of the insert submerged in AGM medium. HBMECs were harvested, pelleted, resuspended in 500 pl MV-2 medium, and seeded on the apical side of the coated insert at 220 000 cells / cm2. Meanwhile, AGM medium from the basal side of the insert was replaced by 1200 pl mixture of 50% AGM and 50% MV-2 medium. Cells on the insert were cultured until the desired date for the experiment wherein the medium was refreshed every two days.

[0131] For the mono-culture BBB model, the abovementioned astrocyte seeding part was omitted. For the generation of the hiPSC-derived iBBB transwell model, details can be found in the hiPSCs differentiation method section.

[0132] Barrier integrity assay (BIA)

[0133] Dextran-FITC (FD20S, 20 kDa, Sigma) was prepared in basal MV-2 medium at 25 000 nM (equals 0.5 mg / ml). The transwell BBB inserts (prepared as triplicates) were first washed with PBS (with Ca2+ / Mg2+) to remove non-adherend cells and residual medium. For BIA, 500 pl dextran medium was added in the inserts (apical side) whereas 1200 pl basal medium was in the outer wells (basal side). The transwell plate was placed in the incubator for 1 hr, where at different time points (20, 40, 60 min.) 200 pl medium sample from the basal side was taken followed by the addition of 200 pl fresh medium and gently shaking the plate. For quantitation, 200 pl medium sample was transferred to the assay plate. The dextran-FITC signal was read by the fluorescence plate reader (FP-8350, Jasco) with an excitation / emission setting of 490 nm / 525 nm. Serially diluted dextran samples were made to generate the standard curve. The apparent permeability (Papp) coefficient was calculated using the equation:

[0134]

[0135] in which dQ / dt is the transport rate (nmol / ml x min), CO is the initial concentration of dextran in the insert (nmol / ml), and A is the surface area of the insert (cm2). Lower Pappindicated better BBB integrity.

[0136] For BBB optimization in terms of barrier integrity, the BBB inserts were tested with different endothelial media including MV (C-22120, Promocell), MV-2 (C-22121, Promocell), EndoGRO (SCME004, Merck), and with different FBS levels in the medium. Co-culture with human astrocytes was tested in comparison to mono-cultured BBB insert. For barrier enhancement, the inserts were treated with serum starvation medium (MV-2 basal medium + 0.5% FBS) with or without db-cAMP (1 mM) / forskolin (10 pM) for 2 hrs in the incubator. BIA was performed to check the effects of the abovementioned actions on BBB integrity.

[0137] Cell viability and BBB disruption assay

[0138] HBMECs were exposed to the sdAbs and cell viability was determined. Cells were grown in collagen-IV and fibronectin-coated 96-well plate in MV-2 medium until confluency. sdAbs (F12, H7, R2) prepared at 1000 nM in 100 pl MV-2 medium were given to the cells and incubated in the incubator for either 3 hrs to check acute cytotoxicity or 48 hrs to check long-term cytotoxicity. After sdAb incubation, cells were washed three times with cell medium and treated with 100 pl Alamar Blue reagent Resazurin (R7017, Sigma) prepared at 50 pM in MV-2 medium for 3-4 hrs in the incubator. After Resazurin incubation, the supernatant was collected, centrifuged to remove floating cells, and transferred to the assay plate. The fluorescence signal was measured with the fluorescence plate reader at the setting of excitation / emission: 530 nm / 590 nm.

[0139] Barrier disruption by the sdAbs was checked on the mono-cultured transwell BBB model. Confluent hBMECs in the insert were incubated with 1000 nM sdAbs or 500 nM CRM197 on the apical side for 3 hrs in the incubator. The insert was washed once with cell medium and followed by the BIA.

[0140] hiPSCs differentiation into BMEC-like cells

[0141] Differentiation of hiPSCs into BMEC-like cells followed a published protocol but with minor modifications

[0032] , Briefly, hiPSCs were washed once with PBS and incubated with Accutase (A6964, Sigma) for 10 min in the cell incubator at 37°C followed by pipetting for less than 10 times to generate homogeneous single cells. Cells were pelleted and resuspended in StemFlex supplemented with 10 pM Y27632 (#72302, STEMCELL) and then seeded in Matrigel precoated 6-well plate at the density of 25,000 cells / cm2. For the next 3-4 days, Stemflex medium without Y27632 was refreshed daily to allow cells to expand. When cell confluency reached around 95%, differentiation was initiated by switching to DeMedium 1 [DMEM / F12 (#11320033, Thermo) + 1x NEAA (#11140050, Thermo) + 0.5 x GlutaMAX (#35050061, Thermo) + 0.1 mM beta-mercaptoethanol (2-ME) (#21985023, Thermo) + 6 pM CHIR99021 (SML1064, Sigma)] for 24 hrs and this day was marked as day 0-1 of differentiation. For day 1-6, DeMedium 2 [DMEM / F12 +1 x NEAA + 0.5 x GlutaMAX + 0.1mM 2-ME + 1 x Va+B27 (#17504044, Thermo)] was used with daily medium change. On day 6, DeMedium 3 [human endothelial serum-free medium (ESFM) (#11111044, Thermo) + 20 ng / ml FGF-2 (#100-18b, Peprotech) + 10 pM all-trans retinoic acid (#554720, Sigma) + 1 x Va+B27] was used and medium remained unchanged for 48 hrs. On day 8, single cells were obtained using Accutase (~20 min) and mechanical pipetting. Cells were pelleted and resuspended in DeMedium 3 and pre-selected at the ratio of 1:1 in a 6-well plate pre-coated with 50 pg / ml collagen-IV and 25 pg / ml fibronectin (co-diluted in PBS) for 1 hr in the incubator. One hour later, floating cells were removed and the adherent cells were lifted again, pelleted and resuspended in DeMedium 3. To generate the transwell iBBB model, cells were seeded onto Matrigel pre-coated insert (1.13 cm2area, 0.4 pm pore size) at a density of 1x106cells / cm2. For IFM and QPCR, cells were seeded into a Matrigel precoated well plate at a density of 800000 cells / cm2. The medium was switched to DeMedium 4 (hESFM + 1 x Va+B27 + 100 pM db-cAMP (D0627, Sigma)) 24 hrs later. From now on, DeMedium 4 was refreshed every two days. On day 11, cells in a well plate were used for either immunofluorescence microscopy (IFM) or QPCR. On day 16, which was 7 days after re-plating onto the transwell insert, the iBBB model was used for sdAb transcytosis assay.

[0142] sdAb transcytosis assay

[0143] sdAb transcytosis assay was done in the hBM EC-based co-culture transwell BBB model on day 8. Before the assay, the BBB inserts (prepared as triplicates) were treated with serum starvation medium plus 10 pM forskolin for 2 hrs to perform barrier enhancement. Three Alexa-647 labeled sdAbs (F12, H7, R2) were prepared at the concentration of 1000 nM in serum starvation medium. Meanwhile, 20 kDa FITC-dextran at 10 pM was included in the assay medium as an internal control to monitor BBB leakage. For sdAb transcytosis assay, 500 pl sdAb solutions were added into the inserts (apical side) and 1200 pl assay medium was added into the outer wells (basal side). The sdAb transcytosis assay took 3 hrs, where at different time points (60, 120, 180 min) a portion of 400 pl medium sample was collected from the basal side followed by replenishing 400 pl fresh medium and gently shaking the plate. All samples, if not immediately analyzed, were frozen at -20°C. To detect sdAb transcytosis, medium samples were heated at 70°Cfor 10 min with 4x SDS-PAGE loading buffer (#1610747, BioRad) followed by loading onto 4-12% Bis-Tris gel (NW04120BQX, Thermo Scientific) and run in MES buffer (NP0002, Thermo Scientific). In parallel, serial diluted samples of each sdAb were made and run in SDS-PAGE to generate standard curves. The polyacrylamide gel was imaged by ChemiDoc MP imaging system (BioRad) with a 647nm laser line. Band intensities of the sdAbs were measured using Imaged software followed by correction with Alexa-647 labeling efficiency. To monitor simultaneous BBB leakage during sdAb transcytosis, 200 l spare medium sample was loaded to the assay plate and the dextran-FITC signal was measured by a Jasco fluorescence plate reader, as described before.

[0144] Specifically, for sdAb transcytosis on hiPSCs derived iBBB model, hESFM medium was used as assay medium, and no barrier enhancement was performed beforehand.

[0145] Immunofluorescence microscopy (IFM)

[0146] All IFM steps were carried out at RT unless specified otherwise. A confluent monolayer of either hBMECs or hiPSC-derived BMEC-like cells in the well plate was washed twice in PBS and fixed either in 4% PFA for 10 min or in methanol on ice for 20 min. Cells were then washed three times in PBS and permeabilized in PBS containing 0.1% Triton X-100 (T8787, Sigma) for 10 min. Cells fixed with methanol did not require a permeabilization step, nonetheless. Specifically, no permeabilization was required if the BBB surface marker was to be stained. After permeabilization, cells were blocked in PBS containing 2% BSA for a minimum of 1 hr. Cells were incubated with primary antibodies for VE-cadherin (vascular endothelial cadherin, #33168, Abeam, 1:400), Claudin-5 (#4C3C2, Thermo, 1:50), ZO-1 (Zonula Occludens Protein 1, #61-7300, Thermo, 1:200), Occludin (#33-1500, Thermo, 1:50), GLUT-1 (glucose transporter- 1, SPM498, Thermo, 1:100), P-gp (p-glycoprotein, #235954, Abeam, 1:100), proHB-EGF (membrane-anchored heparin-binding EGF-like growth factor, #66792, Abeam, 1:200), Caveolin-1 (#2910, Abeam, 1:400), Clathrin (MA1-065, Thermo, 1:300) in blocking buffer overnight at 4 °C. For IFM of astrocytes in the co-cultured transwell BBB model, additional antibodies were used including GFAP (glial fibrillary acidic protein, #76214, Dako, 1:200) and S100P (S100 calcium-binding protein B, #76198, Dako, 1:400). Next day, cells were washed three times with PBS, incubated with fluorescence-labeled secondary antibodies including Goat anti-Rabbit Alexa-488 (A-11008, Thermo, 1:500), Goat anti-Rabbit Alexa-555 (A-21428, Thermo, 1:500), Goat anti-Mouse Alexa- 568 (A-21422, Thermo, 1:500) in blocking buffer for 2 hrs in the dark followed by another 10 min incubation with Hoechst in PBS to stain the nuclei. Cells were finally washed three times in PBS and images were acquired by confocal microscopy (Yokogawa). For quantification, 12 random fields per well were taken at 20* magnification and the images were analyzed using Imaged software.

[0147] Specifically, for IFM in the transwell BBB model, membranes of the insert were first cut off and placed in the wells prior to the IFM procedures. For confocal imaging, the stained membranes were inverted and mounted in the wells. For 3D image reconstruction, images were taken at 60 x magnification across the whole cell layer with a 0.2 pm slicing interval, and 3D images were generated using Imaged software. Western blotting (WB)

[0148] A confluent monolayer of either hBMECs or hiPSC-derived BMEC-like cells in a 6-well plate or transwell insert was washed once with ice-cold PBS and lysed on ice for 20 min in RIPA buffer (#89901, Thermo Fisher) supplemented with Halt Protease and Phosphatase Inhibitor Cocktail (#78441, Thermo Fisher). The cell lysate was centrifugated at 12 000 g for 20 min at 4°C. The supernatant was collected, and the protein concentration was determined by Pierce BCA assay kit (#23225, Thermo Fisher). For SDS-PAGE, a total of 25 pg protein was heated at 70°C for 10 min with 4x reducing SDS-PAGE loading buffer (NP0009, Thermo) followed by loading onto 8-16% Tris-glycine gel (XP08165BOX, Thermo) and ran in Tris-glycine buffer (LC2675, Thermo). For WB, protein samples in the polyacrylamide gel were transferred to 0.2 pm PVDF membrane (#1704156, BioRad) using a Trans-Blot Turbo Transfer System (#1704150, BioRad). The membrane was placed in a tray, washed once with double-distilled water (ddH2O), and then blocked in Tris-Buffered Saline (TBS) containing 3% BSA on an orbital shaker for 2 hrs at RT. For the primary antibody, Rabbit anti-Human p-actin (ab8227, Abeam, 1:2000) and Rabbit antiHuman proHB-EGF (ab185555, Abeam, 1:1000) were diluted in TBST (TBS with Tween20) containing 3% BSA. The membrane was transferred to a 50 ml falcon tube and incubated with antibody mix overnight at 4°C on a roller. The next day, the membrane was washed 6 times in TBST with 5 min per wash. For the second antibody, Goat anti-Rabbit Alexa-488 (A-11008, Thermo, 1:5000) was diluted in TBST containing 5% skim milk. The membrane was incubated with the second antibody for 2 hrs at RT on a roller in the dark followed by another 6 times of membrane washing in TBST in the absence of light. The membrane was then imaged by the ChemiDoc MP imaging system (BioRad) using a 488 nm laser line and band intensities were analyzed by Imaged software.

[0149] Statistical analysis

[0150] All data were presented as mean ± standard deviation (SD). Statistical analysis was performed by student's t-test or one-way ANOVA method using GraphPad Prism 9 software. For multiple comparisons, either T ukey’s multiple comparisons tests or Dunnett’s multiple comparisons tests were performed. P values < 0.05 were considered statistically significant. Results

[0151] Selection of proHB-EGF targeted sdAbs

[0152] Two llamas (SNL 166 and SNL 167) were primarily immunized 4 times with a protein mixture of human pre-proHB-EGF (ectodomain, 20-148 aa) and sHB-EGF (64-148 aa) (Fig. 1a), followed by a booster shot of isolated plasma membrane vesicles from COS-7 cells transiently transfected to express full-length human HB-EGF-GFP. The molecular size of HB-EGF-GFP was analyzed by WB and found to have the expected molecular weight. Llama sera were collected 28 days later for titer evaluation of the heavy-chain-only antibodies. Both llamas were found to have developed a humoral response to the two injected proteins (Fig. 1b). SNL 166 showed a higher immune response towards pre-proHB-EGF compared to SNL 167, whereas the immune response towards the sHB-EGF was equal in both llamas. Notably, as sHB-EGF is generated by ectodomain shedding of the membrane-anchored HB-EGF (proHB-EGF), the inventors hence could deduce that there was also the generation of the heavy-chain-only antibodies against proHB-EGF from the immunized llamas. Total mRNA of the peripheral blood lymphocytes (PBLs) of the llamas was then isolated, transcribed into cDNA, and the VHH regions were inserted into a phagemid vector. After 3 rounds of phage selections, VHHs from 25 resultant clones showed specific binding to HB-EGF (either the pre-proHB-EGF or sHB-EGF or both; data not shown) and were subsequently sequenced. This revealed 12 unique sdAb clones with a typical VHH sequence. These candidates were grouped into 4 families according to similarities of the complementarity-determining region (CDR) sequences (data not shown). Five sdAb candidates coming from different families (named A11, C10, G9, F12, H7) were selected for production. All five sdAbs were able to bind to both pre-proHB-EGF and sHB-EGF (Fig. 1c). With comprehensive consideration, 3 sdAbs (A11, C10, G9) were excluded due to either repetitively low production yield or comparably low binding affinity for the target protein. F12 and H7, which belonged to the same family, were selected for further study.

[0153] Binding affinity of fluorophore-labeled sdAbs

[0154] To facilitate further studies, F12 and H7 were labeled with Alexa-647 dye. As control sdAb, R2, which targets the inorganic compound RR6 [29, 30], was in-house produced and labeled as well. All three sdAbs were conjugated to maleimide-Alexa-647 via a C-terminal cysteine on the sdAb, as previously described

[0031] , This conjugation approach was found to not affect the binding of the sdAb to its target [33, 34], All three sdAbs achieved a degree of conjugation (DOC) above 90%, and the free dye percentage was all below 10% (Fig. 1d). Binding affinity of the sdAbs post labeling was evaluated by the protein binding assay. In addition, CRM197, an innate ligand for proHB-EGF, was labeled with Alexa-647 and used as a positive control in the binding assay. Both F12 and H7 were found to preserve their binding ability to sHB-EGF after Alexa-647 labeling (Fig. 1d). The calculated KD for Alexa-647 labeled F12 and H7 was 14.93 nM (95% Cl = 10.55-20.97) and 7.78 nM (95% Cl = 6.38-9.46) respectively, which suggested the binding affinity of H7 to sHB-EGF was two times higher than F12. R2, as a negative control, did not bind to sHB-EGF at all. Notably, as a positive ligand for sHB-EGF, Alexa-647 labeled CRM 197, however, exhibited much reduced binding affinity (7238 nM) than the free form CRM197 / DT ever reported (27 nM)

[0035] , This might be the consequence of labeling Alexa-647 to random lysine of the CRM 197, which resulted in a suboptimal paratope accessibility in terms of binding sHB-EGF.

[0155] To test sdAb binding on cells, the inventors generated a COS-7 cell line transiently expressing the full-length clone of human HB-EGF. As the protein was fused to a GFP tag, the distribution of HB-EGF can be clearly visualized in the cells. The cell membrane was found to be enriched with linear GFP signal (Fig. 1e). Besides, GFP signal was also found in the cytoplasm, which might represent newly synthesized protein. To confirm that the GFP signal located on the membrane also represented proHB-EGF, the inventors labeled the cells with a commercial proHB-EGF antibody and found it perfectly colocalized with the membrane-localized GFP signal (Fig. 1e). Moreover, as COS-7 cells do not express HB-EGF, the commercial antibody did not label un-transfected cells. Hence, the transfected COS-7 cell is a reliable cell platform to test proHB-EGF binding specificity. As a result, the inventors concluded that both F12 and H7 bind specifically to the cell membrane that presented GFP fluorescence, indicating the binding to proHB-EGF (Fig. 1e). On the other hand, negative control R2 did not show any binding to the cells (Fig. 1e).

[0156] sdAbs have no impact on cell viability or BBB integrity

[0157] The safety of this proHB-EGF targeting sdAbs is a crucial prerequisite for further in vivo applications. Here the inventors investigated the sdAb cytotoxicity on the hBMECs. Cells were incubated with 1000 nM sdAb for either 3 hrs or 48 hrs followed by measurement of metabolic activity of the cells using Alamar Blue reagent. None of these sdAbs were found to have an acute or long-term impact on cell viability (Fig. 2a). Moreover, the inventors tested whether the incubation of sdAbs would cause barrier disruption in a transwell-cultured endothelial layer. The results showed that there was no significant difference as compared to the non-treated control (Fig. 2b). On the other hand, incubation of CRM 197 was found to induce substantial BBB leakage (Fig. 2b), which was consistent with earlier studies [20, 36, 37], All together, the inventors demonstrated that the proHB-EGF targeting sdAbs have no obvious cell toxicity or adverse effect on BBB integrity.

[0158] sdAbs binding proHB-EGF on hBMECs

[0159] The inventors confirmed the presence of proHB-EGF in the hBMEC culture by WB (Fig. 3a). The hCMEC / D3, which is a widely used cell line for BBB modeling, was used as a positive control for proHB-EGF.

[0160] Unlike transfected COS-7 cells, on which the bound sdAbs were clearly detected with a strong fluorescence signal (Fig. 1e), binding of F12 or H7 to the well plate-cultured hBMECs led to a faint signal similar to that of R2, indicating absence of any specific binding (Fig. 3b). To check whether proHB-EGF was present on the cell membrane, the inventors compared proHB-EGF staining between live cells and permeabilized cells that were cultured in the well plate. Interestingly, in line with both the proHB-EGF targeting sdAbs (F12 and H7) and the commercial anti-proHB-EGF antibody, proHB-EGF was found to be broadly present in the cytoplasm, but rarely any on the cell surface (Fig. 3b). However, on hBMECs cultured in the transwell, proHB-EGF could be detected on the cell membrane (apical side) as dot-like structures by the commercial antibody (Fig. 3c). Consequently, on the transwell cultured hBMECs layer, F12, and H7 demonstrated superior binding over R2 (Fig. 3d). These results suggested that the polarization of hBMECs in the transwell triggered redistribution of proHB-EGF from cytoplasm to the apical side of the cell membrane, which might be an intrinsic response of endothelial cells to adapt to polarization. Thus, the transwell-based model was considered an optimal platform for evaluating proHB-EGF binding and transport.

[0161] Optimization of proHB-EGF expression in the in-vitro BBB model

[0162] The inventors hypothesized that the increased presence of RMT targets in the model would make it easier to measure the transcytosis efficacy of tested drugs. The inventors optimized the culture conditions to further increase the expression of HB-EGF, which was supposed to proportionally upregulate the amount of proHB-EGF to be transported to the cell surface. The inventors tested the effect of three commercial endothelial media (MV, MV-2, EndoGFR) on the expression of proHB-EGF in hBMEC culture and found that MV-2 medium was able to induce the highest level of proHB-EGF expression (Fig. 4a). Nevertheless, three media induced comparable barrier integrity in the transwell BBB model. Furthermore, the effect of serum level in the MV-2 medium on the expression of proHB-EGF was investigated. It was found that a serum level of 5%, which was originally provided in the MV-2 medium, was useful to maintain the expression of proHB-EGF compared to a low serum level such as 1% (Fig. 4b). Simultaneously, the effects of serum level on BBB integrity were checked, where the 5% serum was found to be also useful for maintaining proper BBB integrity. Based on these results, the MV-2 medium (containing 5% FBS) was chosen for the culture of hBMECs thereafter.

[0163] Cyclic AMP (cAMP) and its analogs (such as forskolin) have regularly been used during the development of in vitro BBB models to induce tight barrier integrity. However, in the presence of serum, the inventors did not see improved barrier tightness by long-term db-cAMP administration. By contrast, this treatment was found to significantly suppress proHB-EGF expression in the hBMECs (Fig. 4b). On the other hand, the inventors found that short-time treatment of serum starvation and cAMP administration [38, 39] could significantly increase BBB impermeability without substantially changing the expression of proHB-EGF (Fig. 4b).

[0164] Glia cells are important for BBB development and maintenance. Co-culture of brain endothelial cells with astrocytes has been widely considered to improve barrier integrity [40-42], The inventors found that the astrocyte co-culture significantly upregulated proHB-EGF expression in hBMECs (Fig. 4c). In parallel, co-culture with astrocytes also greatly enhanced BBB integrity. On the other hand, co-culture with U87 cells, a glioblastoma cell line, slightly decreased proHB-EGF expression in hBMECs (Fig. 4c). Altogether, based on these results, the inventors implemented the optimal cell culture conditions (5% FBS in endothelial medium, astrocyte co-culture, transient treatment of serum starvation and forskolin) to retain good expression of proHB-EGF in the BBB model as well as providing an optimal barrier impermeability.

[0165] sdAb transcytosis on primary cell-based BBB model

[0166] To analyze sdAb transcytosis, the co-cultured BBB transwell model consisting of hBMECs and astrocytes, was used as described above. The BBB model was foundationally characterized by IFM, confirming that confluent brain endothelium was positive for VE-cadherin or ZO-1, and the astrocytes on the backside of the insert were positive for S100P or GFAP (Fig. 5a). Based on BIA data, the model was used for transcytosis assay on day 8 when the barrier integrity was optimal during a monitoring period of 10 days. Before the assay, the BBB model was shortly treated with serum starvation and forskolin stimulation to promote barrier integrity. Alexa-647 labeled sdAbs (F12, H7, R2) were given to the apical side of the BBB, and medium samples from the basal side were collected hourly for three hours. It was shown that both F12 and H7 exhibited significantly enhanced BBB transport, which was approximately two times higher than that of R2 (Fig. 5b and Table. 1).

[0167]

[0168] nonspecific intracellular transport. To exclude sdAb passage possibly caused by BBB leakage, the inventors included 20 kDa dextran as an internal tracer during sdAb transcytosis. Dextran BBB leakage was found to increase over time but it was significantly lower than the passage of F12 and H7 whereas it was close to that of R2. This indicated that the passage of R2 could result from paracellular leakage, meanwhile suggesting an additional BBB transport mechanism for F12 and H7 (Fig. 5b). To further prove the targeted RMT mechanism of F12 and H7, the inventors stained the endothelial barrier with anti-proHB-EGF antibody and examined if there was colocalization between proHB-EGF and the sdAbs. Confocal microscopy revealed that proHB-EGF broadly co-localized with F12 and H7, but not with R2 (Fig. 5c). Importantly, by viewing in 3D, the sdAb-proHB-EGF complex can be found not only at the apical side of the endothelial barrier but was also found in the cytoplasm and at the basolateral membrane (Fig.

[0169] 5d). This provided strong evidence for RMT of the proHB-EGF targeting sdAbs.

[0170] Clathrin-coated vesicles are broadly involved during RMT of the endothelial cells [3, 4], To explore the possible involvement of clathrin pathway in BBB transcytosis of the proHB-EGF targeting sdAbs, the inventors looked for co-localization of clathrin and the sdAbs. Taking F12 as an example, the inventors found it co-localized with or appeared close to clathrin during the transcytosis process. This suggested the possible involvement of clathrin-coated pits during the transcytosis of the proHB-EGF targeting sdAbs.

[0171] sdAb transcytosis on hiPSC-derived iBBB model

[0172] Recently, iPSCs have become a powerful tool to derive various cell types and bear great advantages over the use of primary cells such as easy accessibility and high reproducibility. Differentiation of hiPSCs into induced brain endothelial cells has been reported and the derivative BBB models (termed iBBB model) were intensively explored for the screening of novel BBB penetrating antibodies / sdAbs [43, 44], To provide more evidence for BBB transcytosis, F12 and H7 were also tested on an iBBB model. The inventors differentiated hiPSCs into BMEC-like cells and confirmed the expression of a set of brain endothelial markers such as ZO-1, Claudin-5, Occludin, GLUT-1, and P-gp (Fig. 6a). Expression of proHB-EGF was confirmed in these cells, the abundance of which was significantly higher than that of hBMECs (Fig. 6a, b). A simplified iBBB model consisting of a monolayer of BMEC-like cells was generated in the transwell system. Unlike the hBMEC model, the iBBB model demonstrated superior impermeability, which greatly minimized paracellular leakage of the dextran (Fig. 6c and Table.

[0173] 1). Consequently, the endothelial transport of three tested sdAbs dropped drastically. Nevertheless, F12 and H7 still demonstrated significantly higher iBBB passage than R2. The highest transport on the iBBB model, which was achieved by F12, reached 0.121 ± 0.003% in 3 sdAbs (Table. 1). In comparison, H7 reached 0.079 ± 0.001% in 3 sdAbs. Notably, even though much lower, R2 was also able to cross the iBBB model, as has been seen in the hBMEC model. This indicated again that the endothelial cells were able to transport R2 in a non-specific way. Taken together, with both of the BBB models, the inventors concluded that the F12 and H7 sdAbs were able to undergo RMT via targeting proHB-EGF.

[0174] Implications

[0175] ProHB-EGF is constitutively expressed in the CNS of the human brain and is strongly upregulated after cerebral ischemia and reperfusion injury [9, 12], Increased expression of proHB-EGF in the BBB is also observed following neuroinflammation

[0011] , Therefore, proHB-EGF can be harnessed as an RMT target for shuttling therapeutic molecules into the brain to treat certain brain diseases. Moreover, it is conceivable that the single use of a proHB-EGF targeting ligand could provide a dual-targeting strategy

[0045] where, after passing the BBB by proHB-EGF mediated transcytosis it could continue to bind to proHB-EGF enriched targets such as inflamed neural cells [46, 47] or brain tumor cells [48, 49], Therapeutic molecules that are conjugated to the proHB-EGF targeting ligand may include chemotherapy drugs, protein-based drugs, and nanoparticles with payload [6, 24, 25, 50-53], To date, only CRM 197 has been explored as a BBB shuttle molecule by targeting proHB-EGF. Thus, the potential of this receptor to facilitate BBB transcytosis remains boundless and merits more investigation. Due to lower immunogenicity, smaller size, and better structural stability, sdAbs are favorable for various applications such as in vivo diagnostics, therapy, and drug delivery [22, 23, 52], In the current study, the inventors derived two sdAbs that achieve RMT through the BBB by targeting proHB-EGF. The results provided promising evidence for further development of these sdAbs to serve as a BBB shuttle molecule that carries therapeutic molecules into the brain.

[0176] The protein binding assay on sHB-EGF using Alexa-647 labeled sdAbs revealed a higher binding affinity of H7 than F12, with the KD of the latter being approximately two times higher. In line with this, H7 showed a slightly better binding on the hBMECs than F12. However, the transcytosis assay has shown an inverse outcome, with F12 exhibiting better transport than H7. The superiority of F12 over H7 was minor in the hBMEC-based BBB model but became significantly more pronounced in the hiPSC-based iBBB model. This might be a consequence of both the presence of a tighter barrier in the iBBB model and the intrinsically better transcytosis potential of F12. The latter can possibly be explained by the favorable sdAb binding characteristics of F12, more particularly, the on- / off- rates. Cumulative evidence has already suggested that only a moderate binding affinity to the targeted receptor could result in the best RMT efficacy

[0054] , As a well-studied case, it was found that the TfR targeting antibodies / sdAbs with high binding affinity resulted in low transcytosis efficiency [55-59], Besides, the strong receptor-ligand association could alternatively lead to lysosomal degradation [55-59], To some extent, this reversed relationship between binding affinity and transcytosis capacity has also been seen for the proHB-EGF targeting sdAbs, where H7 did not yield better BBB transcytosis than F12 despite a higher binding affinity. On the other hand, it remains to be determined if F12 has fulfilled its best transcytosis potential. This could be investigated further by introducing site-directed mutagenesis in general paratope regions (e.g., CDR3) of the sdAb to modulate the binding affinity without affecting the binding specificity. For instance, sdAbs may be tuned to be pH-sensitive by introducing histidine residues in the CDRs. This strategy was found to increase the antibody-receptor dissociation in the acidified endosome and hence the improved release of shuttle antibodies at the basolateral side of the endothelial cell

[0058] , Thus, a panel of F12 variants differing in amino acid (s) in the CDR3 region could be generated and tested comprehensively for the correlation between binding strength and transcytosis efficiency [59, 60],

[0177] Despite being the negative control molecule, the inventors noticed that R2 could also be actively endocytosed / transcytosed by the brain endothelial cells (Fig. 5c). Considering that the endothelial cells are equipped with abundant adsorption / uptake / transport mechanisms [61-64], it is common to see a low level of unspecific BBB uptake / passage of the negative control materials during both in vivo and in vitro transcytosis experiments

[0065] , Notably, this could even hold for the active-targeting molecules being tested. For example, the inventors also found that some F12 traveled through the brain endothelial cells without complexing with proHB-EGF, which suggested non-specific transport of F12, as well. This could be attributed to the static experimental condition which increases the chance of the endothelial cells to capture and transport these cargos in a receptor-independent way such as adsorptive-mediated transcytosis (AMT)

[0064] , In addition, despite that the inventors showed the proHB-EGF targeting sdAbs could co-localize with clathrin during BBB transcytosis, it remains a question whether these clathrin pits were encapsulating the sdAb-proHB-EGF complexes or not. As clathrin pathway also participates in non-RMT mechanisms including the AMT

[0064] , it requires further research to confirm which transport pathways are involved in the RMT of these proHB-EGF targeting sdAbs. Essentially, a tri-localization IFM analysis of the sdAb, the proHB-EGF, and the pathway participators such as clathrin or caveolin, could provide more evidence. Besides, specific pathway inhibition assays would also be relevant to fully determine the pathways involved

[0066] , The availability of transcytosis receptors resulting from expression level and surface density on the apical surface are important parameters to consider when developing an RMT strategy. For an in vitro BBB model designated for drug pre-selection study, the low presence of the receptors makes it difficult to reflect the real transcytosis efficacy of the tested molecules. From both literature and our own study, it was found that various culture conditions can affect the expression and cellular distribution of proHB-EGF [11, 67-69], One interesting observation of the current study was that proHB-EGF had less display on the endothelial surface when cultured on the 2D surface (such as in a well plate), making it inaccessible for the proHB-EGF-specific sdAbs. However, under polarized culture conditions (such as in the transwell culture insert), there was more proHB-EGF localized to the cell surface. The mechanism currently remains unclear to the inventors. As endothelial cells are innately growing in a polarized environment on which the activity of RMT is heavily based, the polarized culture approaches are therefore necessary for reproducing this specialized endothelial phenotype in vitro

[0070] , In terms of the expression level of proHB-EGF, the inventors found that the hBMECs express lower proHB-EGF than the hiPSC-derived BMEC-like cells or hCMEC / D3 cell line. The inventors optimized the culture conditions to upregulate proHB-EGF in the hBMECs and the most efficient booster was found to be the astrocyte co-culture. Astrocyte co-culture was shown to prevent BBB dedifferentiation by rescuing the downregulation of various functional receptors of the in vitro cultured brain endothelial cells [42, 71-73], However, proHB-EGF currently is not physiologically considered as a “receptor” as there has been no evidence so far of any endogenous ligands recognizing it in the human body. Thus, the mechanism of elevated expression of endothelial proHB-EGF by astrocyte co-culture seems to be different. Nonetheless, it has been stated earlier that the astrocyte was a necessary inducer of proHB-EGF expression on the BBB

[0013] and that the up-regulation of proHB-EGF in the brain endothelial cells under diseased conditions also depends on the presence of astrocytes

[0013] ,

[0178] The inventors have noticed that the co-cultured hBMEC-based BBB model showed non-negligible paracellular leakage during sdAb transcytosis assay. Nonetheless, the imperfect barrier integrity of primary brain endothelial cells during in-vitro culture has been an unsolved issue [74, 75], In comparison, the hiPSC-based iBBB model yielded superior barrier impermeability that was revealed by the clear presence of various tight junction molecules together with their smooth continuity at cell borders. The inventors did not include astrocyte coculture in the iBBB model as the addition of astrocytes was found to yield a much higher barrier impermeability than the monoculture iBBB model, which raised the concern that the transcytosed sdAbs will fall below the detecting threshold by our current method. Besides, the expression of proHB-EGF in the monoculture iBBB model has reached a level similar to the cocultured hBMEC-based BBB model, which was sufficient for the transcytosis experiment. Due to high paracellular resistance, the iBBB model has greatly limited the passage for the 20 kDa dextran and the negative sdAb R2. However, the low transcytosis efficiency of the proHB-EGF targeting sdAbs achieved in the iBBB model, as compared to the hBMEC-based BBB model, was also noticeable. Notably, most hiPSC-derived brain endothelial cells nowadays should be rather addressed as BMEC-like cells since transcriptomics revealed contamination of epithelial or neuroepithelial transcripts within these cells [76, 77], Importantly, proHB-EGF enriched in the neuroepithelial population was not even speculated to exert a transcytosis function [9, 78], Furthermore, the population of generated brain endothelial cells was found to constitute only a small portion

[0076] , Together, these might explain why the sdAb transcytosis resulted from the hiPSC-iBBB model was lower than that in the hBM EC-based BBB model despite a similar proHB-EGF expression. On the other hand, inconsistent transcytosis results gained from comparable BBB models were not rare. For example, one study found that the hiPSC-iBBB model showed significantly less transcytosis for the TfR targeting antibody than that from an immortalized cell-based BBB model, even though the iBBB model expresses much higher TfR

[0079] , Nonetheless, despite the immaturity and impurity of hiPSC-derived brain endothelial cells ever reported, the advantages of having high barrier integrity and necessary expression of various functional BBB transporters [44, 80] still make these iBBB models suitable fortransport studies.

[0179] Example 2. Comparison of sdAb with different targets

[0180] To analyze the transcytosis of sdAbs against different targets, the co-cultured BBB transwell model consisting of primary hBMECs and astrocytes can be used as described earlier herein. The BBB model is foundationally characterized by IFM, where the confluent brain endothelium is positive for VE-cadherin or ZO-1, and the astrocytes on the backside of the insert are positive for S100P or GFAP. Based on BIA data, the model can be used for transcytosis assay on day 8 when the barrier integrity is optimal during a monitoring duration of 10 days. Before the assay, the BBB model is to be shortly treated with serum starvation and forskolin simulation to promote barrier integrity. Alexa-647 labeled sdAbs can be given to the apical side of the BBB and medium samples from the basal side can be collected hourly for three hours.

[0181] sdAbs targeting HB-EGF exhibit significantly enhanced BBB transport and can be found in the cytoplasm at the basolateral membrane at much higher concentrations than the sdAbs targeting different targets. See Figure 7.

[0182] Indeed, imaging studies have for example shown that anti-Tfr nanobodies deliver only 0.15- 0.2% of the injected dose to the CSF, with the majority accumulating in bone marrow and placenta (see Balligand et al., eLife 2025;14: RP104302: https: / / doi. Org / 10.7554 / eLife.10430). Also, the IGF receptor (IGFR) pathway is not suitable as an effective or selective BBB transporter. In a human in-vitro BBB transcytosis assay, an anti-IGF1R antibody was shown to be recycled back to the apical membrane rather than undergoing productive transcytosis, thereby failing to deliver cargo across the barrier (Sade et al 2014, PLoS ONE 9(4): e96340,

[0183]

[0184] In contrast, proHB-EGF is a much more attractive transcytosis target at the blood-brain barrier (BBB) because it differs from these well-studied alternatives like the transferrin or insulin receptors. These classical receptors suffer from competition with their abundant endogenous ligands, which lowers the efficiency of drug transport and risks interfering with essential nutrient supply. By contrast, proHB-EGF is a membrane-anchored precursor that internalizes readily but does not have a known endogenous ligand in the circulation, meaning there is no natural competition for binding sites. It is also constitutively expressed in brain endothelial cells and becomes further upregulated under pathological conditions such as ischemia and neuroinflammation, which could increase transport opportunities in disease states. Compared with CRM197, the only other molecule previously used to exploit proHB-EGF, the approach according to the present disclosure avoids toxicity issues and non-specific caveolar transport, while still harnessing receptor-mediated uptake. Together, these properties make proHB-EGF a compelling alternative target that promises more specific, safer, and potentially more efficient BBB shuttling (see Qiu et al, Journal of Controlled Release, Volume 383, 10 July 2025, 113852).

[0185] A later publication by Qiu et al (J Control Release. 2025 Sep 23:387:114257), builds on the transwell work by introducing a microfluidic chip to study nanobody transport under flow. This environment reduces nonspecific background and better reflects physiological barrier properties. The data again support proHB-EGF as a viable receptor, showing that the nanobodies selected against it can cross in both static and dynamic models, reinforcing its advantages compared to conventional targets.

[0186] Bibliography

[0187] 1. Pardridge, SDABS. M., A Historical Review of Brain Drug Delivery. Pharmaceutics, 2022. 14(6).

[0188] 2. Sweeney, M. D., et al., Blood-Brain Barrier: From Physiology to Disease and Back.

[0189] Physiol Rev, 2019. 99(1): p. 21-78.

[0190] 3. Pulgar, V. M., Transcytosis to Cross the Blood Brain Barrier, New Advancements and Challenges. Front Neurosci, 2018. 12: p. 1019.

[0191] 4. Haqqani, A. S. and D. B. Stanimirovic, Brain delivery of therapeutics via transcytosis: types and mechanisms of vesicle-mediated transport across the BBB. Drug Delivery to the Brain: Physiological Concepts, Methodologies and Approaches, 2022: p. 71-91. 5. Zhao, P, N. Zhang, and Z. An, Engineering antibody and protein therapeutics to cross the blood-brain barrier. Antib Ther, 2022. 5(4): p. 311-331.

[0192] 6. Pardridge, SDABS. M., Blood-Brain Barrier and Delivery of Protein and Gene Therapeutics to Brain. Front Aging Neurosci, 2019. 11: p. 373.

[0193] 7. Ohshima-Hosoyama, S., et al., A monoclonal antibody-GDNF fusion protein is not neuroprotective and is associated with proliferative pancreatic lesions in parkinsonian monkeys. PLoS One, 2012. 7(6): p. e39036.

[0194] 8. Couch, J. A., et al., Addressing safety liabilities of TfR bispecific antibodies that cross the blood-brain barrier. Sci Transl Med, 2013. 5(183): p. 183ra57, 1-12.

[0195] 9. Oyagi, A. and SDABS. Hara, Essential roles of heparin-binding epidermal growth factor-like growth factor in the brain. CNS Neurosci Ther, 2012. 18(10): p. 803-10. 10. Raab, G. and M. Klagsbrun, Heparin-binding EGF-like growth factor. Biochim Biophys Acta, 1997. 1333(3): p. F179-99.

[0196] 11. Yoshizumi, M., et al., Tumor necrosis factor increases transcription of the heparin- binding epidermal growth factor-like growth factor gene in vascular endothelial cells. J Biol Chem, 1992. 267(14): p. 9467-9.

[0197] 12. Kawahara, N., et al., The gene for heparin-binding epidermal growth factor-like growth factor is stress-inducible: its role in cerebral ischemia. J Cereb Blood Flow Metab, 1999. 19(3): p. 307-20.

[0198] 13. Gaillard, P. J., A. Brink, and A. G. De Boer. Diphtheria toxin receptor-targeted brain drug delivery, in International Congress Series. 2005. Elsevier.

[0199] 14. Malito, E., et al., Structural basis for lack of toxicity of the diphtheria toxin mutant CRM197. Proc Natl Acad Sci U SA, 2012. 109(14): p. 5229-34.

[0200] 15. Qiao, J., K. Ghani, and M. Caruso, Diphtheria toxin mutant CRM197 is an inhibitor of protein synthesis that induces cellular toxicity. Toxicon, 2008. 51(3): p. 473-477.

[0201] 16. Lee, J. SDABS., et al., Mechanistic aspects of the deoxyribonuclease activity of diphtheria toxin. Biochim Biophys Acta, 2005. 1747(1): p. 121-31.

[0202] 17. Bruce, C., et al., Diphtheria toxin and its ADP-ribosyltransferase-defective homologue CRM197 possess deoxyribonuclease activity. Proc Natl Acad Sci U SA, 1990. 87(8): p. 2995-8.

[0203] 18. Varol, B., et al., The cytotoxic effect of diphtheria toxin on the actin cytoskeleton. Cell Mol Biol Lett, 2012. 17(1): p. 49-61.

[0204] 19. Tosi, G., et al., Exploiting Bacterial Pathways forBBB Crossing with PLGA Nanoparticles Modified with a Mutated Form of Diphtheria Toxin (CRM197): In Vivo Experiments. Mol Pharm, 2015. 12(10): p. 3672-84.

[0205] 20. Lin, SDABS., et al., DT(270-326), a Truncated Diphtheria Toxin, Increases Blood- Tumor Barrier Permeability by Upregulating the Expression of Caveolin-1. CNS Neurosci Ther, 2016. 22(6): p. 477-87.

[0206] 21. Emmerson, C. D., et al., Enhancement of polymeric immunoglobulin receptor transcytosis by biparatopic VHH. PLoS One, 2011. 6(10): p. e26299.

[0207] 22. Muyldermans, S., Applications of Nanobodies. Annu Rev Anim Biosci, 2021. 9: p. 401- 421.

[0208] 23. Oliveira, S., et al., Targeting tumors with nanobodies for cancer imaging and therapy. J Control Release, 2013. 172(3): p. 607-17.

[0209] 24. Ruiz-Lopez, E. and A. J. Schuhmacher, Transportation of Single-Domain Antibodies through the Blood-Brain Barrier. Biomolecules, 2021. 11(8).

[0210] 25. Zheng, F, et al., Applications of nanobodies in brain diseases. Front Immunol, 2022.

[0211] 13: p. 978513.

[0212] 26. Okita, K., et al., A more efficient method to generate integration-free human iPS cells.

[0213] Nat Methods, 2011. 8(5): p. 409-12.

[0214] 27. Kaplan, M., et al., EGFR Dynamics Change during Activation in Native Membranes as Revealed by N MR. Cell, 2016. 167(5): p. 1241-1251 e11.

[0215] 28. Roovers, R. C., et al., Efficient inhibition of EGFR signaling and of tumour growth by antagonistic anti-EFGR Nanobodies. Cancer Immunol Immunother, 2007. 56(3): p. 303-317.

[0216] 29. Frenken, L. G., et al., Isolation of antigen specific llama VHH antibody fragments and their high level secretion by Saccharomyces cerevisiae. Journal of biotechnology, 2000. 78(1): p. 11-21.

[0217] 30. Dolk, E., et al., Induced refolding of a temperature denatured llama heavy-chain antibody fragment by its antigen. Proteins, 2005. 59(3): p. 555-64.

[0218] 31. Pronk, S. D., et al., Single domain antibodies as carriers for intracellular drug delivery:

[0219] A proof of principle study. Biomolecules, 2021. 11(7): p. 927.

[0220] 32. Grifno, G. N., et al., Tissue-engineered blood-brain barrier models via directed differentiation of human induced pluripotent stem cells. Sci Rep, 2019. 9(1): p. 13957. 33. Kijanka, M., et al., Rapid optical imaging of human breast tumour xenografts using anti-HER2 VHHs site-directly conjugated to IRDye 800CW for image-guided surgery. Eur J Nucl Med Mol Imaging, 2013. 40(11): p. 1718-29.

[0221] 34. De Groof, T. SDABS. M., et al., Nanobody-Targeted Photodynamic Therapy Selectively Kills Viral GPCR-Expressing Glioblastoma Cells. Mol Pharm, 2019. 16(7): p. 3145- 3156.

[0222] 35. Brooke, J. S., J. SDABS. Cha, and L. Eidels, Diphtheria toxin receptor interaction:

[0223] association, dissociation, and effect of pH. Biochem Biophys Res Commun, 1998. 248(2): p. 297-302.

[0224] 36. Wang, P., et al., Diphtheria toxin mutant CRM197-mediated transcytosis across bloodbrain barrier in vitro. Cell Mol Neurobiol, 2010. 30(5): p. 717-25.

[0225] 37. Wang, P, et al., CRM197-induced blood-brain barrier permeability increase is mediated by upregulation of caveolin-1 protein. J Mol Neurosci, 2011. 43(3): p. 485-92.

[0226] 38. Fukuhara, S., et al., Cyclic AMP potentiates vascular endothelial cadherin-mediated cell-cell contact to enhance endothelial barrier function through an Epac-Rap1 signaling pathway. Mol Cell Biol, 2005. 25(1): p. 136-46.

[0227] 39. Nitz, T., et al., Serum-derived factors weaken the barrier properties of cultured porcine brain capillary endothelial cells in vitro. Brain Res, 2003. 981(1-2): p. 30-40.

[0228] 40. Canfield, S. G., et al., An isogenic blood-brain barrier model comprising brain endothelial cells, astrocytes, and neurons derived from human induced pluripotent stem cells. J Neurochem, 2017. 140(6): p. 874-888.

[0229] 41. Appelt-Menzel, A., et al., Establishment of a Human Blood-Brain Barrier Co-culture Model Mimicking the Neurovascular Unit Using Induced Pluri- and Multipotent Stem Cells. Stem Cell Reports, 2017. 8(4): p. 894-906.

[0230] 42. Delsing, L., et al., Barrier Properties and Transcriptome Expression in Human iPSC- Derived Models of the Blood-Brain Barrier. Stem Cells, 2018. 36(12): p. 1816-1827. 43. Georgieva, J. V., et al., Antibody screening using a human iPSC-based blood-brain barrier model identifies antibodies that accumulate in the CNS. FASEB J, 2020. 34(9): p. 12549-12564.

[0231] 44. Ribecco-Lutkiewicz, M., et al., A novel human induced pluripotent stem cell blood-brain barrier model: Applicability to study antibody-triggered receptor-mediated transcytosis. Sci Rep, 2018. 8(1): p. 1873.

[0232] 45. Gao, SDABS., Perspectives on Dual Targeting Delivery Systems for Brain Tumors. J Neuroimmune Pharmacol, 2017. 12(1): p. 6-16.

[0233] 46. Jin, K., et al., Heparin-binding epidermal growth factor-like growth factor: hypoxiainducible expression in vitro and stimulation of neurogenesis in vitro and in vivo. J Neurosci, 2002. 22(13): p. 5365-73.

[0234] 47. Tanaka, N., et al., Heparin-binding epidermal growth factor-like growth factor mRNA expression in neonatal rat brain with hypoxic / ischemic injury. Brain Res, 1999. 827(1- 2): p. 130-8.

[0235] 48. Shin, C. SDABS., et al., HBEGF promotes gliomagenesis in the context oflnk4a / Arf and Pten loss. Oncogene, 2017. 36(32): p. 4610-4618.

[0236] 49. Mishima, K., et al., Heparin-binding epidermal growth factor-like growth factor stimulates mitogenic signaling and is highly expressed in human malignant gliomas. Acta Neuropathol, 1998. 96(4): p. 322-8.

[0237] 50. Pothin, E., D. Lesuisse, and P. Lafaye, Brain Delivery of Single-Domain Antibodies: A Focus on VHH and VNAR. Pharmaceutics, 2020. 12(10).

[0238] 51. Wouters, SDABS., et al., VHHs as tools for therapeutic protein delivery to the central nervous system. Fluids Barriers CNS, 2022. 19(1): p. 79.

[0239] 52. Tsitokana, M. E., et al., Targeting the Brain with Single-Domain Antibodies: Greater Potential Than Stated So Far? I nt J Mol Sci, 2023. 24(3).

[0240] 53. Abdolvahab, M. SDABS., et al., Targeted drug delivery using nanobodies to deliver effective molecules to breast cancer cells: the most attractive application of nanobodies. Cancer Cell Int, 2024. 24(1): p. 67.

[0241] 54. Kouhi, A., et al., Brain Disposition of Antibody-Based Therapeutics: Dogma, Approaches and Perspectives. Int J Mol Sci, 2021. 22(12). 55. Niewoehner, J., et al., Increased brain penetration and potency of a therapeutic antibody using a monovalent molecular shuttle. Neuron, 2014. 81(1): p. 49-60.

[0242] 56. Haqqani, A. S., et al., Intracellular sorting and transcytosis of the rat transferrin receptor antibody 0X26 across the blood-brain barrier in vitro is dependent on its binding affinity J Neurochem, 2018. 146(6): p. 735-752.

[0243] 57. Yu, SDABS. J., et al., Boosting brain uptake of a therapeutic antibody by reducing its affinity fora transcytosis target. Sci Transl Med, 2011. 3(84): p. 84ra44.

[0244] 58. Sade, SDABS., et al., A human blood-brain barrier transcytosis assay reveals antibody transcytosis influenced by pH-dependent receptor binding. PLoS One, 2014. 9(4): p. e96340.

[0245] 59. Wouters, SDABS., et al., Identification and in vivo characterization of a brainpenetrating nanobody. Fluids Barriers CNS, 2020. 17(1): p. 62.

[0246] 60. Zimmermann, I., et al., Synthetic single domain antibodies for the conformational trapping of membrane proteins, elite, 2018. 7: p. e34317.

[0247] 61. Mayor, S., R. G. Parton, and J. G. Donaldson, Clathrin-independent pathways of endocytosis. Cold Spring Harb Perspect Biol, 2014. 6(6).

[0248] 62. Sandvig, K., S. Kavaliauskiene, and T. Skotland, Clathrin-independent endocytosis: an increasing degree of complexity. Histochem Cell Biol, 2018. 150(2): p. 107-118.

[0249] 63. Knox, E. G., et al., The blood-brain barrier in aging and neurodegeneration. Mol Psychiatry, 2022. 27(6): p. 2659-2673.

[0250] 64. Herve, F., N. Ghinea, and J. M. Scherrmann, CNS delivery via adsorptive transcytosis.

[0251] AAPS J, 2008. 10(3): p. 455-72.

[0252] 65. Haqqani, A. S., et al., Multiplexed evaluation of serum and CSF pharmacokinetics of brain-targeting single-domain antibodies using a NanoLC-SRM-ILIS method. Mol Pharm, 2013. 10(5): p. 1542-56.

[0253] 66. Wang, SDABS., et al., Disruption of clathrin-dependent trafficking results in the failure of grass carp reovirus cellular entry. Virol J, 2016. 13: p. 25.

[0254] 67. Hieda, M., et al., Membrane-anchored growth factor, HB-EGF, on the cell surface targeted to the inner nuclear membrane. J Cell Biol, 2008. 180(4): p. 763-9.

[0255] 68. Morita, T, et al., Shear stress increases heparin-binding epidermal growth factor-like growth factor mRNA levels in human vascular endothelial cells. Biochem Biophys Res Commun, 1993. 197(1): p. 256-62.

[0256] 69. NasdAba, D. and S. Higashiyama, Dual intracellular signaling by proteolytic cleavage of membrane-anchored heparin-binding EGF-like growth factor. Cytokine Growth Factor Rev, 2004. 15(1): p. 13-9.

[0257] 70. Worzfeld, T. and M. Schwaninger, Apicobasal polarity of brain endothelial cells. J Cereb Blood Flow Metab, 2016. 36(2): p. 340-62.

[0258] 71. Hayashi, SDABS., et al., Induction of various blood-brain barrier properties in non- neural endothelial cells by close apposition to co -cultured astrocytes. Glia, 1997. 19(1): p. 13-26.

[0259] 72. Ito, R., et al., A Human Immortalized Cell-Based Blood-Brain Barrier Triculture Model:

[0260] Development and Characterization as a Promising Tool for Drug-Brain Permeability Studies. Mol Pharm, 2019. 16(11): p. 4461-4471.

[0261] 73. Watson, P. M. D., et al., Bioengineered 3D Glial Cell Culture Systems and Applications for Neurodegeneration and Neuroinflammation. SLAS Discov, 2017. 22(5): p. 583-601.

[0262] 74. Lyck, R., et al., Culture-induced changes in blood — brain barrier transcriptome:

[0263] implications for amino-acid transporters in vivo. Journal of Cerebral Blood Flow & Metabolism, 2009. 29(9): p. 1491-1502.

[0264] 75. Sabbagh, M. F and J. Nathans, A genome-wide view of the de-differentiation of central nervous system endothelial cells in culture. Elite, 2020. 9: p. e51276.

[0265] 76. Lu, T. M., et al., Pluripotent stem cell-derived epithelium misidentified as brain microvascular endothelium requires ETS factors to acquire vascular fate. Proceedings of the National Academy of Sciences, 2021. 118(8): p. e2016950118.

[0266] 77. Lippmann, E. S., et al., Commentary on human pluripotent stem cell-based blood-brain barrier models. Fluids Barriers CNS, 2020. 17(1): p. 64. 78. Romano, R. and C. Bucci, Role ofEGFR in the Nervous System. Cells, 2020. 9(8). 79. Piantino, M., et al., Brain microvascular endothelial cells derived from human induced pluripotent stem cells as in vitro model for assessing blood-brain barrier transferrin receptor-mediated transcytosis. Materials Today Bio, 2022. 14: p. 100232.

[0267] 80. Kurosawa, T., et al., Expression and Functional Characterization of Drug Transporters in Brain Microvascular Endothelial Cells Derived from Human Induced Pluripotent Stem Cells. Mol Pharm, 2018. 15(12): p. 5546-5555.

[0268] 81. Mager, I., et al., Targeting blood-brain-barrier transcytosis - perspectives for drug delivery. Neuropharmacology, 2017. 120: p. 4-7.

[0269] 82. Song, SDABS. SDABS., et al., Transcriptomic comparison of human and mouse brain microvessels. Sci Rep, 2020. 10(1): p. 12358.

[0270] 83. Deo, A. K., F. P Theil, and J. M. Nicolas, Confounding parameters in preclinical assessment of blood-brain barrier permeation: an overview with emphasis on species differences and effect of disease states. Mol Pharm, 2013. 10(5): p. 1581-95.

[0271] 84. Uchida, SDABS., et al., Quantitative targeted absolute proteomics of human bloodbrain barrier transporters and receptors. J Neurochem, 2011. 117(2): p. 333-45.

[0272] 85. Zhang, SDABS., et al., Differential expression of receptors mediating receptor- mediated transcytosis (RMT) in brain microvessels, brain parenchyma and peripheral tissues of the mouse and the human. Fluids Barriers CNS, 2020. 17(1): p. 47.

[0273] 86. Zhang, SDABS., et al., Purification and Characterization of Progenitor and Mature Human Astrocytes Reveals Transcriptional and Functional Differences with Mouse. Neuron, 2016. 89(1): p. 37-53.

[0274] 87. Saito, M., et al., Diphtheria toxin receptor-mediated conditional and targeted cell ablation in transgenic mice. Nat Biotechnol, 2001. 19(8): p. 746-50.

[0275] 88. Wang, SDABS., et al., Emerging trends in organ-on-a-chip systems fordrug screening.

[0276] Acta Pharm Sin B, 2023. 13(6): p. 2483-2509.

[0277] 89. Ingber, D. E., Human organs-on-chips for disease modelling, drug development and personalized medicine. Nat Rev Genet, 2022. 23(8): p. 467-491.

[0278] 90. Picollet-D'hahan, N., et al., Multiorgan-on-a-Chip: A Systemic Approach To Model and Decipher Inter-Organ Communication. Trends Biotechnol, 2021. 39(8): p. 788-810. 91. Tekin, SDABS., et al., Effects of 3D culturing conditions on the transcriptomic profile of stem-cell-derived neurons. Nat Biomed Eng, 2018. 2(7): p. 540-554.

[0279] 92. Mandrycky, C., B. Hadland, and SDABS. Zheng, 3D curvature-instructed endothelial flow response and tissue vascularization. Sci Adv, 2020. 6(38).

[0280] 93. Pediaditakis, I., et al., Modeling alpha-synuclein pathology in a human brain-chip to assess blood-brain barrier disruption. Nat Commun, 2021. 12(1): p. 5907.

[0281] 94. Sances, S., et al., Human iPSC-Derived Endothelial Cells and Microengineered Organ- Chip Enhance Neuronal Development. Stem Cell Reports, 2018. 10(4): p. 1222-1236.

[0282] 95. Kim, J., et al., Fungal brain infection modelled in a human-neurovascular-unit-on-a- chip with a functional blood-brain barrier. Nat Biomed Eng, 2021. 5(8): p. 830-846. 96. Cucullo, L., et al., Immortalized human brain endothelial cells and flow-based vascular modeling: a marriage of convenience for rational neurovascular studies. J Cereb Blood Flow Metab, 2008. 28(2): p. 312-28.

[0283] 97. Walter, F. R., et al., A versatile lab-on-a-chip tool for modeling biological barriers.

[0284] Sensors and Actuators B: Chemical, 2016. 222: p. 1209-1219.

Claims

CLAIMS1. Single domain antibody that binds Heparin-binding EGF-like growth factor (HB-EGF) for use in imaging, treating and / or diagnosing a brain disease,wherein the single domain antibody is in combination with a therapeutic compound, imaging compound and / or a diagnostic compound.

2. Single domain antibody for use according to claim 1, wherein the brain disease is selected from a disease characterized by undesired protein aggregation, amyloidosis, Alzheimer’s Disease, Parkinson’s Disease, Huntington’s Disease, brain tumor, Hunter syndrome, Mucopolysaccharidosis type II (MPS II), encephalitis, headache, cluster headache, migraine, neurodevelopmental disorder, autism and epilepsy.

3. Single domain antibody for use according to any one of the previous claims, wherein the single domain antibody is a Variable domain of the Heavy chain of a Heavy-chain-only (VHH) antibody.

4. Single domain antibody for use according to claim 3, wherein the VHH antibody has the following structure with complementarity determining regions (CDR) and framework regions (FR):(N-terminus) FR1 - CDR1 - FR2 - CDR2 - FR3 - CDR3 - FR4 (C-terminus).

5. Single domain antibody for use according to claim 4, wherein the single domain antibody comprises:- CDR1 having an amino acid sequence of SEQ ID NO:30;- CDR2 having an amino acid sequence of SEQ ID NO:32;- CDR3 having an amino acid sequence of SEQ ID NO:34.

6. Single domain antibody for use according to claim 4, wherein the single domain antibody comprises:- CDR1 having an amino acid sequence of SEQ ID NO:44;- CDR2 having an amino acid sequence of SEQ ID NO:46;- CDR3 having an amino acid sequence of SEQ ID NO:48.

7. Single domain antibody for use according to any one of the previous claims wherein the single domain antibody comprises an amino acid sequence of SEQ ID NO:3 or SEQ ID NO:5, most preferably SEQ ID NO:3.

8. Single domain antibody for use according to any one of the previous claims, wherein the therapeutic compound, imaging compound and / or the diagnostic compound is conjugated to the single domain antibody.

9. Single domain antibody for use according to any one of the previous claims, wherein the therapeutic compound, imaging compound and / or a diagnostic compound is conjugated to the single domain antibody by means of- genetic fusion;- Cys-Maleimide conjugation;- NHS ester chemistry;- Click chemistry, preferably azide-alkyne cycloaddition, or strain-promoted azide-alkyne cycloaddition (SPAAC);- Protein ligation, preferably by means of sortase, intein, or transglutaminase.

10. Single domain antibody for use according to any one of the previous claims, wherein the therapeutic compound is one or more of- cytotoxic agent;- RNA, preferably siRNA, RNAi, miRNA, and preferably capable of modulation of protein synthesis;- growth factor, epidermal growth factor (EGF), insulin-like growth factor (IGF);- protein, preferably a binding peptide that binds one or more of Epidermal Growth Factor Receptor variant III (EGFvIll), beta-amyloid, Tau, alpha-synuclein, prion protein and / or wherein the binding peptide is chosen from single domain antibody, affibody, Designed Ankyrin Repeat Proteins (DARPin) or monoclonal antibody or fragment thereof;- enzyme, to replace deficient or malfunctioning extracellular enzymes, preferably chosen from iduronate-2-sulfatase (I2S);- cytokine preferably chosen from lnterleukin-6, Interleukin-10, Tumor Necrosis Factor-alpha, Transforming Growth Factor-beta;- neurotrophic factor, preferably chosen from Nerve Growth Factor, Brain-Derived Neurotrophic Factor, Glial Cell Line-Derived Neurotrophic Factor, Neurotrophin-3, Ciliary Neurotrophic Factor, proteins such as cytokines and neurotrophic factors, to promote neuronal cell survival and repair; and / or- a therapeutic compound comprised in a nanoparticle, preferably a liposome, micelle, nanoparticle aptamer (NANAP), polymeric nanoparticle.

11. Single domain antibody for use according to any one of the previous claims, wherein the diagnostic compound is one or more of- peptides for detection of protein aggregates;- imaging agent, preferably chosen from18F,68Ga,89Zr, gadolinium, ultra small iron particles, fluorescent dyes (IRDye800cw).

12. Single domain antibody for use according to any one of the previous claims, wherein the single domain antibody is fused or coupled to a protein domain that extends in vivo half-life of the single domain antibody, preferably wherein the protein domain is chosen from albumin (binding) domain, and fragment crystallizable region (Fc region).

13. Single domain antibody for use according to any one of the previous claims, wherein the single domain antibody is used in combination with radiation therapy, chemotherapy and / or one or more immunomodulating agents, wherein the immune modulating agent is preferably an immune checkpoint inhibitor.

14. Single domain antibody for use according to any one of the previous claims, wherein the single domain antibody is comprised in a pharmaceutical composition.

Images