This thesis investigates the use of the variable regions of novel shark antibody fragments. Sharks possess a heavy chain only antibody, known as the immunoglobulin novel antigen receptor (IgNAR). The variable domain of IgNAR (VNAR) can potentially be used as a molecular Trojan horse to deliver therapeutics to the brain. Currently, as many as 98% of therapeutics fail to reach the brain, due to the presence of the blood-brain barrier (BBB) at the level of the cerebral capillary endothelium. The BBB is a highly specialised interface that keeps the internal environment for the central nervous system (CNS) constant and provides the brain with nutrients, but prevents entry of potentially harmful substances. To overcome the problems surrounding delivery of therapeutics to the brain, VNARs have been engineered that target the Transferrin 1 receptor (TfR1), which is highly expressed on the luminal side of the capillary endothelium. The hypothesis is that by binding to the TfR1 receptor, VNARs can transcytose (pass through) brain endothelial cells to the brain parenchyma and potentially deliver therapeutic cargo. In this thesis, VNARs were selected using a functional in vitro phage display methodology, screening a library of bacteriophage (virus that infects and replicates within Bacteria and Archaea) that display VNAR on their surface, henceforth referred to as VNAR phage, for those that were transported across a hCMEC/D3 (human BBB cell line) monolayer, mimicking permeability in vivo. The hCMEC/D3 cells were grown on Transwell® filters and at 7 days post-seeding, VNAR phage were added to the apical chamber of the cells (representing the luminal side). After 1 and 3 hours, medium was collected from the basolateral chamber of the cells (representing the abluminal side), which contained transported VNAR phage. These VNAR phage were recovered from the medium and added to the apical chamber of hCMEC/D3 cells grown on Transwell® filters in the next cycle of selection. Three cycles of selection were carried out and the transported VNAR phage were then assessed for binding to mouse and human TfR1 by ELISA. 182 VNAR phage that were positive for TfR1 binding were sequenced and 30 were re-formatted as VNAR-human-Fc constructs (henceforth referred to as VNAR-Fc) in the form of plasmid DNA, which was then transfected into Expi293F™ (human embryonic kidney (HEK)) cells to produce VNAR-Fc protein. VNAR-Fc binding to mouse TfR1 and human TfR1 was tested again, by ELISA, before nine candidates were chosen for use in in vivo mouse brain uptake experiments. One VNAR-Fc was found in the brain at concentrations ~1.5-fold higher than the baseline concentration seen from the negative control (p = 0.0013 (1-way ANOVA with Dunnett’s post-hoc test)). A library of ~48,000 VNAR mutants was then created from the VNAR-Fc which showed the highest brain uptake (VNAR-Fc 8). This was achieved by mutating the complementarity determining region (CDR) 3 of the VNAR-Fc using partial overlapping mutagenesis. Mutant VNAR-Fcs were then randomly picked and assessed for TfR1 binding. Mutant VNAR-Fcs were sequenced, and 40 candidates were chosen for re-formatting as VNAR-Fc as described above. 21 that bound mouse TfR1 as VNAR phage were chosen and 19 that did not bind mouse TfR1 as VNAR phage were chosen. Both the mouse TfR1 binding and non-binding groups were used for further mouse brain uptake studies, to investigate if higher brain uptake could be achieved by mutagenesis. The amino acid sequences of all VNARs were obtained to compare which mutations may influence TfR1 binding in vitro, and brain uptake in vivo. One VNAR-Fc (VNAR-Fc 8.5) showed roughly twice as much uptake as the original parent VNAR-Fc 8 (p < 0.0001 (1-way ANOVA with Dunnett’s post-hoc test)). It was found that the amino acid residues at positions 1-6 of the CDR3 – IAQLSS (isoleucine, alanine, glutamine, leucine, serine, serine)– were the least important amino acids involved in mouse TfR1 binding as VNAR phage and VNAR-human-Fc, due to amino acids in this region being the most substituted, whilst retaining mouse TfR1 binding. It was also found that amino acid residues 14-16 of the CDR3 – RKH (arginine, histidine, lysine) – were important for mouse TfR1 binding as few mutations in this area were seen in VNAR phage and VNAR-Fcs that retained mouse TfR1 binding. However, it appeared this was not the case for in vivo mouse brain uptake, as IAQ residues were not substituted in the VNAR-Fcs with highest brain uptake, whilst RKH was substituted in some VNAR-Fcs that showed brain uptake. In conclusion, VNARs that bind mouse TfR1 and human TfR1 can be selected from a VNAR phage library using a functional in vitro assay in hCMEC/D3 cells. This method can yield VNARs that enter the brains of mice in vivo, and the mouse TfR1 binding capabilities of VNARs can be enhanced by partial overlapping mutagenesis of the CDR3 region.
Novel Shark Antibodies as CNS Delivery Vehicles
Fleckney, A. L. (Author). 1 May 2019
Student thesis: Doctoral Thesis › Doctor of Philosophy