The aims of this thesis were to i) investigate the effect of the FH Cape Town-1 mutation on the in vitro folding of a concatemer of the amino-terminal (LB1) and second (LB2) ligand binding domain modules and ii) obtain sufficient quantities of a soluble LDL receptor that would retain the ability to bind its native ligands and may be then used to study the functional and structural characteristics of the LDL receptor in greater detail.
A concatemer of the first two LB modules of the LDL receptor (rLB1-2) and a concatemer containing the FH Cape Town-1 mutation (rLB1-2FH1) were expressed in E. coli as thrombin-cleavable fusion proteins with glutathione S-fransferase. Following affinity chromatography and thrombin cleavage, the peptides were refolded by thiol-disulphide exchange in the presence of
calcium ions and then purified by reverse phase perfusion chromatography. Mass spectral analysis suggested that the refolded isomers were frilly oxidized, with each of the 12 cysteine residues disulphide-bonded. To determine whether incorrect disulphide bonds were restricted to the mutant module or if intermodule disulphide bonds were present, the 4-amino acid linker that joins the last cysteine residue of LB1 or LB1FH1 with the first cysteine residue of LB2 was cleaved with pepsin. Digestion of rLB1-2 with pepsin yielded two fragments, with molecular masses of 4343 Da and 4932 Da. Digestion of rLB1-2FH1 with pepsin generated a number of fragments. A single, intense peak had a molecular mass of 4343 Da that was identical to the carboxylterminal fragment generated by pepsin digestion of rLB1-2. Each of
the other peaks yielded a mass of 4760 Da consistent with the amino-terminal fragment of rLB1-2FH1. This mass differed from the corresponding fragment of rLB1-2 by the combined mass of Asp26 and Gly27 It was concluded that the isomers of the amino terminal fragment of rLB1-2FH1 differ only by their disulphide bond connectivities. The absence of a fragment made up of the amino and carboxyl-terminal fragments linked by a disulphide bond demonstrates that all of the disulphide bonds of rLB1-2FH1 are formed within the modules and that no intermodule disulphide bonds exist. To confirm that the carboxyl-terminal fragment derived from rLB1-2 was identical to that obtained from pepsin digestion of rLB1-2FH1,
pepsin-digested rLB1-2 and rLB1-2FH1 were combined and then analysed by RP-HPLC. The carboxyl-terminal fragments co-eluted and had a molecular mass of 4343 Da. The fact that the carboxyl-terminal fragments eluted as a single peak provides strong evidence that they exist in the same form.
As it is difficult to recover and purify a membrane-anchored glycoprotein like the LDL receptor, a truncated soluble derivative composed of the ligand binding and EOF precursor homology domains (amino acids 1-693) was prepared. To enable purification of the soluble LDL receptor by streptavidin affinity chromatography, a fusion protein was made linking a peptide mimetic for biotin, strep-tag, onto the carboxyl-terminus. This soluble LDL receptor is referred to as LDLR693Streptag. CHO-Kl cells transfected with the plasmid directing expression of LDLR693Streptag,
secreted a protein with an apparent molecular mass of 85 kDa under non-reducing conditions. This protein bound to LDL and was recognised by a polyclonal antibody raised against LB1 of the human LDL receptor. LDLR693Streptag could not be purified by streptavidin affinity chromatography.
Further truncations were made to the soluble LDL receptor and the strep-tag was replaced with a hexa-Histidine tag. The soluble LDL receptors were composed of the ligand binding domain (LDLR291Histag), the ligand binding domain and EOF repeat A (LDLR329Histag) and the ligand binding and EGF precursor homology domains (LDLR690Histag). Soluble LDL receptors were produced in a packed bed bioreactor and were successfully purified, from serum-free conditioned medium, by immobilized metal affinity chromatography.
Like the native LDL receptor, the soluble LDL receptors
migrated with anomalous mobilities on SDS-PAGE and their apparent molecular masses increased following reduction due to the disruption of a large number of disulphide bonds. The soluble LDL receptors had acidic pIs of approximately 4.5-5. Results of SDS-PAGE, native PAGE and isoelecyric focussing indicated that there was some heterogeneity within the preparations of soluble LDL receptors. This heterogeneity may be due to differing post-translational modifications. Tunicamycin and neuraminidase treatment showed that the ligand binding domain contained N-linked carbohydrate and sialic acid. Soluble LDL receptors were recognised by the conformation-specific monoclonal antibody IgG-C7, indicating that LB1 had a similar conformation in both the native and soluble LDL receptors.
The major ligands for the LDL receptor are apolipoprotein B-100 (apoB-100) and apolipoprotein E (apoE). The native LDL receptor displays differential binding to the
three common human isoforms of apoE. ApoE3 and apoE4 bind to the LDL receptor with the same affinity however apoE2 displays <1% of the binding of apoE3 and apoE4. Using amino-terminal apoE (N-apoE) complexed to dimyristoyl phosphatidylcholine in a ligand blot assay, Ca2+-dependent binding of N-apoE3 was demonstrated. Binding of N-apoE3 was also dependent on the disulphide bonds being intact, as treatment of the soluble LDL receptors with a reducing agent abolished binding. The amount of N-apoE3 and N-apoE4 bound to the soluble LDL receptors was essentially identical, whereas binding of N-apoE2 could not be detected. These results clearly demonstrate the differential binding of soluble LDL receptors to the apoE isoforms and that the isoform specificities of the intact LDL receptor are the same for the soluble LDL receptors.
Soluble LDL receptors were shown to bind and release LDL. Ca2+-dependent binding of
LDL was demonstrated and LDL-binding was also dependent on the disulphide bonds being intact. LDL binding was also dependent on pH. At pH 5.5, the binding of LDL to the soluble LDL receptors was reduced compared to binding at pH 6.5 or pH 7.5. The ability of both acid pH or removal of calcium ions to release LDL from the soluble LDL receptors was also tested. Decrease in pH resulted in the release of approximately 1/2 of the bound LDL from the soluble LDL receptor composed of the LB and EGF precursor homology domains, consistent with the role of the EGF domain in pH-dependent ligand release. The evidence presented supports the conclusion that the soluble LDL receptors produced in this study are correctly folded and active and that the binding activity is biologically relevant.