Abstract
The two main aims of this thesis have been; 1. To examine the structural changes which alter the binding of the protein Intrinsic Factor to the Co corrinoid Vitamin B12 (cyanocobalamin, CN-Cbl). The alteration in binding was assessed by changes in the magnitude of the equilibrium constants for the formation of the IF-(CN-Cbl) adduct including conjugates of CN-Cbl with the enzyme horse radish peroxidase HRP and the oligopeptide calcitonin. 2. To compare the transport of IF adducts of CN-Cbl, its analogues and conjugates and of HRP across filter grown monolayers of Caco-2 cells (a cell line derived from a human colonic carcinoma which provides an in vitro model for ileal epithelial cells). The Caco-2 cells were used to investigate the feasibility of exploiting the IF-(CN-Cbl) pathway for the delivery of molecules, in particular drugs, across the ileal epithelial cell barrier. Conjugates with both HRP (MW 40K) and calcitonin (with 32 amino acid residues) were prepared by carbodiimide condensation with the carboxylic acid groups (a) produced by hydrolysis of the amide side-chain d and (b) placed on the Co as the organo-ligand Co(CH2)3CO2H. A detailed study of the conjugate of HRP and (CO2H-d-)CN-Cbl showed that the stoichiometry in the conjugate is 1:1. The enzymic activity of the conjugate is the same (98%) as that of the native enzyme.Equilibrium constants K for the binding of corrinoids by the commercially available (Sigma) porcine IF were determined at 4°C by published methods involving 57Co-labelled CN-Cbl. These data provided a quantitative measurement of binding ability and the binding curves provided a qualitative assessment. A value of K - 3 x 1010 M-1 was determined for CN-Cbl itself and was consistent over a wide concentration range, indicating similar values of K for the different isoproteins. Consistent values and simple binding curves were also observed for hydroxocobalamin (B12a) and the d monocarboxylic acid, but anomalous curves were observed for several organocobalamins, indicating slight differences (Δ log K < 2) in their binding to the different isoproteins. Little change in the value of K was produced by attaching HRP either to the β-ligand ( log K < 2), supporting previous conclusions that this ligand position is exposed to the solvent, or to side-chain d (Δ log K < 2), demonstrating that this site is also not attached to the protein. To examine the specificity of IF for the cobalamins (corrinoids possessing the full nucleotide side chain moiety as in CN-Cbl itself), values of K were determined for the corrinoid lacking the nucleotide side-chain (viz. cobinamide Cbi) alone and for the side-chain moiety (a-ribazole and ribazole-3-phosphate) alone and the two combined together. No significant binding (K < 103 M-1) was observed for Cbi a-ribazole or ribazole-3-phosphate alone, i. e. they do not competitively inhibit the binding of CN-Cbl to IF. A similar and significant degree of binding was, however, observed for Cbi in the presence of either a-ribazole or ribazole-3-phosphate (i. e. the phosphate plays no role in binding) and a more detailed study of the interaction of Cbi with a-ribazole showed that IF bound Cbi with K = 2. 5 x 105 M-1, which was dependent only on the concentration of Cbi provided that the concentration of a-ribazole was > 10-9 M. It is suggested that IF can exist in several significantly different conformations (P1 -P3) and that binding occurs in at least two major stages; that the resting P1 cannot bind Cbi but must first bind a-ribazole to form P2-Rib. , which is then able to bind Cbi to form Cbi. P3. Rib which finally loses the weakly bound a-ribazole to leave Cbi. P3 (or Cbi. P4) as the final product; and that the nucleotide side-chain plays the same initiating role in the binding of CN-Cbl. This appears to be the first detailed study of the cooperative interaction in the binding of a co-factor or ligand dissected into two (or more) segments and suggests that the specificity of IF for the cobalamins is at least partly a kinetic effect, depending on an essentially catalytic role of the a-ribazole moiety (whether free or bound in the nucleotide side-chain). Studies of the rate of transport by functionally tight and polarised monolayers of Caco-2 cells of HRP, CN-Cbl and various analogues and conjugates revealed the existence of a well-defined 'HRP' path which was saturable (attaining a maximum flux of 0. 6 fmol min-1 at a concentration of 12 nM HRP), bidirectional and unaffected by the presence of IF, in addition to the previously reported 'CN-Cbl' pathway involving the IF-(CN-Cbl) complex, which was also saturable (with a lower flux of 0. 01 fmol min-1 attained at the far lower concentration of 16 pM CN-Cbl) but unidirectional. For the HRP-Cbl conjugate linked through side-chain d it was shown that the CN-Cbl and HRP parts were transported at similar rates (i. e. probably remained linked together), that the maximum flux was similar to that of HRP and that transport was bidirectional, i. e. the HRP-Cbl conjugate was transported using the HRP pathway; The rate of flux was lowered by the addition of IF to that observed for IF-(CN-Cbl) and lowered further by the addition of IF-(CN-Cbl) in a manner consistent with simple competition for a single site favouring CN-Cbl over the HRP-Cbl conjugate by a factor of 60, close to the ratio (40) of their binding constants for IF, i. e. the IF-(HRP-Cbl) complex is transported via the 'CN-CbI' pathway. This shows that both the HRP and CN-Cbl pathways can be used for 'piggy-back' transport of other compounds (though the transport of CN-Cbl by the far larger HRP is trivial compared to the transport of the HRP protein by the Cbl-IF complex) and that the same compound (viz. the HRP-Cbl conjugate) can be switched from one pathway to the other by the presence or absence of IF.