Binding of Hydrolysis-Prone Ti(IV) in Medicinal and Mineral Form to Glycoprotein Human Serum Transferrin

Abstract

Human serum transferrin (Tf) is a bilobal glycoprotein responsible for the transport and metabolism of two ferric ions. However, at only 30% saturation in human serum, reports have found that Tf can bind a variety of other metals. Titanium, as it is similar ionic size and Lewis acidity to Fe, unsurprisingly is one of these metals that can coordinate to Tf. The hydrolysis prone nature of Ti allows us to visualize features of its binding that have otherwise been unseen in previously utilized spectroscopic methods used to detail the Ti-Tf interaction. This thesis explores the binding motifs of both titanocene dichloride and titanium oxide and reveals interesting binding profiles outside of the traditional binding lobes of Tf. Titanocene dichloride (TDC) is a hydrolysis prone anticancer agent that has been shown to cause a ligand to metal charge transfer as it binds into the lobes of Tf. While UV/Vis proved the characteristic 2:1 metal to protein binding, ICP-OES detected spectroscopically mute interactions revealing higher binding equivalents were still prevalent in solution, up to approximately 50 equivalents of TDC to Tf, far past that seen in controls due to hydrolysis. Further studies showed an inability of TDC to cause a typical lobe closure of Tf as would have been seen with Fe. Lobe closure of Tf in the presence of TDC directly analyzed through protein volume SAXS measurements and denaturing urea-PAGE gels. Additionally, studies with Fe2Tf, or Fe(III) bound Tf, still showed solubilized TDC, which speaks further on the surface binding properties of Tf and possible new pathways of Ti cellular introduction. In addition to being used as drug agents, titanium has many commercial applications in cosmetics, sunscreens, and medical devices such as implants. The introduction of TiO2 into human serum and complexation with Tf is highly probable and a previously untapped area. Exploring this interaction showed not only the dissolution of Ti seen spectroscopically in the binding pocket and a higher concentration of Ti found through ICP-OES, but interesting results through transmission electron microscopy (TEM) that show the changing morphology of the TiO2 particle as it reacts with Tf over time. TEM images showed the particle sizing degrading to smaller particles over the course of 168 h. TEM was also able to detect selective uptake of nanoparticles by Tf. An equilibrium of Ti-Tf uptake was found in ICP-OES and UV/Vis results, yet the ability of Tf to change the features of TiO2 particle’s size could not have been understood fully without TEM. The formation of smaller particles and their introduction into human serum or transport with Tf could be the means of cellular accumulation of metal particles. Finally, to analyze possible surface interactions on Tf, deglycosylation techniques were implemented. Tf has two glycan chains residing on the C-terminal lobe that can be cleaved off with an enzyme PNGase F. Preliminary experiments had varied results among the binding capabilities deglycosylated Tf has with TiO2 and TDC. The differences among these two could help reveal further information about the binding lobe capacity or the site directed binding on the surface of the protein in coordination complexes versus oxides. The superstoichiometric interaction of TDC with Tf was uninterrupted by deglycosylation of Tf. However, uptake of TiO2 by Tf was affected by the removal of the glycan chains on the surface of the protein. Experiments with Fe2Tf could help decipher exact coordination and complete the telling of this story of superstoichiometric Ti binding. In our results, deglycosylated Fe2Tf did not prevent TDC interacting with the protein surface, but ICP-OES results showed far less solubilized Ti from TiO2.