Modeling of Radiometal Complexation Formation with Bifunctional Coupling Agents Using Ligand-Metal Interaction Profile

Introduction

Metal-based radiopharmaceuticals are of particular interest and widely used in both therapeutic or diagnostic medicinal purposes (Bhattacharyya & Dixit, 2011). Diagnostic metal radiopharmaceuticals are used for single photon emission computed tomography (SPECT) and positron emission tomography (PET) applications as imaging agents (Bhattacharyya & Dixit, 2011; Wadas, Wong, Weisman, & Anderson, 2010). Bifunctional coupling or chelating agents (BFCs), as a fundamental critical component of a radiometal-based radiopharmaceutical, are needed for the radiolabeling of biomolecules and target-specific delivery of metal radiopharmaceutical (Liu, 2008). The metal radiopharmaceuticals must be stable enough over the time and live longer to reach its destination for the diagnostic application or therapeutic procedure (Bartholomä, 2012). A BFC ligand binds the radiometal ion in a tight stable coordination complex so that it can be properly directed to a desirable molecular target in vivo (Price & Orvig, 2014). Thermodynamic formation constants (K_ML =[ML]/[M][L]) of a complex, as a useful index of the binding strength of the complex, provide important information on relative affinities of ligands for a specific metal. An ideal BFC should be able to form a stable radiometal chelate with high thermodynamic stability and kinetic inertness (Liu, 2008).

The computer models are useful for the inspection of trends in complexation phenomena. They provide alternative, robust and computationally cheap prediction of desired property in the absence of extensive experimental or computed data (González-Díaz & Prado-Prado, 2008). The main aim of the study was to develop the quantitative structure-property/activity relationships (QSPR/QSAR) models for the analysis of the interaction space of complexation formation of radiometals and different organic ligands. Most of the previous QSAR studies on the complex formation process are based on density functional theory (DFT) descriptors (Deng et al. 2012; Hancock and Bartolotti 2005; Varbanov et al. 2012). However, DFT calculations generally involve time-consuming tasks with high computational cost. Puzyn and coworkers have demonstrated (Puzyn et al. 2008) that the relatively costly DFT calculations could not always improve the quality of the models.

We applied same fundamental principles of proteochemometrics modeling, that has been widely used for the analysis of protein-ligand interactions (van Westen, Wegner, IJzerman, van Vlijmen, & Bender, 2011; Wikberg, Lapinsh, & Prusis, 2004) to develop QSAR models. The chemical descriptors of both small ligands and radiometals ion were used in a combined model (van Westen et al., 2011). In addition to the derived descriptors of ligands and targets, cross-term or interaction-term descriptors can be introduced to the joint models which can describe the effects of both ligands and targets and the specific interactions between them (Lapinsh, Prusis, Gutcaits, Lundstedt, & Wikberg, 2001; Lapinsh, Prusis, Mutule, Mutulis, & Wikberg, 2003). Therefore, this approach can model the interaction complex or the ligand-target interaction space (Fernandez, Ahmad, & Sarai, 2010; van Westen et al., 2011). As QSAR approaches, different statistical methods, including multiple linear regression (MLR) and support vector regression (SVR) have been used to modeling of complex formation of ⁶⁷Ga(III), ⁶⁴Cu(II) and ¹¹¹In(III) radiometal ions with BCFs of diverse structures containing acyclic, hetrocyclic and macrocyclic moieties.