**4. Molecule-molecule or ion-molecule interactions in active molecule design**

Non-covalent interactions are central to biological structure and function. In considering potential interactions of molecules and/or ions and their receptor, the focus has been on hydrophobic interactions, hydrogen bonding and ion pairing. Although hydrogen bonds are by far the most important interactions in biological recognition processes, the cation-π interaction is a general, strong, non-covalent binding force that occurs throughout nature, being energetically comparable or stronger than a typical hydrogen bond.

Cooperativity in multiple weak bonds (hydrogen bond and ion-π interactions among others) has been considered and studied at the MP2/6-311++ G(d,p) computational level [Alkorta et al., 2010]. Due to the presence of a great number of aromatic rings containing heteroatoms in biological systems, this effect might be important and help to understand some biological processes where the interplay between both interactions may exist.

Using Molecular Modelling to Study Interactions Between Molecules with Biological Activity 195

BE = Ecomplex – (Eion + Eligand) (2)

One of the most important aspects of ion complexation is ion selectivity, which is considered in terms of the more or less favourable binding energies. The binding energy (BE) is defined as the difference between the energy of the complex and the energy of the free ligand and

Metal ion affinity is enhanced if the host molecule has a unique conformation that is optimal for complexation, that is, with all the binding sites positioned to structurally complement

Density functional theory based on electronic structure calculations is computationally affordable. It has very good predictability power for various structural and thermodynamic properties of a molecular system, and has therefore been used to model M+-crown ether complexes [Ali et al., 2008] and collarenes acting as ionophores and receptors [Choi et al., 1998]. In both cases, the most stable equilibrium structure for complexes are estimated based on PM3 semi-empirical calculations followed by B3LYP calculations using the G-311++G(d,p) basis set of functions. Currently, the Protein Data Bank (PDB) contains over 25,000 structures that contain a metal ion. Thus, methodologies to incorporate metal ions into the AMBER force field have been developed there [Hoops et al., 1991; Reichert et al.,

Molecular modellisation of Cu(II) and Zn(II) coordination complexes has been studied by our research group, among many others. The parameters used by us were checked against a known X-ray structure and the data obtained agreed quite well with similar deviations published for other theoretical results [Miranda et al., 2005]. The models obtained were useful in explaining the differences observed among the complexes obtained in different environments. Our cation metal parameters have also been used to help in the data elucidation of coordination metal complex structures [Rodríguez-Ciria, 2000; Rodríguez-

Coordination and complexation of ions by aromatic moieties have been studied, taking into account the different characteristics of the electronic charge distribution on the aromatic frame as an addition of coulombic potentials [Albertí et al., 2010]. This type of interaction is quite common in biology as signalling in the nervous system is generally mediated by the binding of small molecules (neurotransmitters) to the appropriate receptors, which usually

An organic ammonium ion never exists as a sole cation; an anion is always associated with it. Depending on the polarity and hydrogen donor/acceptor abilities of the solvent, the association strength is different [Marcus & Hefter, 2006]. Strongly coordinating counter ions such as chloride generally lead to weaker binding constants upon recognition of the associated cation, when compared to weakly coordinating counterions such as iodide or

ion:

the metal ion [Lumetta et al., 2002].

2001; Peters et al., 2010].

Ciria et al., 2002].

contain a cationic group at physiological pH.

perchlorate [Gevorkyan et al., 2001].

Computer-assisted drug design (CADD) has contributed to the successful discovery of numerous novel enzyme inhibitors, having been used to predict the binding affinity of an inhibitor designed from a lead compound prior to synthesis [Reddy & Erion, 2005]. A free energy simulation technique known as the thermodynamic cycle perturbation (TCP) approach [Reddy et al., 2007], used together with calculations of molecular dynamics, offers a theoretically precise method to determine the binding free energy differences of related inhibitors.

Many small molecules are transported across cell membranes by large integral membrane proteins, which are referred generically as transporters. Selection among competing alternatives is always interesting and cation-π interactions are strongly involved in substrate recognition by many transporters. Drug transporters are able to carry small molecules or ions across membranes, being an important target for pharmaceutical development [Zacharias & Dougherty, 2002].

The regulation of metal ions plays a major role in enzymes, allowing to catalyse a range of biological reactions. Identification and characterisation of the metal ion binding sites and their selectivity have received immense attention over the past few decades [Ma & Dougherty, 1997]. It is evident from earlier studies that metal ions can bind to aromatic groups in a covalent as well as non-covalent fashion. Non-covalent interactions between metal ions and an aromatic ring, which are considered strong cation-aromatic interactions, are increasingly being recognised as an important binding force relevant to structural biology [Meyar et al., 2003; Elguero et al., 2009]. However, in many cases, the cation is the side chain protonated nitrogen of a basic amino acid. Reddy et al. have made available a web-based cation-aromatic database (CAD) including metal ions and basic amino acids [Reddy et al., 2007b].

Macrocyclic entities that act as ion receptors and carriers exhibit a large number of conformations in crystals and solutions, depending on the nature of their environments and of the complexed ion. To ensure the formation of the most favourable cavity for a given ion, as well as to enhance the binding and release of the ion during transport at interfaces, flexibility in the ligand structure is of utmost importance.

Complexation studies of ions with macrocycles are well documented in the literature. Some representative trends in these studies would include the following: taking into account the existence of hydrogen-bonded water molecules [Hill & Feller, 2000; Durand et al., 2000; Fantoni, 2003] and sometimes using molecular dynamics and free energy perturbation studies [Varnek et al., 1999]. The complexation phenomena have also been studied in cases where the ligand can exist as different conformers able to complex the cation [Hashimoto & Ikuta, 1999].

One of the most important aspects of ion complexation is ion selectivity, which is considered in terms of the more or less favourable binding energies. The binding energy (BE) is defined as the difference between the energy of the complex and the energy of the free ligand and ion:

194 Bioinformatics

inhibitors.

[Zacharias & Dougherty, 2002].

[Reddy et al., 2007b].

Ikuta, 1999].

[Alkorta et al., 2010]. Due to the presence of a great number of aromatic rings containing heteroatoms in biological systems, this effect might be important and help to understand

Computer-assisted drug design (CADD) has contributed to the successful discovery of numerous novel enzyme inhibitors, having been used to predict the binding affinity of an inhibitor designed from a lead compound prior to synthesis [Reddy & Erion, 2005]. A free energy simulation technique known as the thermodynamic cycle perturbation (TCP) approach [Reddy et al., 2007], used together with calculations of molecular dynamics, offers a theoretically precise method to determine the binding free energy differences of related

Many small molecules are transported across cell membranes by large integral membrane proteins, which are referred generically as transporters. Selection among competing alternatives is always interesting and cation-π interactions are strongly involved in substrate recognition by many transporters. Drug transporters are able to carry small molecules or ions across membranes, being an important target for pharmaceutical development

The regulation of metal ions plays a major role in enzymes, allowing to catalyse a range of biological reactions. Identification and characterisation of the metal ion binding sites and their selectivity have received immense attention over the past few decades [Ma & Dougherty, 1997]. It is evident from earlier studies that metal ions can bind to aromatic groups in a covalent as well as non-covalent fashion. Non-covalent interactions between metal ions and an aromatic ring, which are considered strong cation-aromatic interactions, are increasingly being recognised as an important binding force relevant to structural biology [Meyar et al., 2003; Elguero et al., 2009]. However, in many cases, the cation is the side chain protonated nitrogen of a basic amino acid. Reddy et al. have made available a web-based cation-aromatic database (CAD) including metal ions and basic amino acids

Macrocyclic entities that act as ion receptors and carriers exhibit a large number of conformations in crystals and solutions, depending on the nature of their environments and of the complexed ion. To ensure the formation of the most favourable cavity for a given ion, as well as to enhance the binding and release of the ion during transport at interfaces,

Complexation studies of ions with macrocycles are well documented in the literature. Some representative trends in these studies would include the following: taking into account the existence of hydrogen-bonded water molecules [Hill & Feller, 2000; Durand et al., 2000; Fantoni, 2003] and sometimes using molecular dynamics and free energy perturbation studies [Varnek et al., 1999]. The complexation phenomena have also been studied in cases where the ligand can exist as different conformers able to complex the cation [Hashimoto &

flexibility in the ligand structure is of utmost importance.

some biological processes where the interplay between both interactions may exist.

$$\text{BE} = \text{E}\_{\text{complex}} - (\text{E}\_{\text{long}} + \text{E}\_{\text{ligand}}) \tag{2}$$

Metal ion affinity is enhanced if the host molecule has a unique conformation that is optimal for complexation, that is, with all the binding sites positioned to structurally complement the metal ion [Lumetta et al., 2002].

Density functional theory based on electronic structure calculations is computationally affordable. It has very good predictability power for various structural and thermodynamic properties of a molecular system, and has therefore been used to model M+-crown ether complexes [Ali et al., 2008] and collarenes acting as ionophores and receptors [Choi et al., 1998]. In both cases, the most stable equilibrium structure for complexes are estimated based on PM3 semi-empirical calculations followed by B3LYP calculations using the G-311++G(d,p) basis set of functions. Currently, the Protein Data Bank (PDB) contains over 25,000 structures that contain a metal ion. Thus, methodologies to incorporate metal ions into the AMBER force field have been developed there [Hoops et al., 1991; Reichert et al., 2001; Peters et al., 2010].

Molecular modellisation of Cu(II) and Zn(II) coordination complexes has been studied by our research group, among many others. The parameters used by us were checked against a known X-ray structure and the data obtained agreed quite well with similar deviations published for other theoretical results [Miranda et al., 2005]. The models obtained were useful in explaining the differences observed among the complexes obtained in different environments. Our cation metal parameters have also been used to help in the data elucidation of coordination metal complex structures [Rodríguez-Ciria, 2000; Rodríguez-Ciria et al., 2002].

Coordination and complexation of ions by aromatic moieties have been studied, taking into account the different characteristics of the electronic charge distribution on the aromatic frame as an addition of coulombic potentials [Albertí et al., 2010]. This type of interaction is quite common in biology as signalling in the nervous system is generally mediated by the binding of small molecules (neurotransmitters) to the appropriate receptors, which usually contain a cationic group at physiological pH.

An organic ammonium ion never exists as a sole cation; an anion is always associated with it. Depending on the polarity and hydrogen donor/acceptor abilities of the solvent, the association strength is different [Marcus & Hefter, 2006]. Strongly coordinating counter ions such as chloride generally lead to weaker binding constants upon recognition of the associated cation, when compared to weakly coordinating counterions such as iodide or perchlorate [Gevorkyan et al., 2001].
