Molecular modelling and computational chemistry

Page of

DILRAJ KAUR SEHMI

M PHARM 3

MOLECULAR MODELLING AND COMPUTATIONAL CHEMISTRY

Molecular modelling and graphics are extremely important tools used in disciplines that require an understanding of the structure and function of molecules. The ability to visualise molecules and in turn derive important structural information from them has been a great step in science, allowing the creation of novel drugs or the better understanding of current molecules of interest.

The production of a 3-D model needs input regarding the Cartesian coordinates of individual atoms in space. These are obtained from long used procedures such as X-ray crystallography, NMR spectroscopy or mathematical manipulation. For visualisation of the model generated by this information, molecular graphics is necessary so as to display the model on a computer screen.

The basic input required for generation of the 3-D model will depend on what sort of data can be extracted from the molecule in question. If the compound conforms to the requirements for X-ray crystallography then the information obtained will pertain to the structure as a solid crystalline phase. NMR spectroscopy is generally carried out on molecules in solution and currently cannot provide 3-D structures for molecules larger than 25kDa. When one cannot get any experimental data theoretical approaches to calculation are available.

All molecular modelling techniques rely on assessing the energy of the system, either through quantum mechanics, approximates solutions to the Schrödinger wave equation or molecular mechanics which uses empirical energy terms. The usage of these methods will depend on the available computer power and storage therefore less accurate measures may be used on occasion.

Geometry optimization is a technique used for locating a stable conformation of a model using energy minimisation. Geometry optimisation procedures can be used in addition with energy minimization procedures so as to obtain favourable lower energy structures which are similar to the original structure s with regard to the potential energy. The multiple minima problem arises when one has a large molecule with quite a bit of conformational flexibility. This increased flexibility leads to many minima on the potential energy surface which is impossible to overcome in large molecules. The ability of a geometry optimization to converge to a minimum will depend on the starting geometry, the potential energy function used, and the settings for a minimum acceptable gradient between steps (convergence criteria). As a general rule, this should be preformed before performing additional computations or analysis of a model.

Considering the amazing amount of data involved in molecular modelling, molecular graphics makes it quite easy to study the molecular system allowing us to manipulate the model as seen fit so as to gain a greater understanding of it.

Aims

To generate molecular models of small drug molecules using computational approaches.
To understand the molecular properties important for the action of drugs
To display the 3-D coordinates obtained from X-ray crystallography, of a protein molecule and to study the secondary structural features

1: Generating a 2-D representation of the drug molecule.

This is a preview of the whole essay

Aims

To generate molecular models of small drug molecules using computational approaches.
To understand the molecular properties important for the action of drugs
To display the 3-D coordinates obtained from X-ray crystallography, of a protein molecule and to study the secondary structural features

1: Generating a 2-D representation of the drug molecule.

The structures of three ACE inhibitors, Succinyl praline, Captopril and Enalapril were drawn using ISIS Draw. During this exercise, the various other capabilities of ISIS Draw were also explored apart from the basic drawing features. The 2D structure shows the order of arrangement of atoms with respect to each other and gives a rudimentary idea of the 3-D structure. The structures drawn in 2-dimensional form were then exported to ViewerLite to view the 3-dimensional structures.

2: Generating a 3-D representation of the drug molecule

In ViewerLite, the geometry of the molecules is optimised using molecular mechanics and energy minimisation techniques. This will generate Cartesian co-ordinates for all the atoms in the molecule.

ViewerLite gives a better visualisation of the structures and hence, in addition one is also able to calculate bond angles and atom-atom distances. Unlike with models, the structures are not restricted to predetermined bond lengths and angles. The structure can be translated, scaled up or down and rotated for visualisation in various axes. Display options can be altered as well such as adding hydrogen atoms, viewing electrostatic potential or viewing the molecule as line, ball and stick and space filling model.

2a: Geometrical representations

A line representation is the basic 3D version of the molecule which shows the position of the atomic nuclei and the bonds between atoms. The ball and stick model also shows the position of the atomic nuclei and the bonds between atoms. In ball and stick mode, each sphere represents an atom, and the stick represents the bond between atoms, identification of atoms in ball and stick representation is easier since the balls are colour coded where each colour represents an atom of different element. The ball and stick representation does not give an overall impression of the whole shape of the molecule. It depicts the atoms as spheres of radius smaller than the Van der Waals radius which is the distance outwards from the nucleus into which the electrons extend.

The space filling model gives an overall shape of the molecule through representation of the space an atom occupies. These models are constructed by drawing each atom as a Van der Waals sphere with the atoms nucleus being the centre. One can also visualize the stearic effects of the molecule. Space filling models provide a more realistic sense of molecular size but individual bonds are not seen.

2b: Electrostatic surfaces

In reality, molecules have a surface that is accessible to the solvent and other molecules. Across this surface there is a distribution of charge arising from electron fields. The surface can be modelled by generating electrostatic surface for the molecule. The molecular electrostatic potential is the potential energy of a proton at a particular location near a molecule. Electrostatic potential correlates with dipole moment, electronegativity and partial charges. It provides a visual method to understand the relative polarity of a molecule.

According to Figure 1, which shows the surface of Enalapril, negative electrostatic potential (red), corresponds to an attraction of the proton by the concentrated electron density in the molecules Positive electrostatic potential (blue) corresponds to repulsion of the proton by the atomic nuclei in regions where low electron density exists and the nuclear charge is incompletely shielded.

When a proton approaches a positive region of the molecule, the repulsive interaction results in an increasing positive potential energy (blue), as a proton approaches a negative region an attractive interaction results in negative potential energy (red). The more red / blue differences, the more polar the molecule. If the surface is largely white or lighter color shades, the molecule is mostly non-polar.

The electrostatic surfaces of the three ACE inhibitors have similarities and differences. The similarity of all three is that they all have similar regions of positive electrostatic potential due to the common presence of a proline moiety. They all have negative electrostatic potential (red) due to presence of the electron rich carbonyl group next to nitrogen in proline and the electron rich carboxylic acid attached to one of the carbons in the proline.

However since the remaining parts of the molecules are different from each other, the electrostatic potential is different. Overall, Succinyl proline appears to be more polar due to the dark shades of red and blue and less of white as compared to Captopril. Due to presence of carboxylic group in Succinyl proline which is more polar then thiol group in Captopril, it has a lower negative electrostatic potential than Succinyl proline. Enalapril is more extended and has a benzene ring that has a pale red colored core and a blue region on the outside due to the delocalized electrons but, these tend to be generally lighter than the rest of the molecule indicating low polarity.

The active site of the enzyme must have complementary functional groups that promote the binding of the ligand to the active site via favorable interactions. It should have a hydrogen bond acceptor for the amide carboxyl group and a positive region to interact with the electron rich carboxylate oxygen or thiol sulphur in the ACE inhibitors. It should also contain another positive region for the interaction with the carboxyl terminal group.

2c: Geometry measurements. The structure of ACE inhibitor lisinopril was loaded onto ViewerLite. The dihedral angle for a bond and the associated distance between the oxygen and nitrogen atoms was observed. Varying the dihedral affected the inter-atomic distances (Graph 1). Considering the functional groups involved, the nitrogen in NH2 can acquire a proton utilizing the lone pair electrons, and thus acquire an electron deficient state -NH3+. The oxygen in the carboxylic acid group can deprotonate and become electron rich COO -. These two oppositely charged groups will undergo attraction which is more favorable an interaction than repulsion. Therefore the structure will be most energetically stable when the distance between the two atoms is the least. From the graph we can see that the most likely dihedral angle for this bond in the molecule will be 168.9o.

3: Visualising and manipulating the 3-D structure of a protein molecule produced from X-ray crystallography data

It is impossible to visualize atoms and molecules with visible light this is because in order to see an object, its size has to be at least half the wavelength of the light being used to see it. The wavelength of visible light is much bigger than an atom. X-rays, have a wavelength short enough that they can be used to visualize atoms.

The precise position of each atom in a molecule can only be determined if the molecule will form crystals, which can be grown in the lab by a variety of methods. When X-rays hit a crystallized molecule, the electrons surrounding each atom bend, or diffract, the X-ray beam, which then forms an X-ray diffraction pattern. Crystals are used because the diffraction pattern from one single molecule could be insignificant, but the many individual, identical molecules in a crystal amplify the pattern. A computer to mathematically interpret this pattern and reconstructs the positions of the atoms.

Scientists have used X-ray diffraction patterns since the early part of this century to aid their studies of molecules. In 1953, X-ray diffraction patterns from crystallized DNA were observed and it was determined for the first time that DNA molecules exhibit a double-helical structure. But it was only in the late '50s, with the advent of computers, that scientists were able to determine the precise three-dimensional atomic structure of large molecules, such as proteins and enzymes, knowing their structure allows scientists to better understand how they work and can lead to better drugs and treatments for disease.

The last part of the practical dealt with visualizing and manipulating the 3-D structure of a protein molecule produced from X-ray crystallography, the crystal structure of human angiotensin-converting enzyme in complex with lisinopril and surrounding water molecules. Different display options were available to view the structures and so get clearer visualizations without ‘crowding’.

The structure of was viewed using the ‘solid ribbon’ display option and the structure of lisinopril bound was highlighted. Figure 2 shows part of the structure with the lisinopril bound. The inter-atomic distance between O and N found was found to be 5.50 when bound. From graph 1, the dihedral angle is 149.9o which is close to the angle, predicted of 168.9o. This less stable conformation is feasible because of stabilization resulting from the binding interactions of lisinopril with the enzyme active site.

Figure 2 shows the lisinopril structure (in yellow) bound to the enzyme with the active site (yellow, also in scaled ball and stick mode). This is the secondary structure of the enzyme showing α-helices and β-pleated sheets. The red represents α-helices and the blue represents β-pleated sheets , as annotated. They appear to be anti-parallel sheets .The enzyme comprises of mainly α-helices as can be seen from the secondary structure. The α-helices seem to form a lid on the drug molecule at its active site like a ‘C’ as can be seen on the diagram. The tertiary structure of the enzyme is formed by interaction between α-helices and β-pleated sheets. The activity of the enzyme depends on the tertiary structure.

Figure 3 shows the lisinopril molecule (dark blue) and the active sites on the enzyme (yellow). Also present is a zinc active site on the enzyme where the zinc atom on lisinopril (purple) binds on an α-helix of the enzyme. The zinc atom is bound to the lisinopril molecule at the carboxylate where the Zn2+ binds.

Figure 4 shows the structure of lisinopril aligned parallel to the helix with the zinc active site and a clearer picture of the physical docking of the molecule within the enzyme. It also gives an idea of the various interactions in play. For example, the phenol ring is in contact with a hydrophobic group branching from the helix.

Conclusion

The objectives of the practical were met and the uses and applications of molecular graphics and molecular modeling was also explored via first hand manipulation of ISIS Draw and Viewer Lite. The molecule properties were analyzed using X-ray crystallography and molecular mechanics through which the geometry and energy of the molecule, lisinopril was found. This led to the appreciation of the uses and growth in the field of molecular modeling though the data about molecular structure obtained from these sources does not correspond to the real world.

REFERENCES

Molecular Electrostatic Potential