Prediction of Allosteric Binding Modes in Homologous NMDA-Receptors

Abstract

The NMDA receptors are natural glutamate binders that play critical roles in the function of the
central nervous system, with their dysfunction implicated in serious psychiatric disorders and
neurodegenerative conditions. Understanding their roles in the brain and developing treatments
will benefit from the existence of small molecules specifically targeting different NMDA
receptor subtypes. This study has investigated allosteric binding sites of two homologous NMDA
receptors. Agonist binding domains in GluN1/N2A (crystal structure) and GluN1/N2B
(homology model) were initially subjected to Induced Fit Docking of a group of 15 analogous
ligands. Ten of these are known binders of at least one receptor. Five ligands are newly
synthesised analogues of known binders. Following Induced Fit Docking 23 ligand-receptor
complexes of type GluN1/N2A and 20 ligand-receptor complexes of GluN1/N2B were
investigated with parallel molecular dynamics (100ns) and triplicate metadynamics (20ns).
Results suggest stable binding modes for 11 ligands in GluN1/N2A and 8 ligands in GluN1/N2B.

Popular Abstract

Some vital aspects of biological systems such as ourselves cannot be observed directly. Many
relevant phenomena occur on scales that are either so small, or so large that they cannot be
assessed through experimentation. In trying to understand the intricacies of molecular
interactions or the stability of an entire ecosystem, one often needs to turn to simulations. These
controlled environments are free from the restraints of physical experimentation. Instead they
force us to grapple with the limitations of experimental models, and the limited capabilities of
our computers.
Drawing meaningful conclusions from simulated biological systems remains challenging.
Difficulty often arises because our understanding of many biological processes is incomplete,
and computer hardware lacks the power to model biological complexity in complete detail. There
are nevertheless ways of bridging biology and simulation in a way that advances both.
One way of advancing our understanding of a biological system, and approaching medical
interventions in neurodegenerative disease is by making the biochemical distinction between
related subgroups of human glutamate receptors. To functionally tell these structures apart is a
first step towards studying their roles in the brain. The primary excitatory neurotransmitter of the mammalian brain is found in the form of L-glutamate (Glu). Its functional diversity in the brain
is due to the existence of two families of receptors: the Ionotropic Glutamate Receptors (iGluRs)
and their counterpart, the Metabotropic Glutamate Receptors (mGluRs). These function in
tandem to impact the central nervous system over complex timescales and locations (Reiner and
Levitz, 2018). The normal activity of glutamate receptors in the brain has a central role in
synaptic plasticity and the encoding of memory (Morris, 2013), while dysfunction of
Glu-signalling has been linked to neurodegenerative conditions like Alzheimer's and
Huntington's disease (Lewerenz and Maher, 2015). It is known to play a role in the development
of psychiatric disorders like schizophrenia, bipolar- and major depressive disorder (Li, Yang and
Lin, 2019).
This work approaches the problem of drug discovery within two subtypes of iGluRs, via
computational methods designed to evaluate and specify their interactions with a set of small
molecules, most of which are known to interact with one or both of these receptors.
On the way towards describing, and medically targeting the differences between closely related
protein complexes, a high quality crystal structure paired with molecular simulations is a helpful
starting point. The study of molecular dynamics (MD) centres around computational models of
events at microscopic scales. These simulate interactions between atoms and calculate the
motion that follows. By the now long held insight of Newtonian mechanics, that Force equals
mass times acceleration, it is possible to relate the forces acting on every atomic particle to their
future motion. Given a close enough approximation of these forces, an evolution of a theoretical
system of particles offers insight into the properties of a real one. Forces between particles in the
system are governed by force fields, vital ingredients to these dynamic simulations. These
mathematical models describe things like Van der Waals-forces, electrostatic interactions and the
flexibility of intramolecular bonds, where the force field in question must be sufficiently accurate
to give biologically useful results (Lindorff-Larsen et al., 2012). This offers a detailed model of molecular behaviour, as the dynamics, or time-dependent
behaviour of the system, evolves under the influence of classical mechanics. As Newton's
equations of motion are resolved one step at a time, it becomes possible to track the transitions
between molecular states. Changes in positions, velocities and orientations lend insight into
real-time processes like chemical reactions and molecular binding events (Hollingsworth and
Dror, 2018).
There is a great potential in advanced computational tools like molecular dynamics and its
related methods. As available computing power continues to grow and be more efficiently
utilised (Götz et al., 2012), (Stone et al., 2016), one could expect to see methods like MD spread
into a wider range of disciplines. Aside from the growing power of simulation, another reason
for new opportunities is the increasing availability of crystal structures in neuroscience. Such
protein structures, integral to the processes taking place in the brain, present possible targets for
drug discovery (Hilger, Masureel and Kobilka, 2018), (G. Brent Dawe et al., 2016).

• author: Sellerberg, Anton Julius
• supervisor: Pär Söderhjelm; Rasmus Prætorius Clausen
• organization: Biophysical Chemistry
• alternative title: Computer-Aided Drug Discovery in Human Glutamate Receptors
• course: KFKM05 20241
• year: 2024
• type: H3 - Professional qualifications (4 Years - )
• subject: Biology and Life Sciences; Technology and Engineering
• keywords: Molecular simulations, Drug discovery, Molecular dynamics, Metadynamics, NMDA-Receptors, Computational chemistry, Biophysical chemistry, Neurochemistry, Glutamate receptors, Protein-ligand interaction, Binding mode prediction
• language: English
• id: 9178601
• date added to LUP: 2024-12-11 09:04:35
• date last changed: 2024-12-11 17:15:59