Institute of Neurosciences and Medicine (INM)

Computational Biomedicine (INM-9)

Structural Systems Biology

Neurotransmission is a central event for brain function. As a whole, it can be viewed as a communication of information from the pre-synaptic neuron to the post-synaptic one. Signal transduction cascades triggered by chemical neurotransmitters involve a myriad of receptors and enzymes. INM-9, while keeping its main focus on molecular level descriptions of neurotransmission components (such as receptors and ion channels at the post-synaptic membrane), is making an increasing effort to move towards the systemic level. Key (and mostly unanswered) questions include: when an agonist binds to a neuroreceptor, can we predict its downstream effects, such as epigenetics changes and production of ion current carried out by ion channels? How disease-linked mutations impact on the protein interactome of the synaptic space?

Computational systems biology approaches for modeling signaling pathways and mechanistic models of pharmacokinetics, and pharmacodynamics have significantly contributed to neuroscience drug discovery and development. They have been used, for instance, to translate modes of drug action from in vitro to in vivo, as well as from animal studies to human. Through such models, it is possible to integrate and predict important quantitative and qualitative parameters, such as the drug concentration profile in blood or at the site of action, and cellular signaling downstream of the target sites. The incorporation of molecular structure parameters in such computational systems biology approaches can not only deepen our understanding of the molecular mechanisms of drug action, but also give insight into how genetic mutations can affect the subcellular downstream signaling events by altering the receptor’s ligand binding, activation, and signaling.

To start addressing these fascinating (and yet very challenging) questions, we have initiated an activity towards the integration of a molecular with a systems biology level of description of postsynaptic signaling cascades. Here, unknown key input parameters of the models may be provided by molecular dynamics (MD) and Brownian dynamics simulations. The former includes hybrid quantum mechanics/molecular mechanics (QM/MM), coarse-grain/MM calculations, and all-atom MD-based kinetics calculations. The perspective of combining such simulations and Machine learning tools, together with systems biology, will allow to connect High-Performance Computing (HPC) applications to subcellular investigations, relating molecular-level events to real biological responses.

In this context, we have developed the Structural Systems Biology (SSB) toolkit, a Python library that integrates structural macromolecular data with systems biology simulations to model signal-transduction pathways of G-protein coupled receptors (GPCRs). Our framework streamlines simulation and analysis of the mathematical models of GPCRs cellular pathways, facilitating the exploration of the signal-transduction kinetics induced by ligand-GPCR interactions: the dose-response of the ligand can be modeled, along with the corresponding change in the concentration of other signaling molecular species over time, like for instance [Ca²⁺] or [cAMP]. SSB toolkit brings to light the possibility of easily investigating the subcellular effects of ligand binding on receptor activation, even in the presence of genetic mutations, thereby enhancing our understanding of the intricate relationship between ligand-target interactions at the molecular level and the higher-level cellular and (patho-)physiological response mechanisms. Some application examples include the autism-related A218T mutation of the oxytocin receptor [DOI: 10.1038/s41380-021-01241-8 and the identification of novel antagonists of the adenosine 2 receptor [DOI: 10.1039/d3sc02352d].

We have also developed other tools:

The GPCRs Online Modeling and Docking Webserver (GOMoDo) is a G-protein coupled receptors (GPCRs) online modeling and docking web server, developed by Sandal and collaborators [DOI: 10.1371/journal.pone.0074092]. With a very easy user interface, this biocomputing platform allowed users to effortlessly model GPCR structures and dock ligands to the receptor model, obtaining biological and pharmacological relevant data, in a consistent pipeline: protein sequence alignment, homology modeling and model quality assessment, and docking. One of the novelties that GOMoDo brought to the bioinformatics community was the use of a curated local database of pre-aligned GPCR sequences, plus the possibility of directly docking the ligands into the models. However, at the time of its development and deployment only less than 3% of the human GPCR structures were available. Despite these limitations, GOMoDo is still used today, which prompted us to revamp it by creating pyGOMoDo [DOI: 10.1093/bioinformatics/btad294]. Compared to its predecessor, pyGOMoDo boasts several noteworthy improvements, such as the addition of a new model quality assessment tools, multiple options for docking programs, enhanced utilities for interaction analysis, and up-to-date internal databases. The source code is freely available at the link under the Apache 2.0 license. Tutorial notebooks containing minimal working examples can be found here.