Multiscale modeling and personalized therapies: tackling the complexities of epilepsy with degeneracy concept

By Neha MathurReviewed by Lily Ramsey, LLMMay 9 2023

In a recent article published in the Communications Biology Journal, researchers reviewed epilepsy-related degeneracy at three levels of brain organization: the cellular, network, and systems level.

They described novel multiscale and population-level models for epilepsy, accounting for different processes on a lower level that leads to an event at a higher level.

Study: Degeneracy in epilepsy: multiple routes to hyperexcitable brain circuits and their repair. Image Credit: SewCreamStudio/Shutterstock.com

Background

The multicausal and variable nature of epilepsy makes it one of the most complex brain disorders. Epilespy-induced pathological changes at all three levels interact across multiple brain regions.

Degeneracy enables structurally different biological elements to perform the same function or fetch the same output, which allows living organisms to preserve already evolved functions and concomitantly search for new ones.

At the cellular level, degeneration in epilepsy occurs across ion channels and morphological properties. Likewise, at network and system level(s), epilepsy-related degeneracy occurs across synaptic properties, seizure dynamics, and neuroimmune interactions, respectively.

Thus, despite substantial progress in epilepsy research, researchers find it challenging to disentangle the underlying mechanisms of epilepsy, which has hindered the development of effective treatments.

The human brain exhibits degeneracy where varied subcellular, cellular, and synaptic mechanisms arise from similar physiological states, which has significant implications for understanding epilepsy.

More importantly, delineating the complex interactions underlying epilepsy could help design customized multi-target therapies.

Results

The authors proposed that epilepsy is a conglomerate of multiscale disorders with the potential involvement of degenerate mechanisms at each level.

A sequencing study by Klassen et al. examined >200 ion channel genes in epilepsy patients and healthy controls. They found that healthy individuals also had multiple nonsynonymous mutations in well-recognized monogenic risk genes for epilepsy, suggesting that due to ion channel degeneracy, different channel types partially compensated for each other's defects.

Thus, clinical genetic testing often fails to detect mutations in known disease-causing genes in patients with genetic-origin epilepsies (e.g., childhood absence epilepsy).

A study of ion channel degeneracy in childhood absence epilepsy presented an interesting example of multiscale modeling, highlighting the need for customized single/multi-target pharmacological therapy.

This model demonstrated that augmenting or reducing the activation of T-type calcium (Ca²⁺) or inhibitory γ-aminobutyric acid (GABA)-A synaptic channels, alone or together, transformed physiological neural network activity to seizure-like exonerations.

Individual-level genetic variability in epileptic patients, simulated as the variability in the GABA-A and T-type calcium channel parameters, was linked to childhood absence epilepsy, and the study model presented possible justifications for the failure or triumph of pharmacological therapies directed at these parameters.

However, most importantly, this model predicted the need for multi-target treatment that simultaneously worked on both ion channels.

Another simulation study using multiscale network simulations of the spinal dorsal horn provided new insights into the degeneracy of pathological disturbances linked to hyperexcitable neurons present in epileptic patients with chronic pain.

It showed that under normal functional conditions, the same circuit activity could arise in different models with different configurations of synaptic properties.

However, following similar pathological disturbances, e.g., reduction of cell type diversity, these models exhibited heterogeneous circuit responses.

In degenerate systems, population modeling (also known as ensemble or database modeling) of neurons and neural circuits with varying parameter combinations could give rise to similar behavior.

One disadvantage of population modeling is that it fails to show which model parameter combinations from all theoretically feasible spaces occur in real brains.

However, using these models, implementing intrinsic channel and synaptic variability could help predict multi-target therapy in silico, which could help discover novel antiepileptic drug cocktails with multiple targets.

In a recent simulation study, researchers remarkably used neuronal population models to design multi-target drugs, which could salvage pathological hyperexcitability of neurons in Huntington's disease.

They referred to these hypothetical medicaments as holistic virtual drugs. In epilepsy research, a similar approach could help find therapeutically relevant disturbances of several ion channels that would switch hyperexcitable neurons to normal excitability control phenotypes.

Conclusion

Controversies about the origin of epilepsy have long hindered the development of effective treatments for epilepsy.

In this study, the authors argued that reconciling conflicting hypotheses considering brain degeneracy could manifest in multiple routes leading to similar function/dysfunction.

More importantly, they showed that computational approaches, including multiscale and population models of neurons and neural circuits, could help identify the best-customized multi-target strategies for directing the system toward a physiological state.