Exploring the Future of Neuroscience Drug Discovery

In this interview, Dr. Malika Bsibsi discusses the role of iPSC models in advancing neuroscience drug discovery, focusing on neurodegeneration, neuroinflammation, and innovative in vitro approaches.

Could you start by introducing yourself and giving our readers a brief overview of Charles River Laboratories’ work in neurodegenerative diseases and drug discovery? 

Hi, I’m Malika Bsibsi, Research Leader at Charles River Labs. In my role I lead a team of scientists working on the development and implementation of iPSC-derived cell models for our drug discovery clients.

Charles River Labs has a strong history in supporting neuroscience and neurodegenerative diseases across all stages of drug discovery, such as hit identification, high-throughput screening, lead optimization, in vitro and in vivo pharmacology and efficacy, pharmacokinetics, drug metabolism, and safety and toxicology.  

You’ve worked extensively on advances in cell models related to neurodegenerative diseases and neuroinflammation. Why are in vitro models like these crucial in better understanding these diseases and their mechanisms? 

The processes that cause neurodegeneration and neuroinflammation are complex and involve multiple cell types, so it can be hard to accurately model these processes and to investigate how they are involved in different diseases. Traditionally, researchers and drug developers have used animal models to do this, but there is a continuing push in the pharmaceutical community to find alternatives to animals, particularly in the early stages of drug discovery.

Also, the brains of commonly used rodent models don’t accurately reflect the complexity of the human brain. For instance, in the human brain there are over 90 billion neurons, each with around 7000 synapses to other neurons, and over a trillion glial cells. In mice for example, the ratio of glia to neurons is much lower making them a poor model for human neuroinflammation. Using human induced pluripotent stem cell (iPSC)-derived cells in culture systems enables us to provide our clients with relevant models that have high translation to human physiology and disease. 

Image Credit: Gorodenkoff/Shutterstock.com

Could you elaborate on the in vitro model for Alzheimer’s disease that you’ve developed? What potential does this model hold in drug development to treat Alzheimer’s? 

Our aim here was to develop a robust and reproducible in vitro model for Alzheimer’s drug discovery. To this end, we treated iPSC-derived glutamatergic neurons with β-amyloid aggregates to simulate amyloid pathology. Assessment of toxicity in this cell model showed that these aggregates caused degradation of neurite structures and cell death in a concentration dependent manner.

Neurodegeneration was further confirmed by release of neurofilament light chain (NfL), a key marker of neurodegeneration that has been shown to be increased in the CSF of Alzheimer’s patients. This model can be used in drug discovery to screen a group of novel compounds to determine their efficacy at preventing amyloid-induced neurodegeneration, a key pathology in the progression of Alzheimer’s disease. 

Why is investigating myelination so important in the context of neurological diseases, and what role does it play in disease progression and possible treatment options? 

Myelination is the process by which oligodendrocytes deposit myelin, a lipid-rich substance, around the axons of neurons in the brain, forming an insulting layer that increases the rate at which action potentials travel. This process is vital for neurological functions such as cognition, sensory function, and motor function, and failure or damage in the process can have devastating consequences.

In multiple sclerosis, demyelination in the brain either by autoimmune attack on myelin or failure of oligodendrocyte function leads to muscle weakness, loss of sensation and fatigue. Also there are rare, inherited demyelination disorders such as leukodystrophies and Charcot-Marie-Tooth disease. There is also evidence that demyelination and death of oligodendrocytes may be caused by amyloid and tau pathology in Alzheimer’s disease.

The process of demyelination is therefore a therapeutic target in multiple neurological disorders, and a validated in vitro model of this process such as our neuron/oligodendrocyte co-culture model, is critical to advance drug discovery in these disorders. 

Many of the models you’ve worked on employ induced pluripotent stem cells (iPSCs). What are the main advantages of using iPSCs in modeling neurodegenerative diseases and neuroinflammation? 

There are several advantages to using iPSCs and cells derived from them in drug discovery. Using human iPSCs obviously offers better translation to human diseases, compared to using animal cells for example.

An alternative source of human cells for drug discovery assays is primary cells from post-mortem samples, and while we do have access to these through our relationship with the Netherlands Brain Bank and have developed robust protocols for culture of post-mortem isolated neurons and glia, these cells are a finite resource and generally have low yield. Human iSPCs, either patient derived or commercially available cells, and cells derived from them are a robust, stable and reproducible source of cells for assays for our clients’ projects. 

How to create an iPSC-derived in vitro model to study neuroinflammation

How do in vitro cell models contribute to the drug discovery process, particularly when developing treatments for neurological diseases? 

In vitro cell models, either mono-culture or more complex co-culture models, can be used across different stages of the drug discovery process from target validation, through screening, lead identification and optimization, in vitro efficacy with phenotypic readouts, and even in early safety readouts. At Charles River labs, we use these kind of models widely in answering our clients’ drug discovery questions. 

For example, I already mentioned that we have developed a co-culture model of human iPSC-derived oligodendrocytes and glutamatergic neurons as an in vitro model of myelination. Our clients are using this model to test compounds identified from high-throughput screening, to see how they compare against each other as to their effect on myelination processes and to determine which compound would be best to proceed with. Similarly, our complex model of neurons, microglia, astrocytes and oligodendrocytes in co-culture enables our clients to screen compounds for their efficacy in targeting neuroinflammatory processes. 

We combine iPSC-derived models with our expertise in a wide range of phenotypic readouts. For example, for clients working in epilepsy drug discovery we have developed an iPSC-derived glutamatergic and GABAergic neuron co-culture that develops seizure-like activity when stimulate with seizurogenic compounds, as measured by multi-electrode array (MEA) electrophysiology. Using this model, we can screen compounds for their ability to decrease seizure-like readouts.

What are some other practical applications of in vitro cell models in the field of neurodegenerative diseases? How are these models currently being used in research and clinical settings? 

The examples I gave above explain how we can use these complex cell models in drug discovery, but they can also be widely used in the context of drug safety. For example, we use the MEA system with iPSC-derived cardiomyocytes to test the potential cardiotoxicity of novel compounds and biologics, not just in neuroscience drug discovery but across all therapeutic applications. Similarly, the complex models of neurons, microglia, astrocytes and oligodendrocytes in co-culture can be used to examine the neurotoxicity of novel compounds on a cellular level.  

Outside of the drug discovery and development process, in vitro cell models derived from stem cells are widely used in preclinical, academic research, for example to study developmental processes, and to model processes in health and disease. These models are the same as the models we use in our client projects, but rather than being used to determine compound efficacy, they are used to investigate processes and protein function, for example.  

In the clinical space, stem cells are under investigation for their therapeutic potential. For example, there are currently ongoing clinical trials looking at if iSPC-derived dopaminergic neurons, surgically transplanted into the brain, are an effective treatment for Parkinson’s disease, where symptoms are caused by degradation of dopaminergic neurons in the mid-brain. 

Looking back on your work with in vitro models, what key developments or challenges do you see in the future for using these tools in the fields of neuroscience and drug discovery? 

I think one of the key developments that is happening now with in vitro models is the development and use of 3D models, including organoids. At Charles River we are using organoids outside of neuroscience drug discovery, for example we have developed human gut organoids and validated this model for the development of antiviral compounds against infection with enteroviruses (EV-A71).

In terms of brain organoids, there are a number of challenges to be overcome before this type of model could provide a robust and reproducible model for use in drug discovery, including proportions of different cell types, vascularization, and nutrient supply. Generating brain organoids for use in drug discovery may certainly be a goal, but we also need to consider the question that is being asked and use the model that best answers that question.  

Regarding challenges with the co-culture models we have been developing, one of the key challenges is scaling these models for use in high throughput screening, to provide a translatable model at this very early stage of drug discovery.

At the moment we can culture these models in 96-well plates enabling screening, but not at the level that would truly be considered ‘high throughput’, where tens of thousands of chemical structures can be screened in an efficient and cost-effective way. We are continually refining our cell culture protocols and model set-up, and scaling these types of cell models is something we are actively working on. 

Given the translational nature of your models, how close do you think we are to bridging the gap between in vitro findings and clinical human studies, particularly in areas like neuroinflammation and neurodegeneration? 

There are so many factors that go into whether a new compound or advanced therapy will reach clinical trials, be successful in them and then go on to receive regulatory approval to proceed to treating patients. The therapy must engage with a target, be efficacious and tolerable, have an acceptable safety profile and side effects, and actually reach its target; a key consideration with neuroscience therapies as direct access to the brain is blocked by the blood brain barrier.

The complex, co-culture models and human iPSC-derived models we have been discussing allow us to gather detailed information about many aspects of a therapy’s pharmacology, before processing into animals, leading to better leads optimization and improved focus on therapies that are likely to have clinical effect. As the failure rate of neuroscience therapies is so high, any incremental improvements that can be made in the drug discovery and development process will translate to more therapies reaching patients.

Discover more: Can machine learning help with CNS behavior studies?

About Dr Malika Bsibsi

Dr Malika Bsibsi holds a PhD from the VU Amsterdam University and has spent over 25 years working in in the fields of stem cell biology, neurodegeneration and neuroinflammation.

In this time, she has developed significant expertise in culturing both iPSC-derived and primary brain cells, including neurons, microglia, astrocytes and oligodendrocytes. In her current role at Charles River, Dr Bsibsi brings this expertise to neuroscience drug discovery projects, developing mono-culture and multi-cell culture models and assay readouts, as well as providing strategy recommendations for our clients.