Using Brain Organoids To Model Neurodevelopmental Disorders

Post by Lila Metko

The takeaway

Neurodevelopmental Disorders (NDDs) are characterized by improper brain development and deficits in cognition and behavior. Organoids are three-dimensional cellular structures, grown from human stem cells that may provide a solution for modeling NDDs for potential human therapeutics.   

What's the science?

Research using human-induced pluripotent stem cells (iPSC) is becoming increasingly sophisticated, as it is possible to differentiate these cells into many specialized cell types. Previously, most research with iPSCs has used two-dimensional culture systems, which cannot model complex cell processes such as cell migration and high-complexity cell-cell interactions. This week in Brain, Dionne and colleagues review the ability for three-dimensional, human iPSC derived brain organoids to model many of the NDDs that are difficult to translate from animal models to humans. They discuss how organoids are not without drawbacks: genomes may be altered in the process of cell reprogramming, and there are no standardized procedures to validate the quality of new iPSC lines. Organoids are, however, becoming increasingly more advanced and are a powerful tool for studying NDDs as an alternative to animal models or clinical research in humans.  

How did they do it?

The authors investigated the role of organoids in finding treatments for several common NDDs and summarized many of the advanced experiments that have been done with organoids in recent years. Many of these NDDs have pathology that makes them difficult to model in animals. For example, microlissencephaly, a disorder characterized by lower levels of brain gyrification, is difficult to model in animals because rodent brains are normally lissencephalic. Other disorders, such as Fragile X syndrome, may necessitate organoid models because treatments that have shown high success rates in animals have proven to be untranslatable to humans. Organoids have also proved useful for disorders that are believed to develop pathology in a particular stage of cellular differentiation. Further, there are limitations to studying rodent organs because they develop differently than human organs and the developmental timeline is different. Therefore, organoids give researchers the opportunity to more accurately mimic the developmental pathways of the human brain and apply interventions at particular stages. 

What did they find?

Among the organoid models used for disorders discussed, many produced new data that could not be found in animal models. For example, in research of Fragile X Syndrome, a NDD caused by abnormal translation regulation, animal models showed that mGluR5 inhibition (excitatory) or enhancement of GABAergic signaling (inhibitory) reversed cellular and behavioral deficits. However, these therapies did not show promising results in clinical trials. Organoid models of Fragile X syndrome showed that within targets of the primary protein that is absent in the disorder, FMRP, 66% were human specific. This means that the majority of the targets that the missing protein in Fragile X Syndrome acts on are found only in humans. 

Additionally, there are promising results for using organoid models to map out a more human-like developmental timescale. Angelman Syndrome (AS), is an NDD in which a mutation causes a loss of function in a specific protein, UBE3A, leading to intellectual disabilities and seizures. Organoid models developed from patients with AS, have shown that in these specific organoids, UBE3A localizes to the cytoplasm at a different developmental stage than in a typically developing human. Another instance of organoids better mimicking the development of humans is in Tuberous Sclerosis Complex (TSC). In this disorder, cortical tubers, malformed regions within the cortex, are a major part of the disease pathology but do not develop in mouse models. Research with human-based organoids showed that TSC organoids had a different ratio of glial (supporting) cells to neurons than those in a typically developing human. 

What's the impact?

This review highlights a technology that has the potential to make huge gains in expanding our translational knowledge of neurodevelopmental disorders. These findings suggest that human iPSC organoids have already made big strides towards developing therapeutics for NDDs and will likely continue to in the future. Based on these findings, there is a high likelihood that future neuroscience research will be increasingly dependent upon organoid models as a route to finding effective therapies. 

Access the original scientific publication here.

The Nuanced Relationship Between Neuronal Activity and Blood Flow

Post by Shahin Khodaei

The takeaway

Increased neuronal activity in a brain region increases blood flow to that area. When neurons in a region are active, the signal gets sent up along the vessels that supply blood to that region, causing them to dilate upstream. 

What's the science?

When there is increased neuronal activity in a region of the brain, more blood flows to that area – a process called neurovascular coupling (NVC). This coupling is the basis for functional magnetic resonance imaging (fMRI), which measures blood flow to a brain region as a surrogate for neuronal activity. The regulation of NVC at the spatial level is not well understood – does increased neuronal activity in a small brain area lead to dilation of blood vessels in the same region? Or is the relationship more nuanced? This week in Nature Neuroscience, Martineau and colleagues addressed these questions by studying neuronal activity and blood vessel dilation in small regions of the mouse brain using microscopy.

How did they do it?

The authors used a mouse model and focused on a brain region called the sensory cortex, which is active in response to physical stimulation. Within the rodent sensory cortex, there are cortical “barrels” which become active in response to stimulation of each of the mouse’s whiskers – a cortical barrel for whisker W1, a barrel for the next whisker W2, then W3, and so on. To study the relationship between neuronal activity and blood flow in the brain of mice, the authors removed a portion of the skull directly over the sensory cortex and surgically replaced it with glass. This gave them a window through which they could study the sensory cortex, using microscopes.

The authors performed their experiments on mice whose neurons expressed a fluorescent calcium indicator, meaning that active neurons emitted red light. They then stimulated the whiskers of mice, and used wide field imaging to locate the corresponding barrel for each whisker. Simultaneously, they made use of the fact that oxygenated and de-oxygenated hemoglobin scatters the microscope’s light differently, and were able to characterize blood flow to each barrel. They also used a very high-resolution technique called two-photon microscopy to study the dilation and blood flow through individual vessels in each barrel, and how it changed due to whisker stimulation and neuronal activity.

What did they find?

As expected, when each whisker was stimulated, the corresponding barrel in the sensory cortex showed increased neuronal activity and increased blood flow. Then the authors used higher resolution imaging techniques to study blood vessel dilation in response to whisker stimulation in each barrel. They found that the response of blood vessels to was very heterogeneous: some vessels in barrel W1 dilated when whisker W1 was stimulated, some did not, and some actually dilated when whisker W2 was stimulated. Further experiments showed that blood vessels were not dilating due to increased neuronal activity in their immediate surroundings. Instead, downstream neuronal activity sent a signal up the vessel, causing dilation; meaning that blood vessels dilated in response to downstream neuronal activity. So, in the example above, a blood vessel that was imaged in barrel W1, but was in fact carrying blood toward W2, would dilate in response to neuronal activity in W2 and not W1. 

What's the impact?

This study shed light on the spatial regulation of neurovascular coupling. As the spatial resolution of imaging techniques such as fMRI increase, these findings are incredibly relevant: they suggest that at high resolutions, changes in blood vessels do not report neuronal activity of their surroundings, but instead reflect an integration of neuronal activity downstream. 

Access the original scientific publication here.

Neuroimaging Features Help Predict Treatment Outcomes for Major Depressive Disorder

Post by Meagan Marks

The takeaway

Neuroimaging data shows great potential in predicting treatment outcomes for patients with major depressive disorder, which can help clinicians choose the most effective treatment option.  

What's the science?

Major depressive disorder (MDD) is a mental health condition that is very prevalent and challenging to treat. While a handful of treatment options are available for MDD, their effectiveness varies from person to person. Clinicians currently use various clinical features to choose a treatment for a given patient, yet 30-50% of patients don’t respond well to initial treatments, leading to a trial-and-error approach where different options are tested over several weeks or months to find the most effective one. 

Recent research suggests that neuroimaging assessments – where clinicians scan the brain and analyze the data with machine learning models – may better predict which MDD treatments will work best for a particular patient. This week in Molecular Psychiatry, Long and colleagues review multiple studies to evaluate how well neuroimaging can predict treatment outcomes, which imaging techniques are most accurate, and which brain areas are most useful for prediction.

How did they do it?

To gain a more comprehensive understanding of how neuroimaging data can predict treatment success for patients with MDD, the authors conducted a meta-analysis, examining combined data from over 50 treatment-prediction studies. The authors first selected which studies to analyze based on predefined criteria. Ultimately, the authors included 13 studies on pretreatment clinical features (4,301 total patients) and 44 pretreatment neuroimaging studies (2,623 total patients). 

The authors then extracted and combined key data from each study, running a series of statistical tests to evaluate whether pretreatment clinical features such as mood-assessment scores and patient demographics, or neuroimaging features such as brain region structure and activity were better predictors of successful treatment outcomes. They also assessed which imaging modalities (resting-state fMRI, task-based fMRI, and structural MRI) most accurately predicted patient responses to electroconvulsive therapy (ECT) or antidepressant medication treatments, and which brain regions correlated to the success of these treatments. 

What did they find?

Following their analysis, the authors found that pretreatment brain-imaging features were more effective than clinical features at predicting patient responsiveness and treatment success. Specifically, resting-state fMRI demonstrated greater sensitivity to predictive variables and most accurately identified which patients were likely to benefit from particular treatments. The neuroimaging results revealed that key predictive brain regions were predominantly in the limbic system and default mode network, brain networks that are known to be involved in depression. Notably, alterations in various brain regions within the limbic network were associated with either antidepressant or ECT success, whereas brain regions within the default mode network were primarily linked to antidepressant efficacy.

What's the impact?

This study found that neuroimaging data can reliably predict which treatment options are most effective for patients with MDD, highlighting which imaging modalities and brain regions are best at estimating treatment success. This research could help clinicians accurately identify which patients are most likely to respond to specific treatments, allowing them to consider alternative options when necessary. Additionally, these findings could inspire further research into how neuroimaging might be used to predict treatment outcomes for other psychiatric conditions or diseases. 

Access the original scientific publication here