The brain: that beautiful and complex organ that we all possess and use to control the body, and with which the body interacts in multiple ways. The embodied brain maintains - most of the time - a harmonious equilibrium that results in a robust and resilient system: a viable human being. 

However, what happens when this harmony is disturbed? What happens when the equilibrium is lost by an external attack or the “wear and tear” of parts? A viable system requires that balance be restored in some way. For instance, when the brain is exposed to external pathogens and/or toxins (e.g. drugs), the immune system reacts in an attempt to repel the attack and restore the equilibrium. Even the mere passing of time, e.g., non-pathological ageing, causes a sustained, low-grade activation of the immune system, termed “inflamm-ageing”. 

 

In the brain, the immune system is made up of two main types of glial cells, microglia and astrocytes. When threatened by external or internal noxious stimuli, they produce a stereotypical reaction called neuroinflammation. As is often the case in cell biology, this reaction is a two-sided coin. On the one hand, this activation is desirable and protective, because it helps us combat the threat. On the other hand, excessive or sustained activation of the immune system can damage the very system it is supposed to protect. 

An overreaction can be quite dangerous, with outcomes such as reduced blood and oxygen flow, ischemia, and cell death. Therefore, uncontrolled neuroinflammation has functional consequences and has been linked to the progression and development of various neurodegenerative diseases such as Parkinson's, Alzheimer's and multiple sclerosis. The ability to detect early inflammatory states may offer us a source of important biomarkers for diagnosis and treatment response in neurological and psychiatric diseases. But how can we measure the tone of the neuroimmune system in vivo? The change in morphology of glial cells when confronted with a challenge can serve as a basis. 

 

The main goal of our study has been to develop a tool that allows us to detect when glial cells are affected, non-invasively and in vivo; a tool that will eventually allow us to follow the onset and development of neuroinflammation in humans. So far, various imaging techniques have been proposed for this purpose. One of the most widely used techniques is positron emission tomography. However, it has some disadvantages - the use of radioactive markers and low resolution prevents its use in long-term studies and fragile populations. One alternative is magnetic resonance imaging (MRI), specifically diffusion MRI (dMRI). While previous literature demonstrated its sensitivity to changes in tissue microstructure, so far this technique has not been developed to the level of cell specificity necessary to differentiate between the different glial cell types. 

As the name suggests, diffusion magnetic resonance imaging uses the magnetic properties of water to measure how water molecules move within the brain tissue. In this work we combine this radiological technique with advanced mathematical modeling of how water navigates soft tissues to measure relevant parameters of different cell types. Our hypothesis was that by looking at water diffusing within different glial cells in different reactive conditions, we could infer their morphological changes and identify different inflammatory states. 

 

To test this hypothesis we used animal models. Specifically, we used several well-established models of neuroinflammation in rats through the administration of different pro-inflammatory chemicals. These chemicals, due to their different targeting and activation times, allowed us to specifically activate one glial cell type over another. This strategy provided an innovative tool to obtain unique MRI fingerprints of specific glial cell types. 

 

In our study, we scanned the brains of rats using dMRI and applied our new technique to quantify the inflammatory response. Using invasive techniques that cannot be applied in humans, we tested the accuracy of the prediction by measuring tissue markers of inflammation in the same animals and quantifying the morphological change in different cell types. The combination of both techniques in the same individual animal allowed us to refine the mathematical models that support cell specificity. The results showed that we are indeed able to image non-invasively and in vivo two different fundamental cell types involved in the neuroinflammatory response, astrocytes and microglial cells, and distinguish them. 

 

We believe that this tool, given its ability to differentiate between different brain cell types, will have a very high impact in the clinic. Its characteristics, such as the possibility to scan the brain in vivo and non-invasively, will open the door to the diagnosis, characterization and follow-up of many different and devastating diseases afflicting the nervous system.