HERO or VILLAIN: Unravelling the functions of Astrocytes

Dr. Marian Diamond stood in both shock and awe. She found it hard to believe that the only difference between Einstein’s brain and a normal human brain is the number of ‘glial cells’. They were almost four times more in Einstein’s brain, while the number of neurons remained approximately the same. “Was this just a fluke? Could these glial cells be the reason behind Einstein’s brilliance?” - the famous neuroanatomist from University of California, Berkeley wondered.

Introduction

Glial cells are non-neuronal, electrically inactive cells present in the brain. These cells were discovered in 1856, by German pathologist Rudolph Virchow. Initially, they were considered to be mere connective cells in the brain and hence the name “glia” in Latin, which literally translates to “glue”. It was only later that the varied functions of glial cells were discovered. Glial cells are further classified into - Astrocytes, Oligodendrocytes and the Micro Glia that are present in the Central Nervous System (CNS) and, Satellite cells and Schwann cells that are located in the Peripheral Nervous System (PNS).

Stained astrocytes and their star shaped structure.
Astrocytes are depicted in red and cell nuclei in blue.

Astrocytes, named after their star-shaped structures, are glial cells that mediate most of the neuro-vascular communications in the brain. Like most glial cells, astrocytes were also considered to be connective tissues before their main function was discovered. Unlike neurons which are connected to each other through ‘synapses’, astrocytes communicate among themselves with the help of calcium ion oscillations in gap junctions. They take part in a variety of tasks - signalling using calcium ions and transmitter molecules (cholesterol and thrombin), maintaining homeostasis of the brain, extracellular fluid and ions close to neurons, enhancing neural immunity and mediating neuro-vascular communications. They play an active role in ‘Immuno-defense’ by forming reactive centres where these astrocytes encircle damaged neurons (as in several neurodegenerative diseases) and ensure that the inflammations are properly shielded from the other regions of the brain.

Role of Astrocytes in Neuroimmunology

Astrocytes respond to injuries or diseases in the central nervous system (CNS) via the process - ‘reactive astrogliosis’ which is a defence mechanism to minimise and repair the initial damage. This is characterised by remarkable molecular, cellular, and functional alterations in the reactive astrocytes. CNS injuries lead to two types of reactive Astrocytes – one type being helpful (A2 reactive astrocytes) and other being harmful (A1 reactive astrocytes), Hence the dual role as a “hero” and a “villain”. Generally, they are found to remain benign until provoked by chronic neuroinflammation.
The Astrocytes become pro-inflammatory when they are exposed to certain chemicals such as LPS (Lipopolysaccharides - generally from gram-negative bacteria, an inflammatory stimulus). LPS causes activated microglia to release ROS (Reactive Oxygen Species), resulting in the formation of A1 Astrocytes, which cause the death of neurons. A1 astrocytes produce high levels of inhibitory components such as neurotoxins, particularly soluble CSPGs, (Chondroitin sulphate proteoglycans), which form chemical and physical barriers to axon elongation following CNS injuries. These CSPG deposits form a substantial barrier to regeneration and are primarily responsible for the inability of the brain and the spinal cord to repair internal damages. Soluble neurotoxins rapidly kill a subset of CNS and mature oligodendrocytes (glial cells that provide support and supplement the neurons in the CNS with myelin sheath). A1 Astrocytes are found in abundance in people affected by Alzheimer’s, Huntington’s, Parkinson’s and Amyotrophic related Sclerosis and Multiple Sclerosis. By inhibiting the formation of A1 astrocytes, the death of neurons can be prevented. Researchers around the globe are working hard to unravel the multidimensional roles of reactive astrocytes, as it can revolutionise the treatment of neurodegenerative diseases. Some mechanisms that are actively pursued are prevention of A1 astrocyte formation, promotion of A1 reversion and to block the secretion of soluble neurotoxins.

Glial fibrillary acidic protein (GFAP) tagging - used to identify astrocytes.
An injured brain has more astrocytes than a normal brain.

Astrocytes tend to become anti-inflammatory when they are deprived of oxygen resulting in the formation of A2 Astrocytes as in the case of strokes. These Astrocytes show an increase in the neurotrophic factors and ‘thrombospondins’ at the mRNA level. Thrombospondins belong to a class of proteins that induce the neurons to form synapses, thus promoting the neurons to strengthen synapses. A2 Astrocytes tend to produce substances that support neuron health, growth and survival near the stroke site.

Role of Astrocytes in Vasomodulation

Astrocytes as Neurovascular mediators

Astrocytes act as mediators between the neurons and the blood vessels, thus helping the neurons in getting essential nutrients that are important for neuronal activity. The neurons release chemicals which are known as ‘neurotransmitters’, such as glutamate in case of an excitatory neuron and GABA in case of an inhibitory neuron. Astrocytes, being receptive to these neurotransmitters, respond to them by fluctuating the calcium ion concentrations, releasing vasodilators such as EET (Epoxyeicosatrienoic Acid), which causes the blood cells to dilate and release glucose. The released glucose is then absorbed by the astrocyte which is sent to the neurons as lactate, where it gets converted into ATP.
Almost all the muscles in the human body tend to take blood directly from the blood vessels by releasing vasodilators whenever they are active. Neurons, on the contrary, communicate indirectly with blood vessels using astrocytes acting as mediators.
Let us assume a model in which the neurons are connected directly to blood vessels without the mediating Astrocyte layer (i.e.) assuming that the neurons themselves release vasodilators. Considering the case of two neurons and two blood vessels, of which one neuron is active while the other neuron is inactive and one blood vessel is dilated and the other not dilated. The brain comprises of multiple neurons that become active at varied instances of time. The activated neuron fires and the inactive ones don’t. The activated neurons have a higher energy demand and release vasodilators which causes the blood cells to dilate and release glucose to the neurons. However, these vasodilators released by the neurons tend to act on all the blood vessels, resulting in energy release to all the neurons irrespective of neural activity. This results in an imbalance between the demand and request between the neurons, where the active neurons don’t receive enough energy while inactive neurons tend to receive an equal amount of energy. Hence, a problem arises in considering the one-to-one communication between the neurons and the blood vessels.

Neurovascular interactions without Astrocytes

The presence of the intermediate Astrocyte layer helps in solving the above problem. Astrocytes receive neurotransmitters from selective neurons and then send back lactose to those specific neurons, thereby preventing the imbalance between the demand and supply.

Modelling astrocyte computationally

Through various in-vivo studies it is established that astrocytes play fundamental and multidimensional roles in the function and dysfunction of the brain. Hence, unravelling the complex and sophisticated mechanisms that govern the astrocyte functions is mandatory in understanding the brain operations and their role in CNS pathology. The progress made in technology to solve the numerous differential equations (in some cases up to 20 million equations) in multiple dimensions helped to model the neuron-astrocyte interactions.

In the above model, as shown in the flowchart, the astrocyte-neuron connections are compared to the Deep Belief Networks as proposed by Geoffrey H. Hinton, Simon Osindero and Yee-Whey Teh (2006). These connections are considered to be plastic (i.e.) these connections are assumed to constantly evolve/change with time (Although generally plasticity is assumed only between the neurons when they communicate with each other and not between the neurons and the astrocytes). Plasticity of the astrocyte-neuron connections and the astrocyte-vessel connections are considered to be of crucial importance as it enables a constant learning process for the connections. The weights which are represented by \(w, w^T, w_1, w_1^T, w_2 \) and \(w_2^T\) are then trained by using the Hebbian principle of learning. According to the Hebbian Principle, neuronal and neurovascular communications are considered to strengthen as the number of interactions between them increases. Once the weights are optimised, then the algorithm can be used on real-life examples.
Results obtained from the deep belief network, have shown that, as the number of astrocytes increases, the reproducibility of the input from the output increases. This profound result shows that neural synaptic information could potentially be encoded by the neuro-astrocyte and astrocyte-vascular weights too.

Journey ahead

Over the years, our understanding of the astrocytes has evolved from being considered as mere neuronal glue, to being acknowledged for their role in neuro-vascular communication and to recently being considered for their important role in neural computation. Thus, the perceptions about astrocytes have come a long way. Neuroscientists around the world continue to remain fascinated and enthralled at the extent of astrocyte’s significance in the information processing of our brain. The depths of its influence in neuronal computation continues to awe Scientists. Neuro-astrocyte communication has become one of the most important and sought-after regions of research in the domain of Neuroscience. Indeed, it is going to be an interesting quest.

References

1. A model of indispensability of a large glial layer in cerebrovascular circulation. Rohit Gandrakota, V. S. Chakravarthy, Ranjan K. Pradhan, Neural Computation (2010)
2. Vascular Dynamics Aid a Coupled Neurovascular Network Learn Sparse Independent Features: A Computational Model. Frontiers in Neural Circuits. Philips RT, Chhabria K, Chakravarthy VS. (2016)
3. Equations for InsP3 receptor-mediated \([Ca^{2+}]\) oscillations derived from a detailed kinetic model: a Hodgkin-Huxley like formalism. Yue-Xian Li and John Rinzel (1993)
4. Shaping inhibition: activity-dependent structural plasticity of GABAergic synapses. Carmen E. Flores and Pablo Méndez (2014)
5. Neurotoxic reactive Astrocytes are introduced by activated microglia. Nature 541, 481-487 (2017)
6. Elusive role for reactive Astrocytes in neurodegenerative diseases. Lucile Ben Haim, Maria-Angeles Carrillo-de Sauvage, Kelly Ceyzériat, Carol Escartin (2015)
7. Genomic Analysis of reactive Astrogliosis. Zamanian JL, Xu L, Foo LC, Nouri N, Zhou L, Giffard RG, Barres BA (2012)
8. Image: Stained astrocytes and their star-shaped structure. Archontia Kaminari / CC BY-SA (https://creativecommons.org/licenses/by-sa/4.0)

Author: N Sowmya Manojna
Sowmya is a third-year Biological Engineering student at the Indian Institute of Technology, Madras. She is passionate about pursuing research in the field of Computational Neuroscience, particularly interested in unravelling the function of astrocytes. She is working under the guidance of Prof. V Srinivasa Chakravarthy at IIT Madras.