Scientists have the need to develop and choose the best biological models to understand the basis of the diseases in preclinical studies. Currently, animals and monolayer cell cultures (2D cultures) are widely spread as reference models. Anyway, these models are not completely reliable when it is necessary to comprehend some key events of the development of human pathologies. On one hand, the results of animal models may not be reproducible in humans, and, on the other hand, 2D cultures lack of the tridimensional complexity of in vivo tissues.

For example, a 2D culture of brain cells does not mimic the in vivo scattered distribution of neurons and glia, because neurons, when grown in traditional culture plates, just aggregate forming a bundle on the top of glial cells. This is why scientists are looking with increasing interest to tri-dimensional (3D) cell models. A cluster of 3D neuronal cells is called ‘neurosphere’. Ntera2 (NT2) cell line was the 3D cell model chosen by the authors of this study, whose main goal was to determine whether human neural cells grown as neurospheres presented the same biochemical pathways as the cells forming the human brain.

More concretely, the scientists obtained a neurosphere of scattered neurons and astrocytes derived from NT2 cells, and administered them two chemicals that impaired their functionality. Then, they focused on the changes that occurred in a specific biochemical pathway, the glutamine-glutamate-GABA cycle.

Glutamine-glutamate-GABA cycle can be outlined as follows:

1) neurons release glutamate or GABA (gamma-aminobutyric acid) neurotransmitters via, respectively, glutamatergic and GABAergic synapses;

2) part of the released glutamate or GABA is taken up into astrocytes, that convert glutamate to glutamine, and GABA to Succinyl CoA that enters the Krebs cycle;

3) astrocytes release glutamine, that is taken up into neurons;

4) neurons synthetize glutamate and GABA starting from glutamine.

So, glutamine is an essential precursor of glutamate and GABA neurotransmitters, and there is a tight relationship between astrocytes and neurons. But what happens to the glutamate-glutamine-GABA cycle if neurons or astrocytes are selectively impaired by cell-type specific chemicals?

To answer this question, first scientists added radioactive labeled 13C to the cell culture medium. They added [1-13C]glucose, a substrate taken up specifically by neurons, and [2-13C]acetate, taken up only by astrocytes. Nuclear magnetic resonance (NMR) spectroscopy was used to track the molecules that incorporated 13C, following them across the biochemical transformations into the cells and across their transfers between neurons and astrocytes.

Then, they impaired the metabolism of astrocytes or neurons by exposing them to a molecule with inhibitory effects on their metabolism, methionine sulfoximine (MSO) and acrylamide, respectively.

MSO is known for its inhibitory effects on glutamine synthesis in astrocytes in vivo. 3D cultures confirmed the same behavior: labeled 13C glutamine was strongly reduced in neurospheres treated with MSO compared with control neurospheres. Glutamine decrease had also a negative effect on the content of GABA in neurons, that was 50% less than control. So, GABAergic neurons depend from astrocyte glutamine for GABA synthesis. Glutamate metabolism was not affected, probably because of the activation of an alternative pathway that involves branched-chain amino acids.

Acrylamide is toxic because it impairs the axonal transport of neurons in vivo. So, as expected, 13C GABA decreased due to the decrease of the glutamine uptake by damaged neurons. Surprisingly, labeled glutamate increased, but maybe for the same reason. The collapse of the synaptic regions, in fact, may alter glutamate release from neurons, thus affecting the whole glutamine-glutamate-GABA cycle.


3D culture of neurons and astrocytes has demonstrated for the first time that typical neuronal features in vivo are reliably reproducible in vitro. This cell model represents an improvement with respect to monolayer cultures. It can be employed for further toxicological studies, thus opening the way for a better knowledge of human neural metabolism. It will also be useful for studies on pathological phenotypes impossible to perform so far.



Simão D. et al. Functional metabolic interactions of human neuron-astrocyte 3D in vitro networks. Scientific Reports | 6:33285 | DOI: 10.1038/srep33285



The elaboration of this post has been financed by the project PI15/01082, as a part of the National Plan of I+D+I and co-financed by the ISCIII – General Deputy Direction for Evaluation and Development of Health Research – and the European Regional Development Fund (ERDF).