Have you ever wondered how your brain manages to improve your cognitive skills? Or how an untrained brain differs from the one who learned a new skill? What better way to test this than with one of the most popular video games, Tetris®.

Improving cognitive skills through learning is key to adapting to an ever-faster changing modern working world. On the other hand, cognitive impairment represents a large part of the disease burden of psychiatric and neurological illnesses. Thus, a deeper understanding of the neurobiological mechanisms underlying the learning process of cognitive skills is relevant to one's everyday life and patient care.

However, many aspects are still far from understood. Specifically, it is unknown how adaptations of brain structure, network organization, and metabolic demands interact to support improvements in cognitive performance. In addition, whether behavioral advancements are driven by neuroplastic changes obtained at resting-state or during actual task performance is still unclear.

We addressed these questions in a challenging visuospatial learning paradigm involving the video game Tetris®. PET/MR scans were carried out before and after a 4-week learning period. Adaptations of network dynamics and the corresponding energy metabolism were investigated by applying novel analyses of simultaneous PET/MR neuroimaging data. We combined functional PET¹ with metabolic connectivity mapping (MCM)^2,3 to infer directional interactions across brain regions. Subsequently, simulations were performed to disentangle the role of functional connectivity and glucose metabolism in the learning process.

As a result, four weeks of learning elicited an increased top-down modulation of the salience network on the occipital cortex at rest, which was dependent on the underlying metabolic pattern. Conversely, task execution itself was characterized by a connectivity-driven decrease of the same hierarchical interaction. This divergence between resting-state and task-specific effects explained improved task performance and overall learning success, where participants with a higher difference between resting-state and task-specific effects showed better cognitive performance. We, therefore, conclude that the adaptations at rest and task are complementary and both required for successful skill learning.

These learning-induced interdependent adaptations of functional connectivity and metabolic demands across neuronal systems of different hierarchical levels suggest the formation of a 'skill engram⁴.' The development and storage of such a skill engram are metabolically expensive^5,6. However, once established, the stored engram can be easily retrieved during task execution, enabling efficient cognitive performance. We propose that the resulting improvement in task execution is realized by minimizing prediction errors between neuronal representations of brain regions on different hierarchical levels^7,8.

Thus, disentangling the role of energy metabolism and network organization at rest and during task performance offers unique insights into the plasticity of the human brain. This might be valuable for investigating numerous brain disorders, such as neurodegenerative diseases, traumatic brain injury, and psychiatric disorders.

Take home message

Using multimodal neuroimaging of the human brain before and after a 4-week learning period, we show that functional network reconfigurations during task execution rest upon metabolically engraved neuroplastic adaptations – the 'skill engram.' Crucially, these mechanisms complement each other to support improvements in cognitive performance. We thereby provide a comprehensive account of cognitive skill learning as an energy-efficient retrieval of optimized task representations.

References 

1. Hahn, A. et al. Quantification of Task-Specific Glucose Metabolism with Constant Infusion of 18F-FDG. J. Nucl. Med. Off. Publ. Soc. Nucl. Med. 57, 1933–1940 (2016).
2. Riedl, V. et al. Metabolic connectivity mapping reveals effective connectivity in the resting human brain. Proc. Natl. Acad. Sci. U. S. A. 113, 428–433 (2016).
3. Hahn, A. et al. Reconfiguration of functional brain networks and metabolic cost converge during task performance. eLife 9, e52443 (2020).
4. Josselyn, S. A. & Tonegawa, S. Memory engrams: Recalling the past and imagining the future. Science 367, 1-14 (2020).
5. Harris, J. J., Jolivet, R. & Attwell, D. Synaptic Energy Use and Supply. Neuron 75, 762–777 (2012).
6. Plaçais, P.-Y. et al. Upregulated energy metabolism in the Drosophila mushroom body is the trigger for long-term memory. Nat. Commun. 8, 15510 (2017).
7. Friston, K. A theory of cortical responses. Philos. Trans. R. Soc. B Biol. Sci. 360, 815–836 (2005).
8. Feldman, H. & Friston, K. J. Attention, Uncertainty, and Free-Energy. Front. Hum. Neurosci. 4, 1-23 (2010).